Each series in fifteen to thirty minutes, read or heard.

Invest and set up is the first shelf of the Zatra Handbooks. It takes a reader from an idea to a working business and, for a Tanzanian living abroad, to a business run from a distance. Seven handbooks cover the Mainland and Zanzibar, and each is written for a person who must make a money or business decision. Every handbook asks the same first question: is the activity open to you, and under which law? The Mainland and Zanzibar are kept apart throughout. Handbooks 1 to 4 and 9 apply Mainland law, and Handbooks 7 and 8 apply Zanzibar's own authorities. This summary explains what the series covers, which book to open first, and where the facts most often change. It is general information, not legal, tax or investment advice. Rates, thresholds and dates move, so every figure should be confirmed with the issuer named beside it before anyone acts.

What the series covers

The series has seven titles. Four apply Mainland law to a business in general: Investing in Tanzania 2026 (Handbook 1), Exporting from Tanzania 2026 (Handbook 2), The Complete Guide to TISEZA (Handbook 3) and The Post-Registration Handbook (Handbook 4). Handbook 9 applies the same discipline to a reader living abroad. Handbooks 7 and 8 cover Zanzibar, one on government fees and one on investment opportunities and the ZIPA approval route.

Read in order, the series follows the decisions a founder faces. First, is the activity open to a non-citizen? Second, which vehicle and which incentive? Third, what must be filed and paid after registration? Fourth, how are goods shipped and money moved? A reader whose plan is only in Zanzibar can start with Handbook 7. A reader living abroad can start with Handbook 9 and return to the Mainland books for the legal detail.

Each handbook rests on one principle, which the series states at the front: every figure has an issuer and a date, and a figure without a source is not printed. Several books therefore say that an amount was not found and name the office to ask, rather than offer an estimate. This protects the reader from false precision. It also means the reader must do the final check against the issuer, because the books cannot settle questions that the issuers themselves have left open.

The books run to roughly 150 to 180 pages each. Most can be used by jumping to a chapter, a quick answers section, a master ledger or a checklist. The quick answers are written so that each answer stands alone with its figure, its authority and its date, which makes them useful for a first reading and for later reference.

Start with the activity, not the company

The most important point in Investing in Tanzania 2026 is that the restriction sits on the licence, not on the company. A company registered in Mainland Tanzania may be wholly foreign-owned. But the business licensing rules reserve fifteen activities to citizens under the Order published as GN No. 487A of 2025, gazetted on 28 July 2025. Before any process starts, the planned activity should be tested against that list.

The list covers activities that many newcomers would not expect. Mobile money transfers, repair of mobile phones and electronic devices, domestic postal and parcel delivery, domestic tour guiding, museums and curio shops, and small-scale mining are all reserved. Clearing and forwarding is reserved too, because the shipping regulator requires every shareholder of an applicant to be Tanzanian. Some reservations have carve-outs. Supermarkets and specialised outlets stay open to foreign capital, salons in a hotel or for tourism remain open, and medium and large industry stays open to full foreign shareholding, because scale decides whether manufacturing is open.

The handbook also warns that a company can pass one test and fail another. The reservation uses the meaning of non-citizen in the Citizenship Act, while the fee schedule under the Business Licensing Act uses a different test, under which a company is foreign-owned if it is incorporated abroad or more than 50% held by non-citizens. A company that passes the fee test as local has not thereby answered the reservation. Handbook 2 makes a similar point for export licences: under GN 478B of 2022, one non-Tanzanian owner, shareholder, director or partner makes a company a non-citizen company for the purpose of a fish maw export licence.

For a reader, the practical sequence is simple. Confirm the activity, confirm whether a carve-out applies, and only then choose the vehicle. The establishment budget at the end of the handbook prices the steps that follow, so a plan should not reach the budget stage until the activity question has been settled in writing.

Choosing the vehicle and the tax position

Investing in Tanzania 2026 sets out the main vehicles, a subsidiary and a branch, and shows that the tax outcome is often the same. A resident company pays corporate income tax at 30% of total income, including gains on assets and shares. A subsidiary of a foreign parent pays 30% on its profit and then withholds 10% on a dividend, which gives a combined 37% on distributed profit. A branch of a non-resident company pays 30% and then a 10% charge on repatriated income, reaching the same 37% on remitted profit. The handbook concludes that the vehicle decision turns on liability, licensing and exit, not on the headline rate.

Other rates appear in the same chapter. A company newly listed on the Dar es Salaam Stock Exchange pays 25% for three years, provided at least 25% of its equity is issued to the public. A new assembler of vehicles, tractors, fishing boats or outboard engines may be offered a reduced rate of 10% for five years, and a new pharmaceutical or leather manufacturer 20% for five years, both under a government performance agreement that runs from the start of production. A loss-making company pays alternative minimum tax of 1% of turnover where its losses are unrelieved in the current and two preceding years. Each of these is conditional, and the reader should check every condition against the Income Tax Act as amended.

Debt needs care. Where 25% or more of a company's underlying ownership is non-resident, interest is deductible only up to a 7 to 3 debt-to-equity ratio. Since 1 July 2025, equity for this test includes positive retained earnings as well as paid-up capital. An investor who plans to fund a company largely through loans should model this rule before the financing is agreed, because the excess interest is not deductible.

Chapter 5 sets out the choice of vehicle. Chapter 6 covers company registration, and Chapter 7 the TISEZA certificate, which Handbook 3 tests in much more depth. The chapters should be read together, because the choice of vehicle changes which certificate, which filings and which tax treatment apply.

Incentives: what TISEZA grants and what the tax law grants

The Tanzania Investment and Special Economic Zones Authority began operating on 1 July 2025 under Act No. 6 of 2025, when the Tanzania Investment Centre and the Export Processing Zones Authority were merged. TISEZA registers investors, issues Certificates of Incentives and licenses export processing and special economic zones across the Mainland. The Complete Guide to TISEZA is the reference for anyone who needs to know what the authority can offer and how that offer is taxed.

The guide makes one argument from start to finish: an incentive is accessible only where the tax law that the revenue authority administers grants it, applied as that law words it. Of thirteen reliefs graded, five can be modelled in full. Some reliefs are confined to export income, some are only partly available, and two are named in the investment law but absent from the Income Tax Act. The headline relief, a ten-year corporate income tax exemption on export income for licensed zone enterprises, is real, but goods sold into Tanzania are chargeable at 30%, and the zone licence condition is at least 80% of production exported.

The land route also runs through TISEZA. Investing in Tanzania 2026 reads the Land Act as requiring a TISEZA certificate for a non-citizen's approved investment. Under regulation 14 of GN 466 of 2025, the stated threshold is USD 500,000 for a foreign-owned or joint-venture project and USD 50,000 for a citizen project, which the handbook prints as TZS 1,500,000,000 and TZS 150,000,000. The reader should confirm the current threshold with TISEZA before relying on it.

The costs are set out in Handbook 3. A developer licence is printed at TZS 20 million one-off, and an export enterprise licence at TZS 6 million a year. The book reports 182 projects registered and USD 1.14 billion of investment in the latest published quarter, Q1 2026. Registered project capital is not foreign direct investment, and the handbook takes care to keep the two apart. These figures move, so the reader should check them with the authority when budgeting.

Exporting: permits, customs and getting paid

Exports from Mainland Tanzania leave free of export tax except for ten charged commodities, as Exporting from Tanzania 2026 sets out. Nine sit in the Schedule to the Export Tax Act, Cap. 196, led by raw hides and skins at 80% of FOB or USD 0.52 per kilogram, whichever is higher. The tenth, raw cashew nuts, pays 15% of FOB or USD 160 per tonne under the Cashewnut Industry Act, Cap. 203. The handbook also notes a 10% export tax on crude sunflower oil. Every other commodity leaves Mainland Tanzania free of export tax and is zero-rated for VAT.

The preference a Tanzanian exporter receives depends on the origin of the goods. Goods from an EAC or SADC partner pay nothing where they meet the rules of origin. The European Union admits all products duty-free and quota-free except arms and ammunition, under Everything But Arms, and exporters self-certify origin through the REX system. The United Kingdom arrangement gives 0% on 99.8% of products. The United States AGOA access on eligible lines runs to 31 December 2026, so a plan that extends beyond that date needs a check on renewal. Japan's preference for least developed countries is stated to run until 31 March 2031. Transhipment can break origin, and the handbook explains when.

Permits are the most common cause of delay. Edible oil may leave only on a COPRA permit under GN 91 of 2026, which states the quantity and destination. Raw cashew bought in the Mtwara, Lindi and Ruvuma regions must ship through Mtwara port, and the Cashewnut Board announced that auctions for the 2026/27 season open on 19 October 2026. Buyers need the season's Board buying licence and a TMX bid security before the first session. Plants may not be exported without a phytosanitary certificate, and the Plant Health Act, Cap. 133, sets fines for a body corporate of TZS 10,000,000 to TZS 500,000,000. The reader should confirm each permit with the issuer, because the dates in the book are the dates the authorities announced.

VAT is a point of advantage for exporters. Goods exported from Mainland Tanzania are zero-rated under section 55(1) of the VAT Act, Cap. 148. The exporter keeps its input tax credit, and a refund is paid with the return where at least 90% of twelve-month turnover is zero-rated. Export proceeds have their own chapter, so the reader should plan the bank and customs steps before the goods ship.

The commodity profiles near the end of the handbook test a plan against the permit, the market and the price for a specific product. A reader with one product should go straight to that profile and then return to the general chapters for the customs and payment steps.

After registration: the monthly tax clock

The Post-Registration Handbook starts where most founders' guides stop. A certificate of incorporation confers legal personality and nothing else, and everything after it runs on a clock. A Mainland Tanzanian company must obtain a TIN within 15 days of commencing business, file changes to beneficial ownership within 30 days, register employees with NSSF and WCF, register each workplace with OSHA and hold a business licence for each activity and premises. Missing one of these produces letters, not warnings.

Three dates fall in every month, and they do not move for weekends, closed portals or quiet months. PAYE, SDL and withholding tax fall due on the 7th, VAT and DST on the 20th and excise on the 25th, each for the period before. A nil return is still due on its day. Provisional income tax is paid quarterly, and the final return is due six months after year end. The handbook identifies June as the heaviest month, when the second instalment, the final return and the year-end close fall together.

Business licences follow a fixed cycle. Under the Business Licensing Act, a licence lasts twelve months from issue and can be prepaid for up to three years. Renewal falls within 21 days after expiry and needs a TRA tax clearance certificate. After that a penalty of 25% of the fee applies, plus 2% for each further 30 days, as Investing in Tanzania 2026 records. The same handbook sets out the VAT registration thresholds: TZS 200 million of taxable turnover in twelve months, or TZS 100 million in six months, with the application made within 30 days.

Penalties are priced wherever the book can price them. The compliance calendar gives each task a named owner, and the penalty appendix shows the exposure for each missed date. The book reports that data protection and exemption misuse carry the heaviest settled penalties. Failing to comply with a data protection enforcement notice can bring a penalty notice of up to TZS 100,000,000, and the reader should confirm the figure with the authority named in the chapter. The Finance Act 2026 changed several registration duties, and the book lists them in one chapter so the reader can see what is new.

Each rate and deadline is stated as at 23 September 2026 in the 2026/27 edition. Rates, thresholds and penalties normally move on 1 July each year, and each Finance Act can change them, so the calendar should be checked against the current Act before any date is relied on.

Moving money and land from the Mainland side

Investing in Tanzania 2026 treats money and land as one question. Capital can come in and profits can go out, but the bank, the tax authority and the land registry each ask for proof. The money chapters explain how capital is brought in, how profits leave, and what is withheld on the way out. The land chapter explains who may hold land, on what terms, and how a title is obtained.

Tax at source is the step most owners forget. Handbook 9 uses a dividend withholding of 10% and a 15% withholding on a management fee as an illustrative case. The paying company deducts the tax before the money moves. In that case a dividend of TZS 60 million and a management fee of TZS 24 million give withholding of TZS 6 million and TZS 3.6 million, and the case is stated to be an illustration, not a record of an event.

Land is the restriction most likely to surprise a foreign investor. All land in Mainland Tanzania is public land vested in the President as trustee, and none of it is sold freehold. A citizen or citizen company may hold granted or derivative rights freely. A non-citizen has three routes only: a right of occupancy, a derivative right, or a partial transfer from a citizen in a joint venture, each for investment approved under the 2025 Act. A company with a non-citizen majority is foreign for land purposes, so buying control of a citizen landholding company brings its title under the same rule. On TISEZA land, a derivative right or lease typically runs for 32, 65 or 98 years.

People and buildings are part of the same timetable. Work and residence permits, labour law and payroll, and building approvals each have a chapter. The reader should plan staff, land and the licence together, because one approval often depends on another, and a delay in one moves the others.

Zanzibar is a separate law

Zanzibar has its own investment authority, its own revenue board and its own land law, and a Mainland position does not carry across. Handbooks 7 and 8 treat Zanzibar as a separate system, and neither reproduces a Mainland rate, even for comparison. A reader whose plan touches both should use the Mainland books for the Mainland side and these two for Zanzibar.

Zanzibar Government Fees 2026 is a ledger. A fee is a number only once the issuer has stated it. Each fee line carries one of three marks. CONFIRMED means the amount was read on the issuer's own page or gazette text on 19 October 2026, with the source named. SECONDARY means it was read only in a law firm, payroll or press source and is never printed as an official rate. CONFIRM WITH THE ISSUER means no amount was found, the office to ask is named, and no range or usual figure is printed in its place. Of 42 fee lines in the master ledger, 24 are confirmed, 7 are secondary and 11 have no amount found.

The three illustrative businesses are a 24-seat cafe on Unguja, a 12-room guesthouse and a taxi operator on Pemba. Their owners are described by what they run. The cafe's known government cost in its first twelve months is between TZS 13,191,000 and TZS 18,052,800, which is 16.2% to 22.1% of its sales of TZS 81,600,000, before eight lines with no amount are added. The gap comes from two open readings that the issuers have not settled: what the restaurant levy is charged on, and whether the voluntary health fund is taken.

The investment line comes first. The Zanzibar Investment Promotion Authority is engaged only when capital reaches a floor that the authority states, and the handbook reports four floors from USD 100,000 to USD 2,500,000. Against those floors the cafe stands at 14.6% and the guesthouse at 66.6% or 200%, depending on which floor the issuer applies. The only amount the book treats as certain is the USD 200 application fee. The handbook gives the example of a guesthouse that has spent TZS 520 million on its fit-out before learning of the floor, which is why the investment question should be answered first.

Account 201 covers tax registration and the fifteen-day window that follows it, and Account 202 sets out stamp duty by bands. The cafe pays TZS 204,000 a month in stamp duty, which the handbook shows is TZS 69,000 above the Fourth Schedule floor of TZS 135,000 as printed. The reader should confirm the bands with the revenue authority before budgeting, because the handbook notes that the base of some levies is still open.

Zanzibar opportunities and the ZIPA route

Zanzibar Investment Opportunities 2026 is an atlas in four folds. The first explains how to read it and which law applies. The second maps the zones on the ground. The third covers six sectors: tourism, seaweed, cloves and spice, fisheries, land and property, and light industry with energy. The fourth sets out the scorecard and the ZIPA route, and what follows approval.

The atlas stresses how much sector income swings. The seaweed export value in the source it reads fell 34.1% and then 98.1% in two quarters. Cloves fell 76.0% in one year and rose 12.6 times in the next. A guesthouse, a tour or a boat trip earns on its own occupancy and bookings, and a cold store earns only when goods keep moving through the port beside it. Each opportunity carries a loss ledger that shows what it costs if it fails, so the reader sees the downside before the upside.

Five listed zones are mapped, running from 4.56 hectares to 3,000 hectares, and the atlas says none of them carries a status that could be read in the sources. Zones named on paper are often not ready on the ground. The reader should ask ZIPA and the zone holder for the status of each site in writing, and should not treat a listing as proof of readiness.

The ZIPA route has five steps in the order the authority states: an investment intention form, a full application form, a non-refundable application fee of USD 200 paid with the complete file, supporting documents, and then the decision and certificate. The fee is paid once the file is complete, not before, because a refusal that must be lodged again may cost USD 200 again. The Regulations set 24-hour clocks for site verification and decision, and another 24 hours for the certificate after approval. Two sources read these clocks differently, so the atlas asks the holder to confirm with ZIPA in writing how and from when each is counted.

Before any deposit is paid, the atlas asks thirty diligence questions and provides a route checklist with a document folder. Each item on the checklist is marked for confirmation with ZIPA, which keeps the reader from relying on the atlas as a substitute for the authority's own list.

Investing from abroad

Invest in Tanzania From Abroad treats distance as the main risk. It organises the work around checkpoints: an arrival that establishes who the law sees, an entry that tests what a visit allows, a registration that deals with names and numbers, money in, the operations that run from afar, money out and exit. The early chapters matter most, because a person's status decides which later stops apply.

The first discipline is a walk-away line. The guide's distance budget sets out how much the owner will spend on travel, time and agents before stopping. The plan is treated as a ledger of twelve lines, and the twelfth belongs to the investor, not to a seller or a contractor. A plan that has no line for the investor's own decision is not finished.

The land checkpoint is the most practical part. A video proves that a piece of ground exists, but it says nothing about who owns it, who else claims it, or what it will cost to register. Before any payment, the reader should see the title, check who may hold the land and confirm the registration cost with the authority concerned. The rule that a non-citizen may be granted land only for approved investment applies to the diaspora investor as much as to any other foreigner.

Papers made abroad must be signed, certified and legalised before they are used in Tanzania. Each paper raises four questions: who signs it, who certifies the signature, who authenticates the certifier, and who legalises the result for Tanzania. The guide's checklist of documents from abroad, called the suitcase of papers, helps the reader take the right set. Seven numbers, each issued by a different office, must be in place before the first shilling is spent on the company, and each office asks for the number issued by the one before.

Running the investment from abroad depends on a person on the ground, whether an agent, a manager or a caretaker. Distance turns every small decision into a message and every message into a delay. The guide therefore sets a fixed cadence of week, month, quarter and year, and the cadence sheets show which filings and checks fall in each period. Books, filings and tax can be kept from abroad when they are dealt with on the same day each time.

On the way out, tax at source comes first. A declared dividend and a received dividend are different things, and a national table of declared amounts cannot show why one company's dividend stays at home. The exit checkpoints deal with sale, closure and passing the file on, including the power of attorney, the will and the heir who cannot hold the land.

Reading order and where to check

Each handbook can be read in order, but most readers need only the chapters that answer their own question. A reader with a Mainland company should begin with Investing in Tanzania 2026 for the activity, the vehicle and the tax, then turn to The Post-Registration Handbook for the monthly clock. TISEZA is needed if an incentive is part of the case, Exporting from Tanzania 2026 if goods will leave the country, and Invest in Tanzania From Abroad if the owner lives elsewhere.

A Zanzibar reader should start with Zanzibar Government Fees 2026, because it prices the full cost, and then move to Zanzibar Investment Opportunities 2026 for the sector, site and ZIPA route. Zanzibar Government Fees 2026 has a master ledger organised by authority, which lets the reader find the issuer, the mark and the account behind each line.

Check the date of every figure. The Mainland books state their positions in September and October 2026, and the Post-Registration Handbook states its law at 23 September 2026. The Zanzibar books state positions as at October 2026, with the key fee lines confirmed on 19 October 2026. Rates and thresholds move, and each Finance Act can change them. The issuer named beside each figure is the authority to confirm with before acting.

Two points apply to every reader. First, these books are general information and are not legal, tax, accounting or investment advice, and reading them creates no adviser relationship. Second, for a decision that carries real money or legal risk, the books advise taking advice from a qualified adviser before acting, and an advocate for any dispute or demand.

Figures that change most often

Some figures in this series are more likely to change than others, and they are the first to check. The VAT registration threshold of TZS 200 million in twelve months, the 30% corporation tax rate, the 10% dividend withholding, the 21-day business licence renewal window and the 25% late renewal penalty all sit in Mainland law and move with the Finance Act. The TIN and beneficial ownership deadlines of 15 and 30 days sit in the tax and company rules and should be confirmed at the start of each year.

On the investment side, the TISEZA land thresholds, the developer and export enterprise licence fees, and the ZIPA application fee and floors should be checked with the issuing authority at the time of the application. On the export side, the cashew auction dates, the COPRA permit conditions and the AGOA access period should be checked with the Board, the regulator and the United States authority respectively. Zanzibar fees are the most open of all, and the ledger's own marks show which lines are confirmed and which still need an issuer's answer.

A reader who wants a single check before acting should do three things. Confirm the figure with the issuer named in the book. Confirm the date the figure was stated. Then ask whether the figure applies to Mainland or to Zanzibar, since the series keeps the two apart and a figure from one side does not carry to the other.

Career and consulting holds two handbooks for people who need paid work now. Business Consulting Masterclass is for a student or junior adviser who must win a first paying client. Get a High-Paying Job Without a Degree is for a school leaver, diploma holder or skilled worker who wants a well paid post. Both are first editions, both are written as workbooks rather than reference texts, and both use composite cases with illustrative figures. Both apply Mainland Tanzania law only. Neither covers Zanzibar, so no Mainland rate should be carried across to Zanzibar. The consulting book's cover gives 10 October 2026 as its position date, while its rules page and the job book's rules page were read on 19 October 2026. The later reading is the one this summary follows. The summary explains what each book settles, where the figures are most likely to move, and which points need confirmation with the issuer before a decision is made. It is general information, not legal, tax or employment advice.

What the shelf covers

The shelf has two titles with a shared purpose: each takes an adult from a stated position to a paid decision within about twelve weeks. Business Consulting Masterclass, Handbook 10, is written for students, graduates and junior consultants who have knowledge but no paying client. Get a High-Paying Job Without a Degree, Handbook 14, is written for school leavers, diploma holders and skilled workers who meet job advertisements asking for a degree they do not hold.

Both books sit within the wider Zatra Handbooks series and follow its principles: one decision per book, every figure computed from inputs printed beside it, and every rule carrying an issuer and a date. Neither promises a client, a salary or a result. Both rest on composite people and businesses, described by trade and place rather than by name, and both label every worked example as illustrative.

The scope of each book is Mainland Tanzania. The consulting book mentions Zanzibar once, on its rules page. The job book states that Zanzibar has its own revenue board, labour authority and social security fund, none of which it covers. A figure from either book should not be used for Zanzibar without a separate check against the Zanzibar issuer.

Business Consulting Masterclass: structure and rules

The consulting book is a casebook of fifteen case files arranged in four books of a ledger. Book one opens the practice: the niche and the runway, registration and the tax page, and the five lines a consultant must not cross. Book two runs from the first call to the signature, covering discovery, scope, pricing, the proposal and the engagement letter. Book three covers the work itself: kickoff, change notes, diagnosis, independence from suppliers and the client report. Book four deals with the difficult client, invoicing and cash, and the handover that opens the next engagement.

The book follows one practice through three composite practitioners whose engagements differ in size. The smallest is a final year student taking jobs of TZS 150,000 to TZS 600,000. A former microfinance analyst takes TZS 900,000 to TZS 3,000,000, and a two person firm takes TZS 9,000,000 to TZS 20,000,000. These sizes are inputs chosen for the worked examples. The book states that no market fee for any consulting service could be verified at source, so no market fee is printed.

Two rules run through all fifteen files. The first is that no work starts until the scope, the price and the consultant's own limits are written down and both sides have signed. The second is that commission from a supplier on a recommendation is never taken unseen. A third matter is a limit rather than a rule: a consultant does not audit, practise as an accountant, act as a licensed tax consultant, prepare legal instruments for a fee or do professional engineering work. Ten client documents at the back are supplied for adaptation. They are a proposal, an engagement letter, a scoping questionnaire, a diagnostic report skeleton, a fee quote, a kickoff agenda, a status report, an invoice cover note, a handover memo and a referral letter.

Opening the practice: the niche and the runway

The first case file asks how long savings last before a first fee arrives. The worked practice opens with TZS 4,500,000 of savings, living costs of TZS 900,000 a month and practice costs of TZS 420,000 a month, made up of six lines. The monthly burn of TZS 1,320,000 gives a runway of 3.4 months. The planning figure is three whole months, because the TZS 540,000 left over cannot carry a fourth month. In the worked case, a quarter spent searching without a sale uses 88.0 percent of the savings.

A niche is chosen by a weighted score. Need carries 30 points of 100, reach 25, proof 20, margin 15 and fit 10. Each candidate is scored from 1 to 5 on each criterion, so the best possible total is 500. In the worked case three candidates score 375, 350 and 290. A runway test then removes any niche that leaves less than one month of burn when its first fee arrives. The days to first cash are stated assumptions of 60, 90 and 120 days for the three candidates, and the book says they must be replaced by observed days once calls begin. At those inputs only the fastest candidate passes.

The surviving niche is written as one sentence with five parts: who is served, where, what is delivered, in what form and in how long. The sentence is then read aloud to someone outside the trade. The book warns against naming a sector, such as small businesses or financial management, because a sector invites comparison with every adviser in the district. A student with no case history can still sell a bounded task, a written method and a list of exclusions.

Registration and the tax page

Registration follows a fixed order: the business name, the Taxpayer Identification Number, a bank account kept apart from personal money, and a written answer on VAT. A single owner or partnership applies for a business name through BRELA's online system, and the portal states that the certificate follows within eight working hours. A company route needs name clearance, a memorandum and articles, and forms 14a and 14b. The memorandum is drafted by an advocate. The portal prints no fees, and the book prints none either, so each fee is to be confirmed with BRELA. A business licence follows from BRELA or the municipality, depending on the activity and the place. The issuer for a consultant working from a rented desk is marked to confirm.

The TIN must be applied for within 15 days of commencing an activity. The only source for this period is a law firm summary of the Finance Act 2026, so the book treats it as a working rule to be confirmed with the Tanzania Revenue Authority. The reading of commencing that the book adopts is the earliest plausible day, which is the day paid work is agreed, because a later reading could leave the period already expired. The book shows three possible readings for one worked practice. Their deadlines fall 69 days apart, so the earliest reading is the safe one.

VAT is charged at a standard rate of 18 percent. The TRA page gives registration thresholds of TZS 200 million of turnover in twelve months or TZS 100 million in six months. It also states that the turnover threshold does not apply to professional service providers. The pages do not say whether a business consultant is one, so the book recommends a written question to TRA before the first invoice, stating the activity in the words the invoice will use. For a worked fee of TZS 400,000, the consultant keeps TZS 360,000 under either VAT answer. A registered consultant, however, collects TZS 72,000 of VAT that belongs to the authority, and must not treat that sum as income.

Clients withhold tax of 10 percent on payments for resident professional services. The withheld tax is non-final and is paid by the client within seven days after the end of the month. The book assumes that the withholding base is the fee before VAT, which the pages read did not settle, so the base is on the confirm list. The consultant never handles the withheld money but must record each deduction by client and date, and check it against the withholding certificate. The refund or carry forward of an unused credit, and the period for keeping records, are not stated in the pages read and are also marked to confirm.

Provisional income tax for a resident individual is paid in instalments on the last day of the third, sixth, ninth and twelfth months. An estimate of the year's tax is due within three months of the start of the year, and the balance is due within six months after year end. For 2027 the instalment dates are 31 March, 30 June, 30 September and 31 December, and the balance falls on 30 June 2028. The book prints no presumptive tax figure. A Finance Act 2026 summary reports that the presumptive ceiling for resident individuals rose from TZS 100 million to TZS 200 million at 4 percent. Whether a business consultant sits inside that regime is on the confirm list, and whether a nil estimate must be filed is also open.

Where consulting stops: the five lines

The consulting book draws five lines that a consultant must not cross in Mainland Tanzania. Each rests on a statute or regulator's page read on 19 October 2026. The first is accounting and audit. Under the Accountants and Auditors (Registration) Act, Cap. 286, the practice of accounting and auditing is reserved to registered persons, so the consultant does not audit or certify accounts and refers that work to a firm registered with the National Board of Accountants and Auditors. The second is tax consultancy, where licensed tax consultants are registered and unlicensed practice or use of the title is an offence. The third is legal instruments, where the Advocates Act, Cap. 341, reserves certain instruments to advocates. The fourth is engineering, where only a registered engineer may do professional engineering work under the Engineers Registration Act, Cap. 63. The fifth is personal data, where the Personal Data Protection Act, 2022 applies.

The book applies a five question test in order before any quote is written. The first asks whether the paper will say that figures are true, fair or audited for others to rely on. The second asks whether the work will file, sign or argue anything with the tax authority. The third asks whether the paper will create or change a right or duty between the client and anyone else. The fourth asks whether the work will size, design or approve anything built, installed or loaded. The fifth asks whether the work needs names, numbers or records of staff or customers. A yes to any of the first four sends the request to another profession. A yes to the fifth keeps the task, with its fields, purpose, reader and deletion date written into the scope.

The worked printing press case shows that a request which looks like five refusals can hold a large share of paid work. Of five asks, four are referred to other professionals, and the consultant keeps five days of work: two days to tidy records into a pack marked unaudited, one and a half days for a monthly sales sheet, half a day for a list of commercial wishes and one day for a customer count with no names or numbers. At the worked floor rate of TZS 215,000 a day, the retained work is TZS 1,075,000. Words also matter. A card or proposal should describe the work as business consulting, records and costing, and should avoid accountant, tax consultant, engineer, certified or approved, because those words imply a registered signature.

The book is candid about the limits of the test. The Advocates Act text was not read in full, so the list of instruments and the penalties are not printed. The Accountants and Auditors Act text used is a consolidated version that does not reflect a 2021 amendment the book notes. The personal data rules rest on law firm summaries and not on the primary text. Where the text is incomplete, the line is drawn at the safest reading, and the question is put to the issuer in writing.

From first call to signed scope and price

The first call is timed at forty minutes. The worked call, held in an event hall in Bukoba, moves from facts to opinions, defines success before money is mentioned and ends with a promise small enough to keep. Nothing is priced and no advice is given on the call, because free advice carries no scope.

A scope turns wishes into three parts. A wish is a feeling about an outcome, a deliverable is something that can be handed over on a date, and an exclusion is work the client may assume but will not receive. In the worked case, seven wishes are cut into deliverables, exclusions and acceptance tests. The book treats the exclusions column as the one most often missing and the one that protects both sides. Acceptance is defined by criteria, a response period and the consequence of silence. Each assumption is stated with its consequence.

Price is built from three numbers set before the client is met. The floor is the lowest day rate at which one month of the practice meets its need. In the worked case the month must produce TZS 1,500,000 of pay, and practice costs of TZS 420,000 bring the need to TZS 1,920,000. With nine billable days in about twenty two working days, the raw rate is TZS 213,333, and the floor is TZS 215,000, rounded up to the next TZS 5,000. The plan is then counted in half days with a buffer. The quote in the worked case is 11.5 days at the floor, rounded up to TZS 2,500,000, which is TZS 217,391 a day. A rival quote of TZS 900,000 for the same scope has no stated basis and is treated as an input, not a market price. The walk-away figure governs which scope may be cut, while the rate is held.

The proposal and the engagement letter do the selling and the protecting. The proposal carries the situation, the work, the plan, the fee and the next step in about five pages. The engagement letter fixes ten clauses, each tied to a dispute it prevents, and carries two signatures. Work begins only when both documents are on file. The worked ice plant letter is signed by the client before work starts and is kept with the scope.

The work itself: kickoff, change and diagnosis

The first day converts a signed letter into dated duties. The kickoff agenda runs for ninety minutes and ends with a data request in which every item has a return date. In the worked driving school diagnostic, the plan runs from Monday 1 February to Thursday 11 March 2027, with 18 planned half days and a buffer of five. The report is due in the fifth week and the presentation in the sixth.

Extra requests are common and often fair, so the method is to price them rather than refuse them. A change note names the request, the days it needs, the fee at the rate in the letter and the effect on the date. Work starts only when the approver signs. In the worked honey packers case, a costing job of three days at TZS 400,000 grew by a change note adding two days at TZS 300,000, bringing the fee to TZS 700,000 for 7.5 days. A margin note adds that the rate on the original scope is the figure a change note cannot reach.

Diagnosis starts from records, not from a theory. In the worked driving school case, two records kept by different people, and never added together, showed 624 lesson hours offered and 392 given. The gap of 232 idle hours is then ranked by cause, with no-shows and unsold mornings explaining 140 of the 232. Each cause carries an hours figure, a test that could prove it wrong, a record and a named owner. A cause is accepted only when a record and a second source agree. Hours and effort are ranked separately, because the largest cause can be the cheapest to fix.

Independence is protected by a fixed rule rather than by disclosure. In the worked ice plant case, the regional agent of a refrigeration supplier offered an 8 percent commission on a sale that the consultant's report would recommend. The book argues that disclosing a conflict tends to backfire, because a listener may trust the adviser more and discount the advice less. The rule it prescribes is to write the commission policy in advance, show it to every client, and take no commission that the client has not seen.

The client report is judged by use. The worked report carries findings, evidence, action, a named person and a date on one sheet, with the weak points marked. The first draft's main finding, that the plant should consider expanding its cooling capacity, was rejected because nothing in it could be disagreed with. The long report stays for reference, and any gain is shown as a scenario with a half case beside it.

Money, difficult clients and the next engagement

Most difficult clients are ordinary clients who meet a gap in the paper. The book's remedy is to close each gap in writing before it costs money: a payment date, a deposit, a dated duty list and a clause on independence. Where the paper is silent, the consultant states the rule once, in writing, and then keeps it.

The invoice is followed to the bank. In the worked ice plant case, the first invoice is TZS 6,796,800. This is a milestone fee of TZS 5,760,000 plus VAT at 18 percent, which is TZS 1,036,800. Withholding at 10 percent of the fee is TZS 576,000, so the cash received is TZS 6,220,800. The book states that the withholding base is assumed to be before VAT, and that a different base would change the cash received and the figures in the invoice cover note. A separate case prices a payment 47 days late, including six hours of chasing.

A first year is kept in one ledger dated to the shilling. The worked consultant's fees total about TZS 15,000,000 in the year, and the ledger shows the four provisional tax instalments and the withholding credits against them. Closing an engagement is work, not courtesy. The handover memo records what was delivered, returns or deletes client data, and names the next owner of each open item. The referral letter names a category of professional and the register to check. It takes no payment for an introduction, because a commission on a referral is a separate matter that needs an advocate's advice before any arrangement is made.

Business Consulting Masterclass: points to confirm

Several points in the consulting book are marked as open, and each should be confirmed with the named issuer before acting. The TIN period of 15 days from commencing an activity is a secondary reading of the Finance Act 2026 and is to be confirmed with TRA. Whether a business consultant is a professional service provider for VAT, and from which date registration would apply, is not stated in the TRA pages read. The 10 percent withholding base, whether before or after VAT, and the refund or carry forward of any excess credit are also open.

Other open points include whether a nil provisional tax estimate must be filed and on what form, whether a business consultant sits within the presumptive regime, which body issues the business licence for a practice without its own premises, how long records must be kept, and whether the consultant must register with the Personal Data Protection Commission. The boundaries of the advocate's and tax consultant's work are also to be confirmed with the issuer before any task touches a return or a legal instrument.

Get a High-Paying Job Without a Degree: structure and assumptions

The job book is a workbook meant to be used with a pencil, a phone calendar and one evening a week. It has seven legs and fourteen route cards. Each card opens on a document, a voice message or a scene, explains one idea in a table and ends on a sheet. The sheets are gathered at the back so that they can be copied and filled in on paper. Its six candidates work in six towns, including a warehouse hand in Tanga, a waitress in Moshi and a hardware counter assistant in Songea, with starting pay from TZS 380,000 to TZS 600,000 a month in the stated cases. The sequence follows one candidate through his first ninety days in January 2027.

The book promises a method and no outcome: no job, salary or date. Its working line of TZS 900,000 gross a month is a book convention. It is stated to be 2.3 times the TZS 390,992 mean monthly income from paid employment in the 2020/21 labour force survey, and the reader sets a personal line that can be moved up or down.

Three rules run through the cards. A fee for a vacancy, an interview, a list or a place is a warning sign, and paying an officer to secure a post is corruption, which no card suggests. Employment law is stated only as far as the rules list reaches, and anything beyond it is marked to confirm with the issuer. The cards on the pay slip and on negotiation are education and not financial advice.

The skills ledger and the pay slip

The skills ledger asks twenty statements in five groups: things done, kept, sold, fixed and explained. Each statement is scored 0, 1 or 2, for a total out of 40. A total from 0 to 13 starts with collecting evidence, 14 to 27 starts with the route cards and the pay slip, and 28 to 40 starts with the employer and interview tools. A low score is not a low value. It means the evidence has not been written down yet.

A pay slip is a short sum in a fixed order: gross pay first, then the employee's pension contribution, then tax on what is left, then the amount that reaches the bank. The employer pays a further layer on top. In the worked case of a gross of TZS 900,000, the employee's NSSF share at 10 percent is TZS 90,000. Taxable pay is TZS 810,000, PAYE on the monthly bands is TZS 80,500, and net pay to the bank is TZS 729,500, or 81.1 percent of gross.

The monthly bands used in the book are nil on the first TZS 270,000, 8 percent to TZS 520,000, 20 percent to TZS 760,000, 25 percent to TZS 1,000,000 and 30 percent above that. The book works PAYE on pay after the employee's NSSF share. A table at ten pay levels shows the share reaching the bank falling from 88.5 percent at TZS 380,000 to 70.2 percent at TZS 2,400,000. A rise in gross therefore adds less to the bank than to the offer, which is why offers are compared in net.

Routes into a salary and the choice of employer

Six routes can carry a worker without a degree above the working line. The book compares them by entry cost, months to the first pay step and the stated gross target. Trades and installation costs TZS 300,000 to enter and reaches a first step in nine months. Logistics and clearing costs TZS 250,000 and takes twelve months. Sales and accounts costs TZS 100,000 and takes four months. Hospitality front of house costs TZS 150,000 and takes eight months. Tech support and networks costs TZS 600,000 and takes six months. Records, office and public service costs TZS 400,000 and takes ten months.

Each route is a trade-off between what the ledger already shows, what the entry proof costs and how many months pass before pay changes. Waiting has a price. In the logistics case, each month before the first step forgoes the stated net gain, so three months of delay forgo TZS 1,711,020. The book reports payback in one to three months if the stated gain arrives, and it says the wait before the first step is longer and less certain than the entry cost.

Once a route is chosen, the employer is scored on five criteria: pay structure, ladder, training, lawful basics and need fit. The weights are 25, 25, 20, 20 and 10, so the total is out of 500. A gap of 25 points or more decides the order. A smaller gap is a tie, broken first by a re-weighting test and then by the employer the reader can check more thoroughly. In the worked case, three employers score 355, 285 and 365. The employer with the highest monthly offer, at TZS 1,100,000, scores lowest because its training and lawful basics are weak. Under every set of weights tried, that employer never leads. The lowest offer, at TZS 850,000, leads on ladder and training, and the book notes that a ladder is a promise until someone has stood on it.

Proof, contact and the interview

A certificate is a claim made by someone else, in writing, with a date. It carries value to an employer who cannot watch the work first. The book sets out the National Technical Award ladder from NTA 4, the Basic Technician Certificate, to NTA 10, the Doctorate, under GN No. 223 of 2005. NTA 4 and NTA 5 are the certificate rungs, and NTA 6, the Ordinary Diploma, sits one step below the Higher Diploma at NTA 7. For experienced workers the book describes a prior learning route through VETA. It cites a stated government target of 80,000 artisans reported in a press article of 23 April 2026, a figure to confirm with VETA before it is relied on.

A one page portfolio carries three checkable proofs and counts rather than a list of duties. The book rewrites a line such as responsible for stock as a count, an object and a named referee. Studies cited in the book, one from South Africa and one from Uganda, found that skills which could be shown credibly and certified were valued more by employers. The book notes that both studies come from other countries and that no Tanzanian certificate fee or pay premium was verified.

Contact is planned as twenty four messages over twelve weeks, about two a week, split across three channels. Each message names a need, reaches a named person and asks a small question with a date. In the worked case, a first week of nine general applications produced no replies, because no message named a need and no message reached a person. The funnel then shows the rates: 9 replies from 24 contacts, 5 interviews from 9 replies, and 2 offers from 5 interviews. At those rates twelve contacts give one offer, and twenty four are planned so that a choice of two offers is possible.

The interview is treated as a work sample. The book's four tests carry 100 marks: a work sample worth 40, a need story worth 20, a reliability record worth 20 and questions asked worth 20. The work sample takes six of the ten hours of preparation. The cheapest marks, from the reliability record and the questions, are prepared first. The panel is given something it can inspect before anything is said about the candidate.

The offer, the pay talk and the first ninety days

An offer letter is a short list of promises. The book reads each line as a question with a stated answer, puts unclear points to the employer in writing and runs the pay through a bad month as well as a good one. Nothing is signed until the particulars are on paper. The employer's cost is a second figure. In the worked case a gross of TZS 900,000 costs the employer TZS 1,026,000 a month, a multiple of 1.14. The layers are employer NSSF at 10 percent, the workers' compensation levy at 0.5 percent and the skills and development levy at 3.5 percent. Each rate is to be confirmed with its issuer.

A pay talk is one exact question with a number, a reason and a floor set beforehand. In the worked hotel case, a monthly figure of TZS 850,000 for a front desk post was met with a counter of TZS 1,000,000 and then settled at TZS 920,000, which is TZS 70,000 above the first figure. A written review at six months was added to the margin, and both sides signed the note. The method does not promise a rise. It makes the asking exact.

A minimum wage is a floor to check, not a figure to quote. The book refers to the Labour Institutions (Minimum Wage for Private Sector) Order, GN No. 605A of 2025, which took effect on 1 January 2026 and sets rates by sector. The book prints no single figure, because the sources read disagreed on it. A pay offer below the floor for the sector is a matter to raise with the ministry responsible for labour, not to accept.

Once pay is settled, the ninety day log begins. It is a daily page of five lines kept in pencil, recording arrival and departure times, one thing learned, one piece of work done with a number attached, and what is owed. The log is put before the manager on three review dates, and a written next step is requested each time. In the worked case the first entry is dated Monday 4 January 2027. At the end of the period, four moves are open: stay, ask for a raise, switch employer or build work of the reader's own. Each has a test drawn from the log and the folder, and a move is taken only when its own test passes.

Get a High-Paying Job: points to confirm

Several points in the job book are open. The minimum wage for each sector must be read from the ministry's order, not from the book. The press and a law firm summary gave different figures and different structures, so none is printed. The split of the 20 percent NSSF contribution between employer and employee is not stated on the page read, so the book assumes 10 percent each, and the reader confirms the split with NSSF. The workers' compensation levy of 0.5 percent, paid by the employer, and the skills and development levy of 3.5 percent, which applies to employers with ten or more employees, are to be confirmed with each issuer.

On employment law, the Employment and Labour Relations Act, Cap. 366, is read as consolidated to 31 December 2023. The Labour Laws (Amendments) Act No. 4 of 2025 amended section 14 on contract categories and section 40 on remedies. A law firm summary lists no change to sections 15, 29, 31, 35 or 37. The first year's leave entitlement is to be confirmed with the issuer, and the six month threshold for unfair termination is shown as a planning date only. The PAYE bands are the TRA monthly schedule read on 19 October 2026, and the arithmetic is the book's own.

The book prints no pay scale, no certificate fee and no market salary. Its figures are worked examples computed from stated inputs, and the employer and pay figures are a model, not a benchmark. Zanzibar's position is outside the book altogether.

Reading order and where to check

A reader with a client on the telephone this week should start with the five lines and the first call, then move to the scope and the price. A reader with no niche yet should start with the ledger page and the runway. In the job book, the skills ledger score chooses the first card, and a low score points to the pay slip and the route choice before anything else.

Across both books the rule is the same. Each rate, threshold and date should be confirmed with the named issuer before acting, and the written reply should be kept with the registers or the file. For a decision that carries real money or legal risk, a qualified adviser, an advocate or a licensed tax consultant should be consulted first. The figures most likely to move are the TIN and VAT positions, the withholding base, the provisional tax and presumptive rules, the employer contribution rates, the minimum wage and the fees of each register. The job book's tools may be copied for personal use with the source named, but may not be sold, bundled or re-branded.

Family and business is the Zatra Handbooks shelf for owners who want their household and their enterprise to run on the same written rules. It holds three titles, and each takes one question. The Value of a Good Wife asks how a couple can plan earning, building and serving inside one household budget. Raise Owners, Not Employees asks how children and staff come to run the systems that produce income. Wealth That Lasts Four Generations asks how a family can teach later generations to keep what was built. The shelf shares one habit: the rules and figures are written down before a decision is taken. Its scope is Mainland Tanzania. Zanzibar has its own authorities and its own laws, and none of the three titles applies them. This summary is general information, not legal, tax or family advice. Figures are illustrative unless the source states otherwise, and every rule should be confirmed with the issuer named beside it before anyone acts.

Three titles and one method

Each handbook on this shelf rests on the same five commitments. It takes one decision per book, computes every figure from the inputs printed beside it, names an issuer and a date for every rule, labels illustrative cases as examples, and ends with tools that can be filled in at a desk. The method is deliberately plain. A reader who wants to check a sum can do it with a pencil, and a reader who wants to dispute a rule can see where it came from.

The first and second titles state their positions at 10 October 2026. All three read their rules on Monday 19 October 2026, the date that governs any rule quoted from the shelf. Neither bank, lender, mobile money operator, school nor provider is recommended. Where such institutions appear, they appear as categories, and any product, fee or rate is left to the issuer.

The Value of a Good Wife: one budget, written by both

The first title reads Proverbs 31:10 to 31 in the King James Version as a portrait rather than a checklist. Its fifteen readings are built on the poem's twenty-two verses, and each line is treated as a composite that many women together might draw, not as a set of duties that any one woman must meet. The reader is a working adult of roughly 22 to 55, and the book is written for two people who read it together. Because the poem is framed as a mother's teaching to a son, each reading asks what a line requires of the husband and the household as well as of the wife.

The book is careful with survey figures. It reports that in Tanzania women spend 4.6 hours a day on unpaid domestic and care work and men 1.2 hours, a difference of 3.7 times, drawn from the 2020/21 Integrated Labour Force Survey. These averages are used as inputs to a calculation and never as a target for any couple. They describe a population. A household's own hours are what the two-day clock records.

The working tool is the household budget. The Lushoto couple in the second reading has a joint income of TZS 1,130,000 a month, made up of TZS 650,000 and TZS 480,000. Of this, TZS 185,000 is already promised to parents, to a younger sister's school fees and to a tool loan, leaving TZS 645,000 for home costs such as rent and TZS 300,000 to share. The book then tests an assumed new promise of TZS 50,000 a month. The promised share of income rises from 16.4 percent to 20.8 percent, and the shared remainder falls from TZS 300,000 to TZS 250,000. The rule is one of order rather than generosity: a new promise is made with the page open, and the line it displaces is named before it is agreed.

Three further sheets complete the working set. The two-day clock asks each person to write a working day and a rest day alone, in quarter hours, before the sheets are compared. The twelve first Sundays are a monthly conversation of about 45 minutes, starting on 1 November 2026 and ending on 3 October 2027, each closing with one written decision. The family business role sheet records who does what, who decides, who is paid and whose name stands on which paper. The book recommends that the clock, the budget and the role sheet be redone twice a year, with the old copies kept beside the new.

The law the book reads is limited to Mainland Tanzania. Under section 114 of the Law of Marriage Act, 1971, as quoted by the Court of Appeal in Yesse Mrisho v Sania Abdul (7 November 2019), a court may divide assets acquired during marriage by joint efforts, weighing each party's contribution in money, property or work. Section 60(b), as quoted in the same judgment, presumes equal beneficial interests in property held in joint names, and the presumption can be rebutted. In Bi Hawa Mohamed v Ally Seif (29 November 1983), the Court of Appeal is read as treating domestic work as a possible contribution, with its extent deciding the share. The book does not treat that one case as a rule for other facts. In Hadija Issa Arerary v Tanzania Postal Bank (11 May 2020), a mortgagor must disclose marital status, and the lender must take reasonable steps to verify it. The hours in the two-day clock are a household record that can help show the extent of work. They are not a legal measure.

The book also states the rule on paid maternity leave: 84 days, and 100 days for multiple births, under section 34 of the Employment and Labour Relations Act, Cap. 366, as read through a law-firm summary. A leave beginning on Monday 4 January 2027 would run to Sunday 28 March 2027 for 84 days and to Tuesday 13 April 2027 for 100 days, counting the first day as day one. These are planning dates, not a statement of entitlement. Several other rules were not read for this edition, including sections 56, 58, 59 and 63 of the Law of Marriage Act, the Land Act and Village Land Act, and voluntary NSSF membership. The book prints nothing about them, and the reader confirms them with the issuer or an advocate.

Where a reader fears a partner, the book directs help before any conversation guide. The preface and reading 11 point to where that help can be found, and the couple should use it first.

Raise Owners, Not Employees: fifteen stages from age 0 to 18

The second title is written as an almanac. The reader opens it at the page for the child's age, and the neighbouring stages can wait. There are fifteen stages in five bands of years, followed by a stage for staff that uses the same ladder at any age. Each stage settles four things: what the child does, what the parent or guardian does, what the money looks like in shillings, and where law or safety draws a line. The reader is any parent, guardian or teacher of a child with income. The book does not expect every child to become a business owner, and it counts a young adult who manages a salary well and reads a contract before signing as a good outcome.

Four households carry the figures, each in a Mainland town. The Kyaruzi household in Bukoba lives on TZS 450,000 a month, the Sulle household in Babati on TZS 1,500,000, the Mabula household in Kahama on TZS 2,400,000 and the Kapufi household in Sumbawanga on TZS 9,000,000. The Kapufi household runs a welding and furniture workshop with three staff. All four households and their money are illustrative.

Three kinds of money are kept apart, and so are the three kinds of work behind them. Duty is what a member of the household does because the household is shared, and it is never paid. A paid extra is work beyond duty at home that a child may decline without penalty, and never at the cost of school, homework or rest. Work for an employer outside the home is permitted by law from fourteen only if it is light. A child's own project at home belongs to the child and is run with a parent present.

Pocket money follows a convention of a fixed sum for each year of age, each week: TZS 100 in the Kyaruzi household, TZS 150 in the Sulle, TZS 250 in the Mabula and TZS 600 in the Kapufi. Over a year these come to TZS 88,400, TZS 171,600, TZS 299,000 and TZS 904,800. Set against household income, the book reports them as 1.64, 0.95, 1.04 and 0.84 percent. It treats 2 percent as a ceiling above which pocket money begins to compete with a household's own needs. This is the book's own test, not a statutory or statistical standard.

A weekly sum is divided by purpose. For a child of nine in the Sulle household, TZS 1,350 divides into TZS 675 to spend, TZS 405 to save and TZS 270 to give. The savings share drives the timing: on the book's stated inputs, raising the save share from 30 to 60 percent cuts the wait for a TZS 23,000 balance from 57 weeks to 29 weeks.

The book sets firm limits. It does not ask a child under fourteen to work for anyone outside the home, to sell alone to the public, to carry cash alone at night or to deal with an adult buyer the family does not know. It does not suggest a loan from outside the family, a loan app or a bet, because the minimum ages for these are not confirmed in its sources. A small loan between family members, written down with a date and repaid, is treated as a household matter. No activity is allowed to displace school attendance, homework or sleep.

The age lines come from the Law of the Child Act, Cap. 13 R.E. 2019 and the Employment and Labour Relations Act, Cap. 366. A child is a person below eighteen under section 4(1) of the Child Act. Light work is permitted from fourteen under section 77, provided it is not harmful and does not affect school attendance. Work of more than six hours a day is treated as exploitative under section 78(3). Night work from 20:00 to 06:00 is barred under section 79 of the Child Act and section 21 of the Employment and Labour Relations Act. Hazardous work, including mining, quarrying and machines, is barred under section 82 of the Child Act and section 5(3) of the Employment and Labour Relations Act. Apprenticeship is open from fourteen or after primary school under section 88. Competence to contract depends on the age of majority under the law that applies, and the book reads this with the Child Act's definition, so an agreement signed by a person below the age of majority may be void, and the reader confirms the position with the issuer and a lawyer before treating any such signature as binding. The book also summarises the 2020 SIM card registration regulations, under which a parent registers a minor's SIM and the holder re-registers in his or her own name at majority. That summary is secondary and must be confirmed with the communications regulator, TCRA.

The same ladder is used at the workshop. Adults learn ownership the way children do, through small decisions with a written limit and a record, and the book calls this the staff stair. The last child stage closes on Yohana's eighteenth birthday, Friday 12 March 2027. The handover that day is a list of ten items closing a custody ledger of TZS 1,340,000, which moves responsibility from the parent to the young adult. The book treats the eighteenth birthday as a legal deadline, because promises held by the parent must now be held by the owner.

Wealth That Lasts Four Generations: the daftari and the five ledgers

The third title is written as a daftari, a notebook kept by hand, and it is organised as eighteen folios across four hands. Its method asks for twenty rules and one test for each: could the next hand run the rule without asking. It is written for the working adult in Mainland Tanzania, roughly 25 to 55, who holds something others will need, such as a farm, a plot, a shop, a salary or a savings tin. The book promises a method and a set of sheets. It promises no result.

The method counts five ledgers at each hand-over. Land is counted in acres, or in square metres for a plot. Skills are counted as the number held by two or more people, because a skill held by one person alone is a risk rather than a possession. Name is the number of outside parties who deal on the family's word. Trust is the number of disputes open beyond 90 days, and savings are counted as months of reserve. The measures are simple so that a clerk or a child can count them and the next hand can count them again.

The book's illustrative family, the Mdegela family, is counted in 1962, 1997, 2019 and 2026. Land goes from 6 acres to 8 and stays there. Skills held by two or more rise from 1 to 4, then 7 and 11. Outside parties rise from 3 to 9, 17 and 22. Open disputes read 0, 0, 1 and 0, and reserves read 3, 5, 4 and 6 months. The four hands are making (1962 to 1998), governing (1998 to 2019), keeping (2019 to 2026) and teaching, which concerns the fifteen children and those who follow. The letter of hand-over written in 1997 recorded the farm and the town plot but left out the trees, which only show when something goes wrong. The second hand began after a death and six years of unspoken rules, and its first meeting, with written minutes, set those rules down.

The central act of the second hand is the cut. In 2006 a sum of TZS 18,000,000 arrived for the five owners, and the family wrote its rule before the money came: 30 percent to reinvest, 30 percent to learning and 40 percent to the owners. Each owner received TZS 1,440,000, against TZS 3,600,000 under an equal split. The book is candid that the rule cost each owner TZS 2,160,000 on the day. It was chosen because an equal split would have left two acres of bare land and no money to replant them. A loan of 2011 pledged 40 percent of a plot's value, TZS 12,000,000 of TZS 30,000,000, against a cap of 25 percent, which would have stopped the loan at TZS 7,500,000. In 2016 the family moved from acres to shares, creating a company of 1,000 units so that each branch held a stake it could hold, pass on and count. The governance checklist is drawn from Ostrom's design principles for shared resources, and each rule carries its date, the decision behind it, its cost and any change.

The third hand audits the saying that wealth does not last beyond three generations. The 70 percent figure often attached to it is traced to John Ward's 1987 study of 200 family manufacturing businesses in Illinois, which measured whether majority ownership passed to the next generation. The 70 percent figure is the same fact as Ward's 30 percent continuity rate, not a second finding. Grubman's 2022 review found no independent source for it. Baron and Lachenauer (2021) compared the lifespans of family firms and reached the opposite reading of the same business world. A panel study by Pfeffer and Killewald (2018) followed United States families and found that grandparents' wealth predicts grandchildren's wealth beyond what parents' wealth explains, largely through education and homeownership. The book says this finding is American and applies it to nothing in Tanzania. Its conclusion for the family is that the saying is a warning to test, not a verdict to fear or to dismiss.

The third hand also chooses a manager by scorecard and term. The manager's seat falls open every four years, in 2019, 2023 and 2027. A written card turns a quarrel about people into an argument about weights, which can be read, tested and changed.

The fourth hand creates a fund for every child. On the book's stated inputs, a fund started at birth reaches between TZS 4,680,000 and TZS 97,200,000 by the child's eighteenth birthday. Grants are tied to skills at 14, 18 and 25 rather than paid as birthday gifts, because a skill that two outsiders have seen teaches more. Sixteen of the twenty rules carry a lapse date unless they are re-adopted, and all twenty are read every five years, at review dates set for 13 December 2031, 13 December 2036 and 14 December 2041. A rule that nobody can explain is dropped.

The law the book reads for Mainland Tanzania covers the Trustees' Incorporation Act, Cap. 318, section 2(1), on bodies set up for religious, educational, charitable and similar purposes; the Land Act, Cap. 113 R.E. 2019, sections 159 to 166, on joint occupancy; the Probate and Administration of Estates Act, Cap. 352, read in part only; and section 114 of the Law of Marriage Act, which the first title also uses. Wills, intestacy, estate administration and title registration are left to Handbooks 32 and 37, and the buying and holding of land to Handbook 28. The company rules were not read at source. The book says a private company has at least two and at most fifty members, and that figure is to be confirmed with BRELA or an advocate. Whether a family trust qualifies for incorporation under Cap. 318 is a question for an advocate.

Zanzibar and the limits of the shelf

None of the three titles covers Zanzibar. The Value of a Good Wife states that Zanzibar has its own authorities. Raise Owners, Not Employees states that Zanzibar has its own law for children and for employment and is not covered. Wealth That Lasts Four Generations states that Zanzibar is not covered. A reader in Zanzibar should not apply the Mainland age lines, the pocket money conventions, the marriage rules or the company rules in this shelf to a Zanzibar business or family, and should take Zanzibar advice before using any figure or rule. This summary keeps the two jurisdictions apart for the same reason.

The shelf also states what it will not do. It does not recommend a bank, lender, school or provider, and it prints no product, fee or rate. Where a rule is uncertain, the text says to confirm with the issuer, and the reader is advised to take the safer side. Every case is an illustration, and the book is education rather than legal, tax or family advice.

Where to start and in what order

The Value of a Good Wife is best started with its two-day clock and its first budget, which the book itself describes as the two sheets that make every later reading easier. Each person fills in the clock alone before the sheets are compared. The couple then writes one budget with every outside promise listed, each with a name, an amount and an end date, and sets the buffer line last. The readings on the household's work and the role sheet follow, and the first Sunday conversation starts on 1 November 2026.

Raise Owners, Not Employees is opened at the age of the child, not at the front. Its stage finder lists each stage by age. A parent with a nine-year-old opens Stage 5 and the sheet that goes with it, and leaves the other stages until they are needed. The fridge-door sheets are copied for use at home, and the words to say at each age are given in English and Kiswahili.

Wealth That Lasts Four Generations is read once, in folio order, because each folio builds on the one before. After that, the back of the book is used by year: the constitution, the succession calendar, the letter form, the Daftari Sheets and the twenty rules. A reader who holds family land or a company should read the loan and company folios before drafting any family rule, since the pledge cap and the share structure decide what the rule can do.

The three titles can be read together. The household budget in the first is the family-level counterpart of the ledgers in the third, and the child stages in the second train the young adults who will later receive the handover the third describes. Each title carries its own legal reading, however, and a rule from one should not be carried into another without checking it with the issuer.

Figures and rules to confirm before acting

The most likely points to change are those with a date or an issuer. The age lines for children and light work, the night-work window and the hazardous-work bars are taken from the Child Act and the Employment and Labour Relations Act as read in October 2026, and the book itself asks for confirmation with the issuer. The paid maternity leave rule of 84 days, and 100 days for multiple births, comes from a law-firm summary of the Employment and Labour Relations Act and is to be confirmed with the issuer.

Citations differ between titles. The Law of Marriage Act, 1971 is cited as Cap. 29 in The Value of a Good Wife and as No. 5 of 1971 in Wealth That Lasts Four Generations. The reader should confirm the current citation before quoting either. The section 114 reading relies on the Court of Appeal's quotation of the Act in Yesse Mrisho v Sania Abdul, not on the Act's own text, which the book asks the reader to confirm.

Several sources were read secondhand or in part. The Probate and Administration of Estates Act was read in part, and the company rules were not read at source. The Ward study, the Baron and Lachenauer article and the Pfeffer and Killewald panel were read through secondary summaries, and the book describes them as such. The SIM registration rule in the second title is a secondary summary and must be confirmed with TCRA before anyone relies on it.

The fund and pocket money figures in the second and third titles are illustrative, computed from stated inputs, and they show a method rather than a benchmark. Survey averages, such as the 4.6 and 1.2 hours in the first title, describe populations and are not targets for any household. Planning dates, such as the handover of 12 March 2027 in the second title and the review dates of the third, are fixed in each book and should be checked against the calendar in the year they are used.

Work Smarter With AI is three books on one question: how a small business can hand its repeat work to an AI assistant without handing over its judgement or its customers' data. Each book is written for an owner or manager who already runs a business with staff. Each follows one illustrative business through seven chapters, and each opens with its own count of the work. The titles are Hire AI Before You Hire Staff (Handbook 15), Draft in Minutes, Not Days (Handbook 16) and Drowning in Noise (Handbook 17). The cases are composite and described by type, and every figure is a worked example computed from inputs the book prints. The books state Mainland Tanzania rules at dates in October 2026 and name the issuer beside each. They say that Zanzibar is not covered. This summary is general information, not legal, tax, accounting or business advice. Every rule, threshold, fee and deadline should be confirmed with the issuer named before anyone acts.

What the series covers

The three books share one question: how an owner can use an AI assistant for repeat work while keeping judgement, customer data and the firm's name under human control. They are offered as a set but can be read on their own, since each one starts from a count of the work it describes.

The cases are composite. The print shop in Mwanza, the contractor in Morogoro and the school in Tanga are described by type, not by name, and every figure is a worked example computed from inputs printed in the book. The books state Mainland Tanzania positions at dates in October 2026 and cite the issuer beside each rule. Nothing in the series is legal, tax, accounting or investment advice. For a decision that carries real money or legal risk, the books advise a qualified adviser before acting, and an advocate for any dispute or demand.

One method across the three books

The three books follow the same sequence. The reader counts the work for at least two weeks, using a log kept by the people who do it, and takes volumes from three months of records. Each task is then sorted by the cost of an error and by where its answer lives. Written instructions and a dated reference of facts come next, followed by routines for the repeat work. Every output is checked by a named person before it leaves, and a periodic review logs each error against its cause. The third book adds the step that protects the hours the routines free.

The AI is treated as a drafter, not as the author of record. Its output is checked against the firm's own sources rather than against itself, and the checker is never the drafter. Where a sentence binds the business, such as a price, a bid declaration or a notice to a parent, a named person signs it. The first book notes that the law reads a sent reply as the business's own statement and cites a British Columbia tribunal decision of February 2024, Moffatt v. Air Canada, as an example of that reading.

Error rates fall in each pilot as causes are fixed, and none of the books treats a falling rate as zero. The first book's weekly test of the same hundred past items fell from 14 to 3 errors per 100 over eight weeks, and the check stays because 3 is not zero. The second book's review log moved from 22 to 6 per 100 drafts over the same kind of period. The checks are kept after the pilot for that reason.

Book 1: count the work before buying anything

The print shop's owner had drafted an advert for a second administrator, at an employer cost of TZS 760,000 a month, and put it in a drawer until a decision date. The book's first step is to refuse to price that advert until the work has been counted. A two-week time log and three months of records show 278.0 hours of admin work a month on 220 hours of capacity, a load of 126.4%. The 2025 sales record in the book, TZS 219,960,000, gives the scale of the business against which the hire is judged.

The five routines cost TZS 1,740,000 in year 1 against TZS 9,120,000 for a hire. In the book's no-uplift case, operating profit is TZS 35,443,200 with the routines against TZS 28,063,200 with a hire, and no extra sales are assumed. The book sets a hiring trigger: a hire is made only when the routine load passes 198.0 hours, which is 90% of capacity, for two months running. A reader who cannot yet state the load cannot apply the test.

The safeguards are written rather than assumed. The Instruction Sheet gives the assistant a role, the only facts it may use, rules for common questions, the things it must never answer, and a fixed refusal: that it does not know and a colleague will reply within the hour. The Fact Bank is the price book with a checked-on date on every line, and old price lists and saved conversations holding them are destroyed. Customer names and phone numbers stay out of the tool. The reader is asked to test the sheet on twenty real messages before any draft reaches a customer.

Book 2: brief the draft before the AI writes

The second book's contractor has five writers, and the firm's own count shows 3,352 hours of writing a year against 2,640 hours of capacity, a load of 127.0%. A fluent draft made from a one-line request reads well and may be wrong, because the tool holds none of the firm's facts. In May 2026 the firm's best-written bid claimed two excavators as owned and a quality certificate it did not hold, after an assistant was asked to make the bid stronger. The book uses that case to explain why the brief, not the tool, is where most of the saving and most of the protection come from.

The remedy has two parts. The first is a one-page Draft Brief that states the purpose, the reader, the facts, the sources, the length, the deadline and the signer. The second is a Source Pack of facts the firm can prove: owned and hired equipment, redacted CVs and past projects with certificates. With those in place the planned load falls to 2,016 hours, 76.4% of capacity, with 656 hours, 32.5% of the new load, set aside for checking. A compliance matrix built from the tender document, and a method statement built from the firm's own completed jobs, let a bid go out in 46 hours instead of 70 without the unsupported claims. Letters that carry money follow a letter planner and take 18 minutes instead of 40.

Policies and reports are the other test. A policy starts from sections of law the firm has actually read and goes to an advocate before signature. A report fixes every figure to a source, and the assistant writes only the narrative around it. In the book's model, reports fall from 12 hours to 7 and policies from 20 to 12, which frees 288 hours a year. The book warns that a copied policy describes another firm while carrying this firm's signature, so a policy the firm has not read is a false statement, not a gap to be filled.

Book 3: cut the flow at its source

The third book's school counted every message, alert and call for ten working days without changing a habit. The count found 8,840 items across nine channels and 17 messaging groups, 325 minutes a day, and 134 interruptions a day. Ten class groups and the phone calls took 145 of the 325 minutes. Email, which the director had blamed first, came third. The sort also found an email offering an 8% discount on a TZS 18,400,000 textbook order, which had expired unseen and cost TZS 1,472,000.

The school cut the flow with three written instruments. The Group Rules Card reduced the director's groups from 17 to 6, handed the class groups to class teachers under one set of rules, and opened a broadcast channel for official notices. The Notification Settings Sheet cut lock-screen alerts from 380 a day to 30. The Email Triage Rules cancelled 31 newsletters and set filters, one of which places a supplier folder to be read every day. Items reaching the director fell from 884 a day to 131, 85.2% fewer, and every decision item was kept. The book warns that a ban on forwarding holds only where people are given a better route for their concern.

Reading moved to four fixed windows: 07:15 for decisions, 12:45 for email and the bursar's fee report, 13:10 for news from three named sources, and 16:30 for replies, the AI digest and the next day's list. Message time fell from 325 minutes a day to 138, and interruptions from 134 a day to 42. The 187 minutes saved come to 748 hours a year, valued at TZS 29,920,000 as time value and kept apart from cash. Over five weeks, 60 hours of the protected blocks went to the 2027 budget, to teacher appraisals and to a meals contract that fell 6%, a saving of TZS 9,300,000 a year.

The errors that recur

The first book's pilot found 46 errors in its first 1,000 checked drafts, or 4.6%: 14 wrong prices, 12 tone faults, 9 wrong promises, 6 invented items and 5 wrong customer details. Its list of sixteen recurring mistakes includes buying a tool before mapping the work, letting AI send prices unchecked, keeping an old price list in a drawer, and having no written refusal for gaps in the facts. The second book's review log held 52 errors in 400 drafts, or 13.0%, and the rate falls as each cause is fixed at its source.

The second book separates three kinds of wrong reference: invented, outdated and misread. Its contractor's reference check caught 9 of 52 pilot errors, among them a safety policy that misread section 16 of the Occupational Health and Safety Act, and a complaint drafted under the replaced 2011 procurement Act. The book judges the misread section the hardest to see, because it is real and says something different from what the draft claims. The cure is to read the section at the issuer, not a summary of it or the tool's account of it.

The third book's rumour audit found 14 forwarded claims in ten days: 3 true, 7 false and 4 that could not be confirmed. A false fee rumour put TZS 23,100,000 of tuition at risk in under a day, and one statement on the broadcast channel, citing the board minute, kept TZS 18,150,000 of it. The same book's AI digests carried an error in 7 of 40, or 17.5%, including a trip line that would have announced TZS 950,000 as agreed when 11 of 38 families had objected. The book's point is that a digest is a claim about messages nobody has read, and it must be checked against them before anyone acts.

Mainland law the books rely on

The books apply Mainland Tanzanian rules, which are spread across several statutes. The Personal Data Protection Act, No. 11 of 2022, in force from 1 May 2023, governs personal data. The books read it as requiring controllers and processors to register with the Personal Data Protection Commission, and as allowing a transfer of personal data abroad only on the grounds stated in sections 31 and 32. Data about children is treated as sensitive personal data under section 3, which is why the third book keeps pupils' details out of AI tools at the source.

The Electronic Transactions Act, Cap. 442 R.E. 2022, gives a data message legal effect (section 4) and attributes a message sent by a system programmed to act for a person to that person (section 19). That is the basis for the books' treatment of a sent reply as the business's statement. The Public Procurement Act, No. 10 of 2023, governs bids on the National e-Procurement System of Tanzania, NeST. Under section 69(4) a bidder has five working days after a notice of intention to award to complain. Under section 72 a false representation about qualifications during tender proceedings is a ground for debarment. The Fair Competition Act, No. 8 of 2003, section 16, deals with false or misleading representations in trade.

The Occupational Health and Safety Act, No. 5 of 2003, brings building and civil works sites within its scope (section 3), requires a workplace to be registered before it operates (section 16), requires a compliance licence renewed every twelve months (section 17) and requires risk assessment (section 60). The second book's check found that no section of that Act the firm read requires a written safety policy, so the policy exists for the council and the sites, not because of a section. The Contractors Registration Act, Cap. 235, governs registration classes and limits. The Cybercrimes Act, No. 14 of 2015, deals in section 16 with publishing information known to be false with intent to mislead, and in sections 20 and 23 with unsolicited messages relayed to commit an offence and with cyber bullying.

The Tanzania Revenue Authority requires a fiscal receipt or invoice on each sale from an electronic fiscal device, under the Tax Administration Act, 2015 and the EFD regulations, and the first book's quotes carry a line stating that a quote is not a fiscal receipt. On online content, the third book cites the Online Content (Amendment) Regulations, 2025, G.N. No. 57 of 2025, which add AI-generated content to the prohibited list, and the Online Content (Amendment) Regulations, 2026, G.N. No. 158i of 2026, in force from 1 July 2026, which add a Category C licence for amateur online media. Whether either set of duties falls on a school running its own groups is for TCRA to decide.

Figures and points to confirm before acting

The books' figures are computed from their stated inputs. They are not benchmarks, and a reader should not carry them into a different business without rebuilding the sums. The statutory points most in need of confirmation are listed here, each with the issuer who should confirm it.

The Personal Data Protection Commission should confirm the registration application, its fee and its time limit, which the books do not print. The regulations are cited as G.N. 449C of 2023, published on 4 July 2023, but the books record that another number is cited elsewhere, so the Commission should confirm the correct one before it is printed in any policy. The Commission should also confirm the transfer route for any AI tool that processes data outside Tanzania.

The Public Procurement Act's commencement, which the second book gives as 17 June 2024 under G.N. No. 455 of 2024, should be confirmed with PPRA, as should the five-working-day complaint window, the debarment periods, and the rules and fees of NeST. The complaint example counts that window from a notice received on Monday 7 December 2026, with Wednesday 9 December excluded as Independence Day, and gives Tuesday 15 December 2026 as the last day on the firm's count. That is an illustrative count and must be confirmed with PPRA before anyone relies on it. The Contractors Registration Act text on TanzLII is consolidated to 31 July 2002, so registration classes and limits must be confirmed with CRB.

The third book's online content points should be confirmed with TCRA, including whether the 2025 and 2026 amendments reach a school's own groups and broadcast channel, and whether any later amendment to the Cybercrimes Act changes the sections cited. The first book's fiscal receipt rules are confirmed with the Tanzania Revenue Authority. The books print no fine, penalty, fee or deadline for these matters, and the reader obtains each from the issuer.

Mainland and Zanzibar kept apart

The three books are Mainland books. The first book's local appendix is headed Mainland Tanzania, and the third book's appendix states that Zanzibar is not covered and that no fine, fee or deadline is printed. The only Zanzibar reference in the second book is a composite hotel whose complaint replies took 25 minutes each in English. That case illustrates a brief with a language and voice page for both languages and states no Zanzibar rule.

A reader in Zanzibar should not assume that the Mainland statutes named in this summary, including the Personal Data Protection Act and the procurement rules, apply there unchanged. The reader should take advice from the Zanzibar authorities on the rules that apply on the islands. A reader with businesses on both sides should keep two separate checklists, one for each, and should not transfer a figure or a deadline from one to the other.

Keeping the routines honest

The first book keeps the routines safe at their weakest point, which is the check on each draft. Drafts are checked in slots of fifteen at most. The front desk checks every quote line against the Fact Bank and its version date, with the owner signing off any line not on the list. A Friday review re-reads every tenth sent reply and fixes each cause at its source. Tool 11 logs every corrected draft by type, and each type traces to one line to fix, so that one bad draft is judged against the log and is not used as a reason to drop the routine.

The second book keeps its checks honest with a Draft Review Log kept every Friday and an unannounced sample of drafts. The sample keeps the checker reading the source first, and it matters more as errors fall. An error that reaches a client is corrected in writing within a day, logged with its cost and fixed at its cause.

The third book keeps its arrangements current through a repeated sort. The count is measured by minutes per decision and by channel, and the sort is repeated each Friday, so that a channel which starts to carry decisions again is found while it is still small. The Digest Check Log serves the same purpose for the AI summaries.

Signatures, ownership and the data file

Each book ends with a named owner for every output. The second book's Tool 11 names a drafter, a checker and a signer for each document type, and every bid carries two signatures: the quantity surveyor for every figure, and the owner for every claim and declaration. The signature is the last check, and it rests on the checkers' initials rather than on how the draft reads. An upload on NeST, or a signed PDF sent by email, is the firm's document, and only the signer sends it.

The first book's AI Use Policy is one page and is signed by all six people in the print shop. It lists what never goes into an AI tool, including names, phone numbers, balances and passwords. The shop registers with the Personal Data Protection Commission and confirms the basis of any transfer abroad. Tool 12, the Data Protection Checklist, is built alongside the routines rather than after them, and the book marks each step that depends on the Commission's forms or fees as a point to confirm.

The second book's one-page AI writing policy is signed by every member of staff and reviewed by the advocate with the firm's other five policies. Staff identifiers, health records and other clients' figures never go into an AI tool. Staff CVs go in only from a redacted file built once, which the book prices at TZS 22,400 of writing time and in which 16 × 3 = 48 entries are removed. The assistant can then format a bid without moving personal data.

The time arithmetic across the three books

Each book sets its time savings against a stated baseline, and the reader can check each sum from the inputs given. In the first book, reply time falls from 120.0 hours a month to 64.0, freeing 56.0 hours, once 70% of messages are drafted and checked and 30% are answered under the hand-off rules. Quotes and follow-ups fall from 130.0 hours to 47.7 hours a month. The report falls from 18.0 hours to 5.0, and the two routines together take 9.0 hours a month instead of 28.0. The planning figures bring the load to 128.7 hours, 58.5% of capacity, with the hours for the person's check included.

In the second book the saving is in the first draft, and the checking time is budgeted rather than hidden. The planned load of 2,016 hours, 76.4% of capacity, carries 656 hours set aside for checking. The letter planner cuts a letter from 40 minutes to 18, and the reports and policies together save 288 hours a year. The book treats the net figure as the decision number, not the gross saving, and recommends planning more bids on the cost if none wins.

In the third book the arithmetic starts from minutes a day. The count of 325 minutes a day, which the book puts at 1,300 hours a year, falls to 138 minutes a day. The protected deep-work blocks are a separate use of the freed time, and the book keeps them apart from the saving. The 748 hours a year is valued at TZS 29,920,000 as time value and is never added to cash. The term review keeps cash items, fees kept and time value on separate lines.

Across the three books the same caution applies. A time saving is an estimate until the log shows it, and a value placed on time is not money received. The reader should begin each book with the count and end it with the review, and should compare the two in the same units before deciding.

Projected cases and dates

The books state their positions at different dates, and the reader should note which. The first book is stated at 10 October 2026, the second book's statutory readings were made on 7 October 2026, and the third book's appendix is stated at 8 October 2026. Cases that carry dates after those statements are projected worked examples, not records of events, and the books say so.

The print shop's nine-week case runs from 12 October 2026 to a hire decision on 11 December 2026. The contractor's complaint example uses a notice received on 7 December 2026. The school's four-week digest audit begins on 9 November 2026. In each case the reader should treat the figures as tests of the method. A complaint window, a notice period or a renewal date runs from the date of the notice or the act, not from the date the reader happens to read the rule, and the count of working days should be confirmed with the issuer before anyone relies on it.

Where to begin

A reader who runs a business with staff and is buried in repeat work should begin with Hire AI Before You Hire Staff, because its counting method underpins the other two books. A reader who writes bids, letters, policies or reports for clients or public bodies should move to Draft in Minutes, Not Days, and study its Source Pack and four checks before any AI draft leaves the office. A reader whose day is consumed by groups, alerts and forwards should begin with the ten-day count in Drowning in Noise, since the count needs no purchase and changes no habit while it runs.

The first book's kit of tools may be copied into the reader's own business with the source named, but may not be sold, bundled or re-branded. Its own case runs over nine weeks, and a reader should expect the count and the checks to take a comparable period before the first hire decision is made. Every statutory figure should be confirmed with the issuer named in the book before it is relied on.

Read Every Deal is the fourth shelf of the Zatra Handbooks for readers who must decide on a deal before money or property changes hands. It holds four books, each answering one question the reader faces now: whether a contract is fair in its terms, whether a bargain is real, whether a payment instruction is genuine, and whether a mineral sale is lawful. The books share one method of seven steps and twelve tools, so a reader who learns the method in one book can apply it in the next. The series is written for a Mainland Tanzania setting. Its legal points rest on Mainland instruments read at the issuers in early October 2026, and Zanzibar is not covered by any book in the series. Every case is illustrative, and every money figure is computed from inputs printed beside it. This summary explains what each book decides, where the rules come from, which figures are most likely to change, and in what order to read the books. It is general information, not legal, tax or investment advice. Each rule, fee and date should be confirmed with the issuer named beside it before anyone acts.

What the series covers

The four books form a single line of protection for a person who is about to commit money. The first asks whether the contract in front of the reader is priced fairly once its hidden terms are counted. The second asks whether an offer to buy a business, a plot or shares is a real bargain or a trap. The third asks whether a payment request, a bank change or a voice instruction is genuine before money leaves the account. The fourth asks whether a gold or gemstone sale is lawful and whether the money can be recovered if it is not. Each question is decided on paper, before the signature or the transfer, and not after a loss.

The series sits in the Deals and negotiation category of the catalogue and runs as a numbered set. Book 1 is Read the Game, Not the Words, Book 2 is How to Spot No-Brainer Deals, Book 3 is The Voice Was Fake and Book 4 is Scams in African Mineral Deals. The summary follows that numbering, which is also the order the catalogue gives, although the books stand alone and each can be read without the others.

The shared method

Each book follows the same design rules, set out in the front matter of every title. One book answers one decision. Every figure is computed from the inputs printed beside it. Every rule names its issuer and a date, and the issuer is the authority to confirm it with. Illustrative cases are labelled and described by type rather than by name. Each book closes with a kit of tools to be completed at a desk on the day the decision is made.

The method has a recognisable rhythm across the four books. The reader first maps the parties or the claims, then tests the price or the proof against value, then reads the terms or the controls that decide what happens when something goes wrong, then runs the options through a bad case, and finally writes a rule that sets the walk-away point. The rule is written before the first meeting or the first payment, so that the decision is taken on the reader's own terms rather than the other party's pressure. This order is what gives the series its practical value, because the scam books and the contract book reach the same conclusion from different starting points: the decision is made when the reader can still walk away. The tools are built to be used at a desk rather than read once. Each is a single page with the case figures in one column and a blank column for the reader's own deal, so the same calculation can be run on the reader's own numbers. The kit ends with a one-page rule that is signed and dated. In these books the rule is a short list of numbers: the walk-away price, the worst case that is still tolerable, the period in which the asset could be sold and the names of the people who approve. A rule of this kind is more useful than a general intention to be careful, because it can be checked the next time an offer arrives.

Book 1: reading the game in a contract

Read the Game, Not the Words is written for an owner who has been handed a supply contract, a lease, a partnership or a joint venture and asked to sign within the week. Its worked case is an illustrative sunflower oil mill in Singida. The mill is shown TZS 125,000,000 of operating profit under the distributor's draft, but after interest and 60 days of credit the book reports that it keeps TZS 59,652,307, which is less than it would earn by signing nothing. The lesson is that the headline price is an anchor, and the reader must convert every offer into one effective price per litre or unit after payment days, penalties, guarantees and fees.

The book then moves through the clauses that decide the outcome when the deal goes wrong. Exclusivity, termination on notice, liability caps and governing law are priced in shillings, because the book finds that in the draft they run in the distributor's favour. It treats promises made in the meeting as worth nothing unless they are written into the contract, and values each one on a register before the owner decides. The scenario grid then ranks options by their worst case over the full term, not the base case. In the case, the base case of the draft beats staying as the mill is by TZS 6,418,568 over three years, but in a worst case of a 20% rise in seed from year two the same offer ends with a shortfall of TZS 381,523,351. The decision rule is to sign when an option beats the walk-away on its worst case, ask for changes where a priced clause stands in the way, and walk when no change closes the gap.

The legal points in the chapters rest on the Law of Contract Act, Cap. 345 R.E. 2019, the Arbitration Act, 2020 and the Fair Competition Act, 2003, read for Mainland Tanzania on 7 October 2026. The figures are the illustrative mill's own inputs, not market benchmarks, and the reader should confirm every clause with an advocate before signing.

Book 2: the ten-point test for a bargain

How to Spot No-Brainer Deals addresses the offer that arrives with a deadline. The seller says the price is half the market and must be accepted by Friday. The book supplies a Ten-Point Test for any business, plot or share offer, and applies it to an illustrative agro-dealer in Morogoro who is offered a mini-supermarket at TZS 120,000,000, a plot at half the market and a share tip that promises to double by Friday. On the shop's own records the book values it at TZS 69,900,000, so the asking price stands TZS 50,100,000 above value before any negotiation begins.

The test moves from price to value, then to the seller's reason for selling, the cash the business produces and its payback, the downside and the margin of safety, the title and the paper, and the exit. The payback figures in the case show the difference the test makes. At the asking price of TZS 120,000,000 the shop takes 5.71 years to return its price on maintainable cash, while at the agreed TZS 54,900,000 it takes 2.61 years, a return of 38.25% against the owner's stated hurdle of 15%. The plot is found registered to a third person under a caveat, and the share tip asks for TZS 40,000,000 to be paid to a personal number. The book's authorities for title are the Land Registration Act, Cap. 334, including the official search under section 97(2), the Land Act, Cap. 113 and the Capital Markets and Securities Act, Cap. 79. The agro-dealer closed the shop at TZS 54,900,000 with an earn-out, and walked away from the other two offers without paying a shilling. The exit figures were fixed before the purchase: the shop would take nine months to sell at a fair price, the clean plot six months and the share block 20.8 trading days, while the plot at half the market had no exit at all. The margin of safety against the conservative value of TZS 69,900,000 is 21.46% at the handover price, whereas the asking price of TZS 120,000,000 left none and stood 71.67% above value.

The reader should confirm the search, registration and share rules with the registry and issuer named in the book, and treat any statement about a plot's title as a question for an advocate until the register has been searched.

The Ten-Point Test is also a discipline on the seller's story. The book asks the reader to interview the seller about the reason for the sale and to check that reason against the records. In the case the couple's retirement is said to be the reason for selling, yet their figures leave out a rent rise of TZS 3,600,000 a year. The plot's owner is said to be travelling, and the book notes that this reason leaves TZS 33,600,000 on the table with no visible cause. The share tip's Friday deadline exists, the book observes, only to stop the checks that would expose it.

Stock is a second place where value is inflated. The sellers' figure of TZS 55,000,000 of stock is at shelf price. The book values stock at cost, which at the 18% margin the supplier invoices proved is 82% of shelf price, and the sample count suggests TZS 5,000,000 less is on the shelves than the sellers claim. The Value Sheet therefore takes usable stock at cost from a count taken with both sides present. The reader should not accept a stock figure that has not been counted on the day of handover.

Book 3: payments, voices and invoices

The Voice Was Fake addresses the fraud that works through the payment process rather than the contract. Its opening case is an illustrative trading company in Dar es Salaam. On Friday 2 October 2026 a call in the owner's cloned voice asked for TZS 186,200,000, a sum the book puts at 67.46% of a year's operating profit. A call-back to the payee on a number already on file stopped it seventeen minutes later. The book's central rule is that a firm cannot tell a genuine change of bank details from a fake one by reading the request, and must check with the payee on a number it already holds before any money moves.

The seven steps set out the controls in order. The scam map explains how cloned voices, deepfakes, fake invoices, SIM swaps and fake reversals operate. Change requests are routed through a form and a call-back, and a code word is agreed for the owner's voice. An approval matrix in three tiers puts a second approver on higher payments; in the trading company's illustrative year, 408 of 5,708 payments, or 7.15% of the count, carried a second approver and 95.53% of the money. A supplier master file check is run on the full list of payees, and a channel settings sheet locks bank and mobile money limits. The book's own control is a 24-hour wait on the first payment above TZS 20,000,000 to a new or changed account, which is a recommended design choice and not a legal requirement. The first-hour card then sets five moves in fixed order: the bank, the mobile money operator, the evidence, the police station and TZ-CERT. The book's illustration of a first drill reports 120 minutes to complete the sequence, and a second drill 55 minutes against a 60-minute target.

The authorities are the National Payment Systems Act, 2015, under which settlement is final under section 36(2), the Cybercrimes Act, No. 14 of 2015, the Bank of Tanzania and TZ-CERT. The reader should confirm the current bank recall process, the mobile money reporting line and the TZ-CERT incident channel directly with each provider, as these change more often than the statutes.

The book's figures for the trading company show the scale of the problem. Of eleven change requests that reached the company in the year, three were fake, and together they asked for TZS 203,600,000. None of the three was stopped by a rule in the system. The book treats the lesson as one of design: a request judged on its own content cannot be trusted, so the check must move to a channel the request does not control.

The supplier master file check is the book's most practical exercise. When the trading company checked its 64 payees in October 2026, 57 matched and 7 were wrong: four were outdated, two were mistyped and one could not be explained at all. That last payee was the number through which TZS 4,800,000 had left in March. The check took 26.67 hours of finance time and cost TZS 333,333. The book's point is that the cost of the check is small against the cost of one mistake, and that the check is useful only if it is repeated on a fixed date and the results are signed.

Access control is the other half of the method. The book proposes that no single person can move a large sum alone, that payments above TZS 1,000,000 need two users on two devices, that alerts go to two people, and that each till line carries a daily limit. In the illustrative company that limit was set at TZS 3,000,000 a line, which was 8.21 times the average day. The book explains that settlement under the National Payment Systems Act, 2015 is final, so a setting must be right before the money moves and cannot be corrected afterwards by a simple reversal. The Channel Settings Sheet is the place where each limit is recorded and checked.

The book is written for an owner who travels, and it therefore gives the approval matrix a deputy rule. When the owner is abroad, a named deputy approves within the same tiers, and the matrix states which days the money moves.

Book 4: scams in mineral deals

Scams in African Mineral Deals is the most time-sensitive book in the series, because its rules are set by a regulator and a central bank that revise their notices. The illustrative case is a cashew exporter in Mtwara offered 20 kg of gold dore at 18% below its value. A sample is tested in a back room in Mwanza, and a schedule of six advance charges totalling TZS 73,972,000 arrives before any gold is shown. A ten-minute discount test shows that the price is one no lawful seller could accept.

The book's seven checks begin with the law on who may possess, sell and export minerals. Under section 18(1) of the Mining Act, Cap. 123 R.E. 2019, only a mineral right holder, a licensed dealer, a licensed broker or a Minerals Import Permit holder may possess or dispose of minerals, and only a right holder or licensed dealer may export them. The Mining Commission is established by section 21(1) and is the authority with which the book checks licences and who may sell. The book then sets out the paper trail, the control of the sample by the buyer's own people, and the escrow route, under which the book's model deal releases 90% against proof and holds 10% for the refinery. It records that a test of all 20 bars in the offer would cost TZS 2,120,000, or 0.047% of the price. Two 2024 rules shape lawful gold. Under the Mining (Minerals and Mineral Concentrates Trading) (Amendment) Regulations, 2024, GN No. 539C of 30 June 2024, an authorised miner or licensed dealer must set aside at least 20% of gold for refineries, smelters, jewellers or the Government, through the Bank of Tanzania, before any export permit is issued. The Bank of Tanzania notice of 1 October 2024 sets gold sold to it at a 4% royalty instead of 6%, a 0% inspection fee instead of 1% and 0% VAT, paid within 24 hours of the refining report.

The illustrative exporter stopped on Wednesday 7 October 2026, with TZS 4,250,000 already gone and TZS 521,172,000 kept out of the scheme's reach. The Red Flag Scorecard, which scores any offer against Zatra's twelve red flags, and the Walk-Away and Report Form close the book. The reference prices in the book are illustrative, because the real figure is the daily price the Mining Commission publishes.

The foreign buyer's side of the book is equally practical. An offer of a 50 kg lot, which the book uses as a worked example, shows eleven of Zatra's twelve red flags before any deposit is paid. The Tanzanian contact who asks for money to be passed on is, the book says, the target rather than the helper. On the broker's schedule in the case, TZS 523,972,000 would have been in the seller's hands before a single gram was assayed. On an escrow path the figure is nil until an accredited laboratory reports, and the deposit asked of the buyer in the book's example is TZS 233,580,000, which is 2% of the price.

The gold scam is designed around the discount. The book's test asks whether a lawful seller could offer gold 18% below value at all, given the set-aside, the royalty, the inspection fee and the daily price. The answer in the case is no, and the book uses that single test as the first stop. A buyer who is shown a discount that the lawful route cannot support is looking at the bait, and the rest of the schedule should be read with that in mind.

Authorities, dates and what is not covered

Each book states its own date. The Mainland rules were read at the issuers or public copies in the first week of October 2026. Each book's cover is stated at 10 October 2026. Books 1, 2 and 4 read their Mainland rules on 7 October 2026, and Book 3 read its rules at the issuers on 8 October 2026 and states them at 7 October 2026. The dates matter because a rule read in early October may have moved by the time a reader acts. The series does not cover Zanzibar, which has its own authorities, fees and mineral and investment regimes, and no fee, penalty, hotline or time limit in the series should be applied to a Zanzibar matter.

The books repeatedly ask the reader to confirm a figure with the authority that issues it, rather than with the document offered. The Mining Commission, the Bank of Tanzania, the registry and the advocate are the named checkpoints. Where a book names a provider, such as an accredited laboratory, an escrow agent or a bank, it does not recommend one, and the reader should choose a provider after separate checks.

Reading order and where to start

A reader with a contract on the table should begin with Read the Game, Not the Words, then take the scenario test and the walk-away rule before the first meeting. A reader offered a business, a plot or shares within a deadline should begin with How to Spot No-Brainer Deals and complete the Value Sheet before any money is discussed. A reader who pays suppliers, staff or families by bank or mobile money should start with The Voice Was Fake, because the controls there apply to every payment whatever the deal. A reader offered gold, gemstones or minerals, or asked to pay advance charges to a foreign buyer, should read Scams in African Mineral Deals before any travel or deposit.

Most readers will need two books rather than four. The method is the same in each, so the reader who completes the first book has the habits the others expect: a written walk-away, a list of the paper, and a check on each party through a source the reader already holds.

A family business that is both negotiating a supplier contract and paying suppliers by mobile money is a common case, and it benefits from reading Book 1 and Book 3 together. Book 1 fixes the contract terms and the walk-away. Book 3 ensures that the payments made under those terms reach the right account. Neither book is sufficient alone, because a sound contract can still be undone by a fake change of bank details, and a secure payment process cannot protect a bad price.

Figures and points to confirm first

Some figures in the series are more likely to change than others, and they should be checked before any decision. These include the 20% set-aside under GN No. 539C of 30 June 2024, the Bank of Tanzania royalty and inspection fee terms in its notice of 1 October 2024, the daily mineral price published by the Mining Commission, the 24-hour and TZS 20,000,000 thresholds used as the book's own controls, the land search and registration rules under Cap. 334, and the arbitration and competition rules cited in Book 1. The illustrative amounts, such as the TZS 120,000,000 shop or the TZS 186,200,000 payment, are case figures and are not benchmarks for any market.

The series is general information. A reader should take each decision to the issuer named beside the relevant figure, and to an advocate or a qualified adviser before any contract is signed or any payment leaves the account.

Where a fee was not confirmed, the text says so rather than stating an estimate. Where a rule may have changed, the text says so and names the issuer to ask. Figures in a worked case are model inputs, and the reader should replace them with the reader's own numbers before using a tool.

Keep the Cash is the cash shelf of the Zatra Handbooks. It holds five books for owners and managers who already trade and need to know where the money goes, when it will run short, what the owner may safely draw, whether the accounts can be read, and what a loan really costs. The books follow one another. Book 1 builds a 13-week cash forecast that shows a gap early. Book 2 keeps the daily records that survive a tax examination. Book 3 separates the owner's money from the till and sets a pay the business can carry. Book 4 teaches the reader to read the profit and loss, balance sheet and cash flow statement. Book 5 tests a loan offer before anything is signed. Each book answers one money question, works through one illustrative business, and ends with a routine, a policy or a checklist the reader can keep. Figures in Books 1 to 4 are worked in US dollars, as the models are stated, and Book 5 works in Tanzanian shillings because its offers are quoted that way. The series is general financial education. It is not accounting, tax, legal or investment advice, and it recommends no bank, lender, provider or adviser. Its local rules cover Mainland Tanzania only, at the dates each book states, and every rule should be confirmed with the issuer named beside it before anyone acts.

How the five books fit together

The series runs in a sequence, though each book stands on its own. The reader may begin where the pressure is: a forecast that shows a gap, a till that does not reconcile, a pay question, statements that cannot be followed, or a loan offer waiting on the desk. The books are printed as Book 1 to Book 5, and the handbook numbers are 19, 20, 21, 23 and 24. The summaries below follow the handbook numbers, and each names its book in the series.

Every book follows the same discipline. It answers a single question, works one illustrative business through every chapter, and computes each figure from inputs printed beside it. Each rule carries the issuer, the instrument and the date on which it was read. Where a rule was read only second hand, or was not read at source, the book says so and prints no figure it could not support. The result is a method the reader can repeat with a pencil, a desk and a calendar, rather than a list of answers that will age.

The books also share a habit. Each ends with a routine or a document that keeps the work alive: a Friday cash meeting, a daily close, a monthly pack, a drawings register, a pay policy, a statement reading routine and a test to run before signature. A reader who completes one book has a tool that keeps working after the book is closed, which is the purpose of the series.

Where the cash goes

Books 1 and 2 answer the question of where the money is before the owner asks how much remains. Book 1 starts from the familiar paradox that a business can be profitable on paper and short on Friday. Its trader reports a profit of USD 42,000 on sales of USD 720,000, yet the bank balance falls to USD 2,600 in week 9 of the modelled quarter. The method is to bridge profit to cash through receivables, stock and payables, to build a 13-week forecast with one column for each Friday, and to hold that forecast against a buffer. The buffer in the model is USD 8,000, which covers one payroll of USD 6,000 and one loan instalment of USD 1,500, rounded up. The book then ranks the ways to close a gap by cost and reversibility. Moves within the business's own control come before borrowing, and each option is priced as an annual rate so that the choices sit on one scale.

Book 2 supplies the records that make the forecast reliable and the accounts defensible. Its mini-supermarket sells about USD 200 a day and cannot show the papers behind its 2025 sales. The examination test is the argument of the whole book: each figure must be walked back to a dated document made at the time of the sale. The daily routine is short, a ten-minute close that places cash, mobile money and credit sales on one sheet. The monthly routine is longer, a forty-minute pack that closes the month and keeps a file that can be opened within a day. The book states that its chapters apply in any country, while its local appendix covers Mainland Tanzania only. Where a record has been lost, the reader is told to reconstruct it from outside sources, to label the reconstruction and to date it, and never to present it as the original.

Read together, the two books give the reader the forecast and its raw material. Book 1 says when the cash will be short, and Book 2 says whether the figures that predict the shortage can be trusted.

Owners' pay and reading the accounts

Books 3 and 4 move from the till to the statements. Book 3 concerns the household's claim on the business. Its salon earns USD 1,130 a month of real profit after every cost, yet its owner takes USD 1,340 through forty withdrawals in the same month, leaving USD 840 of rent and supplier bills unpaid. The book proposes eleven evenings of work. The first evenings separate the two purses and build a drawings register. Later evenings measure real profit, test pay levels against a weak month, and set a fixed pay with a quarterly share of any surplus. A written family policy answers requests from the owner's own purse. The method is simple, and its value lies in the pay being chosen before the cash is needed, not after.

The tax treatment of that pay depends on the legal form, and Book 3 keeps the forms apart. In a sole trader or partnership, the owner's money is drawings, and the books must show where it went. In a company, the owner receives a salary as a cost of the company and a dividend from profit after the company's own tax. A dividend paid in a month with no profit is a loan to the owner, not a distribution. The book does not state the withholding or corporate tax rates, because they were not confirmed at source, and the reader is directed to the issuer.

Book 4 takes the reader through the accounts that the owner signs. Its guest house reports a profit of USD 15,292 for 2025 and holds USD 492 in the bank. The book shows how revenue, running costs, depreciation, interest and tax combine into profit, how the balance sheet records what the business owns and owes, and how the cash flow statement explains the gap between profit and cash. Four tie-out checks confirm that the three statements agree, and the book is candid that the checks cannot prove the underlying records are complete. Five ratios are then read as a set. Between 2024 and 2025 the gross margin held near 84 percent, while receivable days rose from 21.4 to 37.3 and cash cover fell from 58.6 days to 2.3 days. A reader who looks only at the margin would miss the cash problem, which is the book's central lesson. The ten questions for the accountant come near the end, each tied to the page and tool where its evidence sits.

Borrowing on fair terms

Book 5 applies the same discipline to credit. The main case is a fabrication workshop in Shinyanga with annual sales of TZS 216,000,000 and three offers for TZS 24,000,000. The book begins with the arithmetic of a flat rate, which charges interest on the original sum throughout, so its true annual cost is higher than the headline figure suggests. A reducing balance loan charges interest only on what remains outstanding. The Mainland rules read in the appendix point to the reducing balance method for the offers in the model, and the book asks the reader to confirm how that applies to each lender tier.

The offer is then tested as a complete document. The book sets seven lines that every offer must state: the amount, the method, the rate, the fees, the total to be repaid, the schedule and the security. An offer that omits a line cannot be fully tested, so the gaps become written questions. The ten tests that follow cover the true annual rate, the cash in hand after fees, the clauses on late payment, early repayment and change of rate, the spare cash calendar, the cover ratio and a cash floor, whether the purchase earns more than its cost, the loss case, the pledge and the lender's paper. The loss case is the critical test for a small firm, since it shows whether an instalment survives a bad month. In the model, the workshop's existing instalment of TZS 1,775,000 outruns its spare cash in several months, by TZS 7,434,000 in the worst month shown, and the book treats that as a failure regardless of the rate.

The legal part of Book 5 covers the protections in the Mainland rules as read. The Microfinance Act, 2018 sets four tiers of provider, and no microfinance business may operate without a licence. For the second tier, the penalty for operating without one is a fine of TZS 20,000,000 to TZS 100,000,000, or two to five years, or both. Early repayment without penalty is required for that tier, and recovery must follow written notice. A mortgage of a matrimonial home requires the signature of the spouse who lives there, and a sale requires notice. A term letting a provider change the rate alone is void. The book prints no rate cap, because none was found in the Act as read. The three offers are assumed terms and not quotes, and nothing in the book ranks a lender or a class of lender.

Mainland Tanzania and Zanzibar

The series keeps Mainland Tanzania and Zanzibar apart. Every local appendix states that it covers Mainland Tanzania only, and that Zanzibar has its own revenue authority, its own issuers and its own instruments, none of which is covered. The general method in each book, including the forecast, the record tests, the pay test, the reading of statements and the loan test, applies wherever the business operates. The local rules do not. A reader in Zanzibar should keep the method and replace each local item with the rule published by the Zanzibar authority, confirmed in writing before use.

Where a Mainland figure appears in shillings, the book prints it as the issuer publishes it and does not convert it. The one exception is the working conversion of TZS 2,600 per USD used in Book 4 to apply the company framework tests. That rate is a convention of the book and not a market rate, and it should not be used for any other purpose. Readers on the Mainland follow the same local rules, but the dates on which they apply should be confirmed at the start of each quarter.

Points to confirm with the issuer

The series prints many figures and dates, and several are provisional by the books' own account. The reader should confirm the following before relying on any of them.

Tax. The presumptive income tax table for individuals was read on the Tanzania Revenue Authority page on 20 October 2026. A law-firm summary of the Finance Act 2026 reports a change in the rate and the turnover ceiling from 1 July 2026, but the Act itself was not read at source, so the change must be confirmed with the Authority. The relief for a new taxpayer in the first twelve months is stated on the page, but its conditions beyond the page were not read. Estimated tax instalments fall on 31 March, 30 June, 30 September and 31 December according to the same page, and each business should confirm its own schedule. The rates of dividend withholding, the corporate tax rate, withholding on directors' fees and the skills levy are not printed, because they were not confirmed.

Payroll and VAT dates. PAYE and the Skills and Development Levy are due by the 7th day of the month after payroll, and VAT by the 20th day of the month after the tax period, as read at the issuers on 6 October 2026. NSSF practice and the Workers Compensation Fund rate are marked for confirmation with the Fund, because the published sources do not agree.

Records and devices. The keeping period of five years, and the longer period while an audit is open, come from the Tax Administration Act, Cap. 438. The penalty for failing to keep documents runs per month, but the value of the currency point is not printed. Whether a given shop must use an electronic fiscal device depends on the business, including the Authority's second-phase rules, which name a turnover figure of TZS 11,000,000 for persons not registered for VAT. That scope must be checked with the Authority.

Consumer credit. A complaint goes first to the lender, which has 14 days to respond, and then to the Bank of Tanzania. The 2019 regulations give the Bank a decision period of 45 days, but the 2025 amendment may change it, so the figure must be confirmed. The Bank's list of licensed institutions, and of digital platforms and apps not approved to 31 December 2025, should be checked on the day of any borrowing. The cooling-off period, the interest method for each tier and the notice periods are to be confirmed with the Bank.

Company accounts. The NBAA sets the reporting framework by entity category. The micro entity test used in Book 4 asks whether two of three conditions are met: turnover not above TZS 800 million, total assets not above TZS 400 million and not more than ten employees, as read. The audit exemption for a private company depends on a figure in the Income Tax Act, 1973, which the series does not print. NBAA Technical Pronouncement No. 3 of 2025 requires audited accounts to be submitted through its verification system for current-year submissions from 1 November 2025, and the cost is not printed. The third edition of the IFRS for SMEs applies to periods beginning on or after 1 January 2027, and whether Tanzania adopts it, and when, must be confirmed with the NBAA.

Lending protections. Book 5 prints no rate cap. The Bank regulations require a reducing balance method and the Bank's written approval for any new fee or fee increase, but the scope of each rule for each tier, the lengths of the reflection and cooling-off periods, and later amendments to the Land Act on matrimonial homes must be confirmed with the Bank and with an adviser.

Using the series well

A sound way to use the series is to start with the number that worries the owner most. A gap in the bank points to Book 1, a till that does not reconcile points to Book 2, a withdrawal that cannot be explained points to Book 3, statements that do not agree with the bank balance point to Book 4, and a loan offer points to Book 5. The reader then builds the tool the book supplies, fills it with the business's own records, and compares the result with the illustrative model. The model is never a forecast of the reader's business. It shows the method, the arithmetic and the questions to ask.

The series is most useful when its tools are kept and reviewed on a fixed day. The Friday cash meeting, the daily close, the drawings register, the statement routine and the signing test each depend on regular attention. A business that keeps them for a full quarter will have better figures than the models, and a clearer view of its cash than the accounts alone give.

Start It Right is the Zatra Handbooks shelf for the first decisions a founder makes: which business to choose, whether customers will pay for it, how the start is funded, and how prices are set so that a busy year ends in profit. It has four books, and the reader is best served by taking them in book-number order. Book 1, Fursa 365, chooses the business. Book 2, One Page, One Business, designs and tests it. Book 3, Build It With Other People's Money, funds it. Book 4, Stop Selling at a Loss, prices it. All four apply Mainland Tanzania law and practice. The cases in each book are illustrative composites built from stated inputs, not records of real firms, and the figures are computed to show the working. Where a rule turns on an authority's timing or treatment, the book says so, and the reader should confirm with the issuer named beside it before acting. This summary is general information, not legal, tax or financial advice.

What the shelf covers and in what order

The four books answer four questions in turn. The first is which business is worth testing, given the reader's money, skills and town. The second is whether customers will pay for the offer, and whether the economics hold on one page. The third is how growth is funded without borrowing more than the lowest month of cash can carry. The fourth is how the price is set so that every sale contributes to the fixed costs, including the owner's own time. The sequence matters because each book takes the output of the one before as its input. A reader who already has a tested offer can begin with Book 3 or Book 4, but a reader who has not yet chosen an activity will find the later books hard to apply.

Every book in the shelf uses the same working method. It opens with a short scope statement, sets out the illustrative case that runs through the chapters, and closes each chapter with a chapter-in-brief and six key takeaways. Each book also carries a kit of twelve tools, a closing section on sixteen mistakes and the case year in full, a list of questions readers ask, a glossary and a summary in Kiswahili. The scope statements are dated, and the books state that cases with later dates are projected worked examples rather than records of events. Positions in the books are stated at 10 October 2026.

The money figures follow one convention across the shelf. Each is computed from stated inputs, the funding book states that its figures exclude VAT and are stated before income tax, and every source of money is printed with its loss case beside its gain, so the cost of each choice is seen as well as its benefit.

The shelf does not recommend any lender, product or provider. Where a book names an authority, such as the Business Registrations and Licensing Agency, the Tanzania Bureau of Standards (TBS), the council licensing system or the Tanzania Revenue Authority (TRA), the reader is directed to that authority for the current rule. The books explain how the steps connect, not what each office will decide on a given day. That distinction is the thread of the whole shelf.

Choosing and testing the business

Fursa 365 treats choice as a scoring exercise and not an act of enthusiasm. A catalogue of 365 entries cannot know the reader's money, skills or town, so the first task is to filter the list to a shortlist of three and score each against six written weights. In the worked case the peanut butter maker scored the preferred entry at 79.0, a laundry at 76.0 and layer chickens at 61.0. The margin between the first two was smaller than one demand point, and the book is explicit that such a lead justifies a test, not a spend. The lesson carries over to any reader who is tempted to commit on the strength of a close score.

The demand test in Fursa 365 is designed to fail cheaply. Three hypotheses are written with pass marks before the work begins. In the case, a shelf audit of 40 shops found 26 stocking the product, and at a market 78 of 120 tasters said they would buy at TZS 6,500, which is 65.0% against a pass mark of 60%. Fourteen shops then signed trial intents for 84 jars, and eleven of them reordered in June. The test cost TZS 127,000 and sold nothing before certification. Licensing follows the test. The case records a business name and a tax number as quick steps, while the TBS step is described as the longest. The TBS application was lodged on 23 February 2026 and granted on 15 May 2026, an interval the book gives as an illustrative 81 days. The regulator's timing is not published, so the reader should plan on the issuer's current practice rather than on that figure.

Fursa 365 then prices for margin rather than hope. The peanut butter maker sells a 400 g jar at TZS 5,000 to shops and TZS 6,500 direct. With a unit cost of TZS 3,000, the shop jar leaves TZS 2,000 and the direct jar TZS 3,500, and a mix of 70% shops and 30% direct gives a blended TZS 2,450 and a gross margin of 44.95%. Fixed costs of TZS 1,300,000, the owner's pay included, need 531 jars a month, or 583 with the 4% tax set-aside the book models. A cut to TZS 4,500 a jar would have pushed break-even to 620 jars. The first customers are named before launch from the demand test, 14 trial shops with signed intents and 150 WhatsApp contacts among them. Eleven of the 14 trial shops reordered within 30 days. The book then holds the twelve-month plan against each month's actual figures on the first Monday. In the case, June lost TZS 544,000, July lost TZS 216,312 and August made TZS 163,280 after the owner's pay, and a groundnut price rise from TZS 4,000 to 4,300 a kg moved the margin from 45.65% to 42.49%. The gap is explained in shillings and a decision is made each month, not left for year end.

One Page, One Business applies the same logic to the whole model. The Arusha laundry's first canvas assumed households would be the core market, but the interviews found only 3 of 20 households would use self-service, while 8 of 20 would pay for collection. Seven of twelve lodges named linen a top-three problem, and lodges became the first segment. The paid pilot with two lodges then found a price of TZS 2,800 a kg against TZS 3,200 planned, and about 1,000 kg a lodge against 1,400 kg a month planned. The canvas, updated after each test, is a record of what was learned. The book argues that a lender should see that record before the loan is discussed, and its bank plan therefore leads with the figures and keeps the canvas as evidence.

Funding the start in layers

Build It With Other People's Money begins from a practical problem: the draft plan was profitable on paper but ran out of cash. The workshop's first instalment was 490.03% of its 2026 monthly profit, and the cash gap fell in a month when the desk materials were paid before the school paid. The remedy is not a larger loan but an earlier source of cash. Customer deposits are the first layer. The book shows that a 50% deposit at order pays for the materials of a custom piece 1.19 times over, and that a 40% deposit at signing from the school pays for the first lot of desks before any board is cut. Deposits in the plan peak at TZS 37,400,000 in April, 32.84% of the stack, and each one is registered, matched to its materials and written with refund terms.

The growth year raises the workshop's materials bill from TZS 96,768,000 in 2026 to TZS 168,288,000 in 2027, with May and June each above TZS 26,000,000. Credit lines must therefore be sized to those months and not to the annual average. Supplier credit is the second layer. Its price is the cash discount that the buyer gives up. The book sets out the timber line as a 2 in 98 discount forgone over 30 days, which is 24.83% a year, and the consignment stock as a 3% premium on TZS 4,000,000 of fittings, costing TZS 504,864 in the year. Both are kept by paying on the agreed day. The third and fourth layers are assets that are used but not owned, and skills that are bought as needed. The book tests each machine against the number of days it would be used, so that a panel saw used every day is bought outright, while machines used less often are bought over time or hired. A truck is costed to own against hire by the trip, with break-even at 257 trips in the case. Partners are admitted only on written terms, set before any money or shares move.

The final loan is sized to the lowest month of cash rather than to the plan. In the case, January 2027 is the lowest month, with two machines and a hire purchase deposit paid. Without a loan that month ends short. The loan is set at TZS 15,000,000, the smallest figure that leaves a month of fixed costs of TZS 9,000,000 in hand. The lender is chosen from the Bank of Tanzania's list, after a free credit report, by comparing key facts statements on the effective rate and the total paid. Against the investor, the book values a 20% share that would claim a fifth of profit in every later year, while a loan of the same size costs TZS 7,244,976 with its fee and then stops. Profit before tax in the final plan is higher than in the draft, and cover rises from 1.30 to 1.90 once the investor is set aside and the loan is reduced.

Setting prices that cover the year

Stop Selling at a Loss starts from the observation that a busy workshop can still lose money. The uniform workshop booked 97.86% of its sellable hours, yet the school set lost TZS 2,775 a set and TZS 13,320,000 in the year. Counting the owner's time and the idle hours, the five products lost TZS 10,660,000 and the idle hours TZS 2,882,000. The book's answer is to cost one unit with every hour counted. Dividing fixed costs by the hours actually booked would raise the rate whenever trade fell, so the book sets the rate on sellable hours and shows idle time on its own line.

The price itself is built from three numbers: a floor of materials and packing, a cost-plus price that keeps 12% after every hour and the owner's time, and a value ceiling set by the buyer's next best choice before VAT. The book is precise that a 12% margin on price is full cost divided by 0.88, and that multiplying cost by 1.12 gives a smaller margin. Discounts are then priced in volume. A discount comes out of contribution, so the extra volume needed to replace it is the discount divided by the contribution percentage less the discount. On the corporate set, 10% off needs 23.04% more sets, and 20% off needs 59.86% more. Each sales channel is given its own terms, earned by a condition such as a minimum order, a delivery run or a payment date, and written down.

Channels are where the hours are lost or won. In the base year the school set sold at TZS 27,900, TZS 24,000 and TZS 25,000 in three channels, all below its full cost of TZS 28,050, and three of the five channels earned less than the TZS 6,700 an hour the workshop needs. The new terms give each channel its own price, earned by a condition: one counter price, 12% to uniform shops for 50 sets or more, school contracts at TZS 29,000 above the floor, 5% to companies paying by invoice date, and 8% to lodges for 100 sets with delivery in the region included. Channel contribution then rises from TZS 120,258,000 to TZS 174,240,400 on fewer hours. The book's rule is that when hours are full, channels are ranked by contribution an hour, and a trade term is given only for a minimum, a deposit, a delivery run or a payment date, written down.

When a rival cuts, the book holds the price and answers with value the workshop can prove. Following a cut to TZS 27,000 would reduce the contribution from TZS 16,000 to TZS 9,000 a set, need 77.78% more sets and lose TZS 1,050 on each set against the full cost. Holding keeps about 1,300 of 1,800 town sets and earns TZS 20,800,000 against TZS 16,200,000 for following. The 750 hours freed make 214 corporate sets worth TZS 10,058,000, a total of TZS 30,858,000. The workshop follows only on school contracts down to the full-cost floor, and never agrees prices with other tailors. The final chapter sets the review calendar. A fabric rise of 8% notified in January is priced into a list reissued for 1 February, which leaves TZS 2,940,000 for the fixed school contracts to absorb, against TZS 7,909,550 if the review waited for year end.

What to confirm before acting

All four books state that their cases are illustrative and that their figures are computed from stated inputs. The inputs are the reader's own to replace. A reader who copies a case figure into a plan without substituting local costs, prices and dates will repeat the error the cases warn against. The first check in every book is therefore to rebuild the working with the reader's own numbers, using the tools in the kit.

Several points need confirmation with the issuer before the reader relies on them. The TBS timing is not published, so the 81 days in the Fursa 365 case is an illustration and not a standard. Business name, tax number, sector approval and council licence rules should be checked with the registering office, TRA, the relevant regulator and the council at the time of application. Build It With Other People's Money names VAT on customer deposits as a point for TRA to confirm, since the book cannot settle it from the model. Its loan rate, lender fee, hire purchase terms, cash discount and investor price are assumptions for the model and not quotes, so each must be obtained from the actual provider. The Mainland rules on lenders, loan agreements, credit reports, leasing and VAT were read at the issuers on 7 and 8 October 2026 and should be read again before use.

Stop Selling at a Loss uses VAT at 18% in its tables, and the reader should confirm the rate in force with TRA before quoting. The price review dates in that book are proposed dates for 2027 set by the workshop, not by any authority. Legal and tax points, including the treatment of an investor's share of profit, the terms of a partnership and the structure of a company, fall outside the books' scope and should be taken to a qualified adviser. The books are written so that the questions are clear enough for that adviser to answer quickly.

Reading the books in use

A reader with no chosen business should begin with Fursa 365 and treat its scoring and demand tests as a condition of spending. A reader with a chosen offer should move to One Page, One Business and run the pilot before any loan application. The funding book comes next, once the cash needs of the first year are known. Pricing is then set against the real hour cost and the real channels. Each book can also be used by chapter, since the chapter-in-brief and the takeaways give the figures and the conclusion in short form for later reference.

Used in this order, the shelf moves a founder from a list of options to a priced, funded and tested business. The books do not promise a result. They show the working behind each decision, name the points where the rules must be checked with the issuer, and leave the final choice to the reader and to the advisers the reader consults.

Run It Tight is the operating shelf of the Zatra Handbooks. It is written for owners, managers and working family members of trading businesses that sell well yet keep too little of what they earn. The shelf has three books, each with one business case, a seven-step method and a kit of twelve tools. Together they answer three questions: where profit leaks out, which few numbers reveal the leaks in time, and how a team can carry the routine when the owner is not in the room. The cases are composite and illustrative, built from stated inputs, and the books state Mainland Tanzanian rules only. This summary explains what each book covers, how the three arguments connect, and which figures and legal references need confirmation before anyone acts. It is general business information. It is not accounting, payroll, legal or tax advice.

What the series covers

Run It Tight has three books, numbered 25, 41 and 43 in the Zatra Handbook series and marked Book 1 of 3, Book 2 of 3 and Book 3 of 3. All three are first editions with positions stated at 10 October 2026. Each follows the same design: an opening that sets out the case, seven chapters that each take one step, a chapter-in-brief with six key takeaways at the close of every chapter, a twenty-statement check at the front, and an after-the-chapters section that holds the kit. The kit contains twelve tools, each printed on one A4 page with the case figures in one column and a blank column for the reader's own business.

The books share one argument. Profit is decided by a small set of flows, numbers and people that ordinary records already touch but no one reads together. Book 1 finds the seven leaks. Book 2 chooses the seven numbers that expose them and places them on one weekly page. Book 3 builds the team and the written limits that keep the routine running when the owner is away. The books can be read in number order, which is the order the series itself gives under Also in this series. A reader whose immediate problem is cash can start with Book 1 and read the others later.

The books have one reader in mind, an adult facing a business decision with real money at stake. They draw their examples from a bakery, a pharmacy and a restaurant. Book 2 and Book 3 each include an appendix that applies the seven steps by trade. A business with a point of sale or a fiscal device gets the most from Book 2, because the method starts from the records those machines already print. A business that keeps only a notebook is shown how to use the same method without one.

Poor Management, Poor Profit: seven leaks in one bakery

The case is a bakery in Dodoma. It bakes 1,000 loaves a night, sells TZS 498.6 million a year and keeps TZS 24.6 million. Its owner works hard and blames flour prices. The book argues that the figures point elsewhere: bread baked and never paid for, discounts nobody approved, schools that pay in 75 days, flour bought at the wrong price, bakers waiting for the mixer, cash taken from the till, and a second oven used one day a week. Each leak receives a measure, a tool and a fix, and the seven chapters follow the same order.

Chapter 1 begins with production, because sales records cannot show what was never sold. Of the 1,000 loaves baked each night, 900 are paid for. The Waste Count found that, of the other 100, 46 came back stale, 17 went missing without a record, 31 were given free to agents and 6 went home. The book values the stale, missing and over-recipe flour at TZS 19,143,095 a year, which it puts at 77.9% of profit. A count signed by two people each morning, a recipe card, a scale at the mixer and a 2% target recover an amount the book models at TZS 14,600,375, without accusing any individual.

Chapter 2 asks what each sale really brings. An agent loaf listed at TZS 1,500 brings the bakery TZS 1,304 once the discount, free loaves, free delivery and bad debts are counted. The price waterfall makes that visible line by line. Chapter 3 treats credit. Customers, and schools in particular, take longer than the terms agreed, and the business keeps supplying them. An ageing list and a stop-supply rule bring credit back inside its terms. Chapter 4 covers purchasing, which leaks through two gaps: a price above what the market offers, and weight paid for that never arrives. A three-quote sheet and a goods received check close both.

Chapters 5 to 7 turn to labour, owner pay and the last leak. Chapter 5 asks what the eight bakers are paid for. The bakery pays for 72 hours a night while bakers wait for the mixer, so the chapter pays for work rather than waiting. Chapter 6 shows that the owner took TZS 41,682,816 from the business in a year, of which only TZS 24,000,000 was recorded as pay. It argues for a fixed salary, recorded as pay, in place of cash taken from the till. Chapter 7 finds the last leak, the cost that earns nothing, and sets the routine that keeps all seven closed: one monthly leak page read on the first Tuesday of every month.

The practical lesson is the order of work. Count at the source first, then price, credit, purchasing, labour and owner pay, because each later measure depends on the earlier count. The bakery is illustrative, so the reader should test each leak against the business's own till records and production counts before deciding that a leak exists. The figures show what a count can reveal, not what any particular bakery will find.

What each book asks the reader to produce

Each book ends its chapters with something the reader can keep, and the three outputs build on one another. Book 1 asks for a Waste Count signed by two people each morning, a recipe card and scale for the mixer, an ageing list for credit, a three-quote sheet for each significant purchase and a goods received check at each delivery. Together they make a monthly leak page with seven measures, read on the first Tuesday of every month.

Book 2 asks for a Definition Card for each of the seven numbers, signed and frozen for a fixed period, a collection calendar that names one record, one moment and one person for each number, and a single landscape page with seven rows, targets and two lines per row. It then asks for a fixed Monday meeting at which every red has a first action, an owner and a day. The pharmacy's cards were frozen until 12 January 2027, which shows how long a definition is expected to last before it is reviewed.

Book 3 asks for a role map that gives each role its outcomes, limits and one boss, a hiring scorecard written before advertising, a ten-day induction plan, a delegation ladder with levels and shilling limits, a set of three to five written standards for each role, and a stay conversation at set intervals. Each output is designed to be checked by a second person, and none depends on the owner remembering a conversation.

What Gets Measured Gets Managed: seven numbers on one page

The case is a pharmacy in Mwanza. It sells TZS 602.8 million a year and keeps TZS 25.2 million. Its point of sale and fiscal machines print 43 figures a week, and the owner spends six hours every Sunday reading them. Yet four of the seven numbers that control profit and cash appear in none of those printouts. The book treats the problem as one of selection and definition rather than of data volume.

Chapter 1 chooses the seven numbers. Each must pass three tests: it moves profit or cash, it can be acted on within a week, and it is cheap to collect. The largest single move in the case is halving the requests turned away, 18 a day or 126 a week, which the book values at TZS 8,845,200 a year. Stock and claims tie up TZS 100,164,822 of cash. The turned-away requests represent TZS 17,690,400 of gross profit a year that the records had never counted.

Chapter 2 makes each number mean one thing. A Definition Card gives the formula, unit, source, cut-off, counter and a worked example, and is signed by the people who use it. The pharmacy's staff had produced four margins from the same machines: a markup on cost of 42.86%, a POS margin of 28.80% and a margin kept of 26.19%. The book also shows that 168 receipts a day were 156 people, which is a reason to define a customer rather than count receipts. The cards were signed on 10 October 2026 and frozen until 12 January 2027.

Chapter 3 sets a collection calendar. Each number has one record, one moment, one person, a fixed number of minutes and one cross-check. Sales come from the POS and the Z reports of the fiscal device. Turned-away requests come from a counter book written at the moment of asking. Stock days come from a 30-item Sunday spot count, near-expiry stock from a daily shelf sweep, and rejections from the claims log. The owner's time falls from 360 minutes to 50 minutes a week, a saving of 268.67 hours a year.

Chapters 4 and 5 design the page and set its signals. The page is one landscape sheet with seven rows. Each row carries the number, its target and two lines, thirteen weekly cells and a status, and colour appears only where a written rule calls for action. An amber line starts a question and a red line forces an action. The targets were set on 26 October 2026 from last year's figures, the moves priced in Chapter 1 and its best weeks. In week 4 near-expiry stock read TZS 7,800,000, TZS 1,800,000 above its red line. The action it fired returned TZS 1,400,000 of one batch for credit and sold 900,000 first, saving TZS 2,300,000. Reds fell from four a week to none by week 6.

Chapters 6 and 7 turn the page into decisions and guard it from distortion. The review is a fixed meeting held every Monday from 07:30 to 08:00, with five minutes for reading the page, fifteen for reds and ambers and ten for new actions, each with one row, one name and one day. Over thirteen Mondays the pharmacy logged 31 actions, closed 26, and reduced claims rejected from 7.5% to 2.4%. Chapter 7 shows the risk of a number rewarded on its own. A bonus of TZS 20,000 a week to each of four counter staff, for keeping turned-away requests low, turned the recorded figure green in weeks 8 and 9, at 30 and 26. A Sunday spot count showed the true figures at 61 and 57. The fix is a paired measure, a spot check, and a team bonus paid monthly only on checked numbers.

The Team That Wins Without You: an Arusha restaurant

The case is a restaurant in Arusha. It sells TZS 677.0 million a year and keeps TZS 52.6 million. Its owner cooks, buys, cashes up and settles every dispute, working 84 hours a week and answering 46 questions a day. Eleven of twenty-one staff left in the year, six of them within 90 days. The restaurant counts each leaver at TZS 614,645, or TZS 6,761,095 for the eleven. Nine staff work without written particulars and without NSSF registration.

Chapter 1 designs the roles. A time log shows 38 hours of work the team is already paid to do and 39 hours that belong to no role. A role map adds a restaurant manager at TZS 24,624,000 a year, gives eleven roles their outcomes, limits and one boss, and cuts the owner's direct lines from 21 to 3. The plan year adds TZS 22,814,150 to operating profit. If only half the gains arrive, the plan leaves a shortfall of TZS 4,879,985.

Chapter 2 hires on a scorecard written before the advert. The restaurant screened 38 applications down to 12 on must-have criteria, interviewed 5 with six weighted questions scored against anchors, paid two finalists TZS 60,000 each for a trial shift, and called two referees for each. The preferred candidate scored 4.25 against 3.95 at interview and 55 of 70 on the shift. The restaurant hired the candidate at TZS 1,800,000 a month from 1 December 2026 on a written contract.

Chapter 3 pays fairly and lawfully. Nine staff, on a payroll of TZS 3,600,000 a month, have no written particulars, no payslip and no NSSF account. At the joint rate of 20%, read on 8 October 2026 and to be confirmed for 2026/27, the unmade contributions come to TZS 720,000 a month. Full compliance costs TZS 17,740,800 a year, and operating profit falls from TZS 52,571,600 to TZS 48,251,600. In return, every worker holds a contract, a payslip and a pension number, and the owner's compliance cost is set out in full. For the manager's post the book adds employer on-costs of 14%: the NSSF employer share of 10%, the Skills and Development Levy of 3.5% and the Workers Compensation Fund of 0.5%, each to be confirmed for 2026/27.

Chapter 4 inducts new staff. A ten-day plan of shadowing, helping and then working alone, a skills matrix, a cook who teaches on a timetable and a day-30 talk bring a new waiter to work alone in 12 days instead of 30. Each leaver avoided is worth TZS 138,472 by the restaurant's count. Aiming for five leavers a year instead of eleven saves TZS 3,687,870.

Chapter 5 delegates by written limits. A ladder gives every recurring decision a person, a level from 1 to 5 and a shilling limit. Single purchases are capped at TZS 500,000, 200,000 and 50,000 for the manager, the head cook and the floor supervisor, with refund limits of TZS 100,000, 30,000 and 30,000. Dismissal stays with the owner. A weekly rhythm of 235 minutes carries the exceptions, and the owner's week falls from 84 hours to 20 between 7 December 2026 and 1 March 2027.

Chapter 6 reviews performance and rewards a checked number. Each role has three to five written standards drawn from records the restaurant already keeps, a monthly 20-minute check-in with its lead, and a review twice a year on a three-point scale. Feedback names the task and the standard, never the person. One team reward is paid on a checked number: 30% of the food cost saved below a 38% level. On a quarter's sales at last year's rate of TZS 151,255,000, a food cost of 36% gives a pool of TZS 907,530, or TZS 113,441 for each of eight kitchen staff.

Chapter 7 keeps people or lets them go lawfully. The restaurant holds stay conversations at weeks four and eight and every quarter after that. Where a person must go, the Act asks for a valid reason and a fair procedure, and the employer carries the proof. The cashier's case in the book took 66 days and cost TZS 701,593 plus notice, against not less than TZS 5,760,000 had it been found unfair.

Figures and points to confirm before acting

The figures in all three books are composite worked examples. They are built from stated inputs, exclude VAT, and show each calculation. They are not statistics or benchmarks, and a reader should replace them with the business's own records. Positions are stated at 10 October 2026. The Mainland rules in the books were stated at 7 October 2026. Book 3 records that its rules were read at the issuer on 8 October 2026, and where a rule could not be read there, the book says so and names who confirms it.

Dates need the same care. Some cases carry dates after 10 October 2026, such as the restaurant's hiring from 1 December 2026 and its owner's week falling to 20 hours by 1 March 2027. Book 1 states that such cases are projected worked examples, not records of events, and the same reading applies to the other two books. The reader should treat them as a plan to test, not as results that have already happened.

Payroll rates need the most care. Book 3 gives a joint NSSF contribution rate of 20%, an employer's NSSF share of 10%, a Skills and Development Levy of 3.5% and a Workers Compensation Fund rate of 0.5%, each to be confirmed for 2026/27. The same book states that NSSF adds 5% for each month or part of a month a sum stays unpaid. Each rate should be confirmed with its issuer before a pay grid is set: NSSF for pension contributions and the penalty, the Workers Compensation Fund for compensation, and the Tanzania Revenue Authority for the Levy and PAYE.

The two books cite the Employment and Labour Relations Act differently, and the difference must be resolved. Book 1 cites Cap. 366 and places the ordinary maximum of six days and 45 hours a week in section 20, with the night premium in section 21(4). Book 3 places the same hours rule in section 19, noting that the 2023 consolidation numbers it section 20. Book 3 also places dismissal in section 37, or section 38 in the consolidation. No section number should be relied on until the current text is checked. The wage order for hotels and restaurants, which Book 3 says the owner confirms with the Labour Commissioner, should be checked in the same way.

The receipt rule has a legal basis, but the TZS 10,000 weekly limit does not. Book 2 cites section 36 of the Tax Administration Act for the fiscal receipt that every supply requires. The book then sets its own rule that any difference above TZS 10,000 in a week is written down and explained. That threshold is an internal control chosen by the book, not a legal limit, and should be read as one.

The scope is Mainland Tanzania only. None of the three books addresses Zanzibar. A business that operates in Zanzibar must check Zanzibar's own labour, tax and receipt rules separately, and must not assume that a Mainland rate or section applies there.

No lender, bank, software product, recruitment agency, trainer or medicine is named or recommended. Where a decision carries real money or legal risk, the books direct the reader to a qualified adviser, to an advocate for any dispute, or to Zatra, before acting.

Reading order and how to use the kits

Book 1 is the natural first step for a business that is busy and selling but short of cash. Book 2 follows for an owner who must decide which numbers to watch each week. Book 3 suits an owner whose team, decisions and hours have outgrown one person. The three books point to one another in their Also in this series lists, and each works alone.

Each book opens with a twenty-statement check that the reader can score before reading the chapters. Each chapter closes with a chapter-in-brief that states the case result and six key takeaways, followed by a question that leads into the next chapter. The structure allows a book to be read in short sittings, with the chapter-in-brief serving as a review before the next chapter begins.

Books 2 and 3 each add a worked year in practice and a list of common mistakes near the back. Book 2 includes thirteen Mondays in practice and sixteen mistakes, and Book 3 includes twenty weeks of stepping back in practice and sixteen team mistakes. These sections show how the method ran in the case business over time, and they are the best place to test the reader's own plan against a full sequence of weeks.

The kits may be copied into the reader's own notebook and used inside the reader's own business with the source named. They may not be sold, bundled or re-branded. Book 2's kit includes a one-page measurement policy, and Book 3's a one-page people policy. Each book also carries a Kiswahili summary and notes by chapter, and the kits can be read online at zatra.co/guides under each book's title.

Where the three books agree

Three ideas recur across the shelf. The first is that the count belongs at the source. Book 1 counts bread at the oven, Book 2 counts turned-away requests at the counter when the customer asks, and Book 3 counts hours in a time log. A sales record cannot show a sale that never happened, so the books treat that gap as the first place to look.

The second is definition before measurement. A number that is not defined on a signed card will be read differently by each person who reads it. Book 2 shows how four margins from one set of machines can disagree. Book 3 uses written standards for each role, so that feedback names a task rather than a person.

The third is a fixed meeting. The bakery reads its leak page on the first Tuesday of each month, the pharmacy reviews its page for thirty minutes at 07:30 every Monday, and the restaurant holds a weekly rhythm of 235 minutes. A record without a fixed time and a named owner is unlikely to change the business.

The fourth is protection from distortion. A bonus on a single number drifts from what it measures, and each book pairs its measures with a spot check, a second record or a written limit. The practical rule for any reader is to check every number against a second source before it drives a decision or a payment.

The fifth is that every number has a named person. Book 2 gives each of its seven numbers one person who makes the record, and Book 3 gives each role its own written standards and its own limit. A figure with no owner is not checked, and a standard with no owner is not enforced. Each book therefore asks who records, who reads and who acts before it asks what the figure says.

Grow Your Wealth is the shelf for a reader who already has something to protect or place: savings that earn little, a lump sum from a sale, a business that needs capital, or a lending group that holds other people's money. It has five books, read in book order: Hide Your Treasures in Appreciating Assets (Book 1), The Secret to Sound Investment (Book 2), The Leverage Secret (Book 3), Successful Money Lending Models in Africa (Book 4) and Never One Basket (Book 5). The series is financial education under the Bank of Tanzania Certified Financial Educator standard. It recommends no bank, fund, share, bond, dealer or plot. Its households, firms and groups are illustrative, and every rate of return in them is an assumption. This summary sets out what each book settles, where the arithmetic is most likely to mislead, and which figures must be confirmed with the named issuer before anyone acts. It is general information, not investment, legal or tax advice.

What the series covers

The five books take a reader through one sequence of decisions. The first asks whether idle money should stay idle, and what inflation takes from it each year. The second asks what an asset is worth before its price is paid. The third asks whether borrowed money, or other people's time, skill or capital, will multiply a result, and what a bad year does to it. The fourth applies the same discipline to a lending group, where the money belongs to members and the lender must hold a licence. The fifth asks how the whole of a household's wealth should be arranged so that no single bet can ruin it.

Each book has the same architecture. It has seven chapters, one step each, and every chapter ends with a short task for the same evening, tied to one tool in the kit. A dated first-year part then follows one illustrative household, investor, farm or group through the year. Every gain is printed beside a loss case for the same asset. A reader who works through a book leaves with a written rule, not a recommendation to buy or sell anything.

Who the series is for

The audience is Tanzanian households and small owners who must make a money decision without formal training in finance. Book 1 suits salaried workers, traders, farmers and couples who hold savings in a box or a current account. Book 2 suits a person with a lump sum to place, such as a trader who has sold stock or vehicles, a retiree with a gratuity, or an owner offered a stake in a friend's business. Book 3 suits owners of working businesses and farms, and families offered a loan or a partner. Book 4 suits officers and members of VICOBA and SACCOS groups, microfinance staff, and people asked to put money into a lending business. Book 5 suits owners who already hold a business, a plot, a few shares or a deposit.

None of the books assumes a bank account full of spare cash. Each assumes only a willingness to write figures down before money moves, and to read the source of each figure. That habit is the series' main teaching goal.

Book 1: idle money and the emergency fund

Hide Your Treasures in Appreciating Assets opens with a Dodoma household that holds TZS 26,000,000 earning nothing: TZS 6,000,000 in a locked box and TZS 20,000,000 in a current account. At an assumed 4% inflation, that money buys TZS 17,564,668 of today's goods by October 2036, a loss of TZS 8,435,332. The book anchors its inflation discussion to the National Bureau of Statistics consumer price index, and cites the year to February 2026, when the index rose 3.15%. A figure that never changes on a statement can still lose buying power every year, and the book makes that loss visible in shillings.

The second chapter sizes the emergency fund before any shilling is invested. Six months of essential spending, which for this household is TZS 10,500,000, stays safe and ready. The book prices the cost of that safety. In a savings account at an assumed 2.5%, the fund gives up TZS 151,442 of buying power at 4% inflation. Chapters three to five then compare assets by real return, calculated as (1 + nominal return) divided by (1 + inflation), minus 1. A unit trust returning a nominal 9.0% is a real 4.81%, not the 5.00% a shortcut gives. Each asset is first stripped of its costs, fees, withholding tax and land rent. On a Treasury bill an 8% discount becomes 7.83% after 10% withholding tax, a rate the book flags for confirmation with the Tanzania Revenue Authority.

The plan in chapter six splits TZS 15,500,000 across four assets and sends a TZS 550,000 monthly surplus to three of them, with three dated goals. The mid case ends in October 2036 at TZS 145,377,882, worth TZS 98,212,088 in 2026 shillings, against TZS 81,500,000 paid in. The low case ends at TZS 106,503,156, and in that case the shares sit below what was paid in. The book treats that shape as the honest result of a ten-year plan, not a failure of the method. The final chapter sets a yearly review on the second Monday of October, at which the real return is measured after money paid in, drift is recorded, and sales are made only for a dated need. The same chapter says each adviser's licence should be checked with the Capital Markets and Securities Authority.

Book 2: how much to pay

The Secret to Sound Investment separates price from value. Its illustrative investor is a lorry owner in Morogoro who holds TZS 90,000,000 after selling two lorries and keeping TZS 6,000,000 for a third lorry and the family. Before placing any money, he writes two numbers for every offer: the price asked and his own estimate of value. Plot A asks TZS 34,000,000 against a working value of TZS 25,200,000. The book takes the middle of three estimates as the working value and the lowest as the floor, and sets a margin of safety as (working value minus price) divided by working value. The idea is credited to Benjamin Graham, chapter 20 of The Intelligent Investor, fourth revised edition, 1973. The percentages in the case, 30% for land, 33% for listed shares and 40% for a private stake, are the book's own and not Graham's.

The book's central warning is that careful estimates go wrong in a known direction. It cites Flyvbjerg and colleagues, who found projects costing 28% more than planned on average across 258 projects, and Buehler and colleagues, whose work took 55.5 days against 33.9 predicted. The lesson is that a single estimate hides its error, because nothing on the page checks it. Three estimates, kept apart, expose the spread.

For a plot, the three estimates are adjusted titled sales, a subdivision residual tested with costs 65% higher, and a lease floor. The title is searched before the price is discussed. For a listed share, value comes from the company's five-year figures rather than the price sheet, and the investor waits for a price at or below a stated maximum. For a stake in a private business, the driving school's TZS 24,000,000 is cut to TZS 9,800,000 of real profit after a manager, the fleet and tax, and a 40% minority stake is valued from there. The book shows the investor declining Plot A and waiting on the share, so that the margin of safety is kept rather than argued away. Its loss cases are printed beside each gain.

Book 3: when leverage pays and when it ruins

The Leverage Secret follows an illustrative layer poultry farm on the edge of Mbeya city, with one house of 3,000 hens and owner's equity of TZS 110,000,000. A bank offers a loan to scale up, and three plans are compared: Plan A with no lever, Plan B with a measured lever, and Plan C with a full lever secured on the family home, which is worth TZS 120,000,000 and sits outside the farm's figures. The 18% loan rate is an assumption. The book sets the Bank of Tanzania's published average beside it for comparison and quotes that average with its month.

The central figures show the mechanism. Plan C borrows TZS 150,000,000 and lifts the normal-year return on equity from 19.91% to 52.86%, but deepens the worst-year loss from 29.06% to 94.05%. The book's test for any lever is the spread, which is the return on what a loan buys less the loan's true annual cost. House 2 earns 33.89% in a normal year against the 18% assumed cost, a spread of 15.89 points. In a worst year the same house loses 25.96%, which is 43.96 points short. A lever multiplies the good year and the bad year by the same factor.

The same test applies to other people's time, skill and capital. A manager paid TZS 10,200,000 a year frees 36 hours a week for premium buyers, and the book prices those hours at about TZS 5,449 each before asking whether they earn more. A vet on retainer at TZS 2,400,000 a year pays only if replacement cost falls and laying rises, and the book prints the case where neither happens. Money that carries no interest still has a price. The book converts it to a yearly rate with one rule: the extra given up, divided by the money held, times 12, divided by the months held. On that rule the pullet supplier's terms in the case cost 27.21% a year.

The stress test widens the single return column to nine cells: three plans, each in a normal, a bad and a worst year. Plan B survives a bad year with TZS 5,375,332 of its reserve intact, but a worst year would still cost it TZS 90,506,313 of equity, with the home untouched. Plan C does not pass the same test. The book closes with a one-page Leverage Rule of six limits, written before signing and checked every quarter, so the owner decides on the worst year before the lender describes the best one.

Book 4: lending as a lawful group

Successful Money Lending Models in Africa follows thirty women traders at the central market in Tabora who have saved and lent together every Thursday since 2022, without registration, on memory and one exercise book. Their last cycle lent TZS 31,200,000 at 10% a month flat, charged TZS 9,360,000 in interest, wrote off TZS 6,900,000 and paid each member TZS 442,000 for TZS 420,000 saved. The book follows their next cycle, from November 2026 to October 2027, as a registered Tier 4 community microfinance group, set beside a SACCOS, a microfinance company and a digital lender.

The legal anchor is the Microfinance Act, 2018. It sorts lenders into four tiers and forbids microfinance business without a licence under section 16(1), and the book states an unlicensed Tier 2 fine of TZS 20,000,000 to TZS 100,000,000. The group's registration file follows GN 678, and its loan standard follows GN 679 of 2019. Under regulation 39(2) of that standard, each loan sits on one signed page showing the amount, both rates, the repayment schedule, the total, the penalty, the security and the signatures. The group adopts that list as its own standard. The book notes that registration costs TZS 150,000 in its illustration, which the reader must confirm with the authority.

Pricing is built from four parts: the members' own money, running costs, a loss allowance and a surplus. The group sets its rate so that each loan covers those parts. At a flat 5% a month the group charges 15.00 per TZS 100 lent, and the book shows how that rate is built from the four cost parts and turned into an effective annual figure. Loan size comes from the borrower's cash, not the amount asked. A dried-fish trader who asks for TZS 1,000,000, on a stall that leaves TZS 450,000 a month after stock, costs and household needs, would give up 85.19% of her spare cash over three months at the group's 5% flat rate. The savings cap of TZS 720,000 would take 61.33%. The loan is reduced to a size the trade can carry.

Arrears are counted every week, and provision is made for loss before any surplus is shared. After a market fire on 2 May 2027, the group's count from 3 June showed that at 31 August TZS 3,870,900 was unpaid on the March, April and May loans. That was 40.19% of the TZS 9,632,401 still owed, classified by the oldest missed instalment under the whole-balance rule. The final share-out is savings plus interest actually collected, less costs and losses. In the base case the box holds TZS 18,230,860 on 28 October 2027, or TZS 607,695 a member, a gain of 26.60% on TZS 480,000 saved. In the fire case the box holds TZS 14,068,000. Both figures belong to the illustration and are not a forecast for any real group.

Book 5: arranging the whole of what is owned

Never One Basket starts from a count. Its illustrative investor is a dental surgeon in Mbeya, aged 47, who opened a two-chair clinic in 2014. She owns a home, the clinic, an idle plot on the edge of Dodoma city, shares of one listed company bought on a tip, and TZS 150,000,000 in the bank. Outside the home and a reserve of six months' spending, her portfolio is TZS 288,000,000, and 87.50% of it sits in two places, the clinic and the bank. A second branch would put 87.50% into one business in one city, and one 40% shock would take TZS 100,800,000, or 35.00% of the portfolio.

The book's rule sets four baskets, business, land, bonds and shares, each with a written purpose, a target weight and a band. The business basket is capped. The clinic is valued plainly, at TZS 120,000,000 on its profit after market pay for the owner, times four, rather than at its headline earnings, so that the bad year is priced into the basket. The land basket must earn rent or have a dated use. The Dodoma plot, carried at TZS 30,000,000 against a cost of TZS 12,000,000 in 2019 and paying nothing, is turned into four rooms at a cost of TZS 26,000,000 in the case. The bond basket dates each Treasury bond to a goal, such as a university fee, clinic equipment or income from age 62. The shares basket replaces a single tip, down 25.0% from its 2021 cost, with new money spread across listed companies and an approved unit trust.

A portfolio is kept in shape by its weights, not its picks. The rule sends new cash to the basket furthest below its target, and the weights are counted once a year in October by a written method. The loss case is printed for each basket: a bond sold early, a plot that stays empty, a business that fails in a bad year, and shares that fall while the rest of the portfolio rises. The model portfolio is an illustration built from stated inputs, not the author's holdings, and no basket is recommended.

Common errors the series corrects

The series names six habits that cost households most. The first is treating a still number as a safe one. Losses weigh more than gains, so an account that never falls looks safer than it is, and the book writes the inflation loss in the same column as every other loss. The second is quoting the headline rate. A 7.0% return on a plot falls to 5.50% after costs, tax and land rent, and to 1.44% after inflation in the book's case. The third is paying the asking price because it sounds reasonable, when the margin of safety has not been set. The fourth is borrowing against the family home on the best year alone, when the worst year decides the outcome.

The fifth error is lending from memory. A group that counts its loans from recollection cannot know what has come back, and a surplus shared from memory is a surplus that does not exist. The sixth is holding everything in one place. A business, a plot and a tip can look diversified on a list and still rest on one economy, one city and one set of decisions. Each book corrects one or more of these habits with arithmetic the reader can repeat without help.

The rule of 70 is used as a quick check. At 4% inflation, the buying power of idle money halves in about 17.5 years, and at 6.8% in about 10.3 years. The book pairs that check with the real return formula so that the reader can see both the speed of the loss and its size.

Reading order and where to start

Book 1 is the shared starting point for any reader with idle money, because it sets the emergency fund and the real return method used throughout the series. Book 2 then asks what to pay, and Book 5 asks how the whole arrangement should look once several decisions are in place. Books 3 and 4 apply to a reader who borrows, lends or is offered someone else's money. A reader offered a loan should read Book 3 before signing anything. A reader who runs or joins a savings group should read Book 4 before the first loan is made.

Each book can be read on its own, but the margin of safety in Book 2 and the weights in Book 5 both assume that the emergency fund from Book 1 is already set aside. The Secret to Sound Investment, which carries Handbook number 39, is Book 2 and is read second, after Book 1 and before The Leverage Secret.

Mainland and Zanzibar

The five books concern Mainland Tanzania. None of them covers Zanzibar, and each says so in its scope note. The Mainland rules the books rely on, including those on Treasury bills and bonds, withholding tax, capital gains on land, land rent and deposit insurance, are stated as at the dates in each book's appendix. A reader whose money, plot, business or lending group sits under Zanzibar's own authorities should take the figures in these books as an illustration of method only, and confirm the local law and fees before acting. The two jurisdictions are not interchangeable, and the series does not treat them as one.

How the first year is organised

Each book closes with a dated first-year part that follows its illustrative subject week by week. Book 1 runs from counting the idle money on Monday 12 October 2026 to the first yearly review on Monday 11 October 2027. Book 2 runs from Monday 12 October 2026 to the first review on Monday 11 October 2027, with decisions taken in sequence. Book 3 runs from three plans on one page on Monday 12 October 2026 to the Leverage Rule re-signed by the owner and spouse. Book 4 runs from the vote to register on Thursday 8 October 2026 to the share-out on Thursday 28 October 2027. Book 5 closes with the October review, which holds the weights to the rule.

These timelines are teaching devices. They show the reader the order of decisions, the points at which a figure must be checked, and the moment when a rule is applied rather than argued. The reader can copy the calendar, but must confirm every date that depends on a public body's notice.

What the books do not settle

The series recommends no bank, fund, share, bond issue, dealer or plot, and makes no call to buy or sell. Firms appear as categories. Public bodies are named for what they do: the Bank of Tanzania, the Capital Markets and Securities Authority, the Dar es Salaam Stock Exchange, the Tanzania Revenue Authority, the National Bureau of Statistics, the Deposit Insurance Board and the Ministry of Lands. The books print no current yield, price, commission, fee schedule or land rent rate. Where one matters, the reader is pointed to the issuer, and a real coupon or index appears only as read at its source, dated and labelled historical.

The return assumptions are never forecasts. A 4% inflation rate, a 2.5% savings rate, an 18% loan rate and a 40% shock are chosen to show how the arithmetic behaves, and each can be replaced with the reader's own figures in the tools provided. The series is financial education under the Bank of Tanzania Certified Financial Educator standard, and it does not replace advice from a licensed adviser.

Figures to confirm before acting

Several figures in the series are dated and will move. The books state their positions at 10 October 2026 on the cover page. The appendix on Mainland tax, land and deposit rules is stated at 7 October 2026, and the Bank of Tanzania and lending readings are taken on 8 October 2026. A reader should note these dates and treat none of the figures as current without checking.

The deposit cover is the first figure to confirm. Book 1 states TZS 7,500,000 per depositor per bank, and the reader should confirm with the Deposit Insurance Board how that cover applies to a joint or business account. The Treasury bill and bond yields, and the 10% withholding rate used in the examples, should be confirmed with the Bank of Tanzania tender results and the Tanzania Revenue Authority. The land rent and any capital gains treatment in Book 1 and Book 5 must be confirmed with the Ministry of Lands before a plot is priced.

In Book 4 the registration cost of TZS 150,000, the Tier 2 threshold and the penalty range should be confirmed with the authority named in the Microfinance Act, 2018, and with the Bank of Tanzania for the digital lending guidance cited. The guidance on loan terms and the twelve items required under GN 679 should be checked against the current regulation before a group adopts them. In Book 3 the Bank of Tanzania's published average lending rate must be read at its source for the month cited, and the 18% loan rate used in the farm case must be replaced by a rate the reader is actually offered.

Finally, the inflation figures should be read from the National Bureau of Statistics at the latest release. The series uses the year to February 2026 at 3.15% as an anchor, and a reader working in a later month should replace it with the newest published index before running the Inflation Erosion Table or the real return calculation.

Land and Legacy is the shelf for families who want to own land safely and pass what they own to the next generation intact. It has five titles. Two cover land: how a plot is bought, checked and kept (Handbook 37), and how a surplus is spread across several plots as a land bank (Handbook 28). Three cover the household and the estate: how spouses share money without a quarrel (Handbook 33), how a family protects its breadwinner with life cover (Handbook 27), and how an estate is counted, willed and administered so that it reaches the heirs (Handbook 32). Read in this order, the series moves from the first plot to the last will. Every title applies Mainland Tanzania law only. Zanzibar is not covered, and no book in the series describes Zanzibar land, succession or family law. Each book states the date on which it read the law at source, 7 or 8 October 2026, and says that it is not legal advice on any particular plot, estate or transaction. Where a fee, rate or timetable could not be fixed from a source the book names, the book says confirm and prints no figure. The reader should treat every figure as an illustration tied to a stated case, and confirm it with the issuer named beside it before acting.

What the five books share

Each book in the series answers one question that a family or a couple faces with money, land or heirs. The subjects differ, but the books share a method that the series states at the front of every title. A decision is built from the household's own figures, each computed from inputs printed beside it. Every rule carries the name of its issuer and a date. Illustrative cases are described by type, and every worked example says that it is an example. Each book ends with a kit of tools to fill in at a desk.

The method matters more than any single figure. A figure printed without its source is not one the reader can rely on, and a rule printed without its date may already have changed. A reader who follows the method can check a book's conclusion against the same issuer and reach the same answer. That is why the series prints so many confirm with notes. They are not a weakness of the books. They mark the points where the reader must check the position before paying.

The series also carries one warning throughout. The most expensive mistakes happen before any money moves: a plot bought without a search, a guarantee signed without disclosure, a policy issued on an answer that was not true, or a will that was never written. The books therefore spend more pages on the order of steps than on the arithmetic, because the order is what protects the family.

Where to start

A reader with a plot in view, or about to pay a deposit, should start with Handbook 37. It follows one family's purchase from search to title, and its checklists are the most direct guide in the series to a lawful residential plot in Mainland Tanzania. A reader with surplus money who is weighing several plots, or a business that wants to hold land as an asset, should go to Handbook 28, which turns the same discipline into a portfolio of up to five plots.

A couple whose money is spread across two names, a business and an informal savings group should start with Handbook 33. Its central tool is a single page showing what both spouses earn, own and owe. A household whose income depends on one earner should read Handbook 27 before any policy is bought, because it prices the gap first and only then chooses the cover.

The reader who has reached the stage of wills and heirs should read Handbook 32 last. Each book can be read alone, but the estate figures depend on the asset register, the land title and the cover that the other books help the family assemble. Most of the numbers the estate book uses were settled earlier in the series.

Buying a plot in the right order

The land books insist on a sequence that protects the buyer. In Handbook 37 a family of two spouses and three children considers three plots in Kibaha in October 2026. The plots differ in the paper they carry: surveyed general land with a Certificate of Right of Occupancy, unsurveyed village land with a Certificate of Customary Right of Occupancy, and a plot with no papers, sold by a seller who holds no letters of administration. The category of land sets the paper, the office that deals with it and the lawful route, so it is fixed before any comparison is made.

The official search is the decisive step. Under section 97 of the Land Registration Act, the search shows the registered owner, any mortgage, any caveat and the land rent position. In the Kibaha case plot A is clean of mortgage and caveat but carries TZS 108,000 of land rent arrears, which the seller must clear before the registry will act. The book's rule is that money leaves the bank only after the search, the consents, the seller's tax certificate and a signed agreement are in hand. A photocopy of a title proves only that a copy was made.

Costs belong in the decision. Plot A's price of TZS 18,000,000 carries costs of TZS 1,063,250, an outlay of TZS 19,063,250, which the family reaches from a saving of TZS 550,000 a month. A double sale would take the whole outlay, the equivalent of 34.7 months of saving. Stamp duty falls due within 30 days of the transfer. The stamp duty bands, registry fees and search fees are not printed in the book and are to be confirmed with the issuer, so the reader should not carry the book's figures for them into a budget.

Handbook 28 applies the same order to a portfolio. Its pharmacist fund, capped at a quarter of net worth, weighs five offers. Scored on published evidence, three offers pass the rule for titled plots and one passes as a long-hold village plot. The official searches then remove one offer, because its registered owner is someone other than the seller, so that deposit is never paid. A road authority line places 300 square metres of another offer inside the road reserve, and the seller will not price the remaining 700 square metres on their own. Two grants are under three years old, so each sale needs the Commissioner's approval under section 37(2) of the Land Act, and village land passing to a non-resident needs council approval after 60 days' notice under the Village Land Act. The lesson is that a price per square metre must be compared all in, on the usable area, and not on the figure a seller quotes for the whole plot.

Keeping land and selling it on a rule

Owning the paper is not the same as keeping the plot. A right carries conditions. A building grant needs planning consent and a permit applied for within six months under section 34 of the Land Act, and a breach can lead to revocation after a warning letter, although remedying the breach ends the liability. Bare land is not free to hold. Handbook 28's pharmacist budgets TZS 10,720,000 over ten years for land rent, clearing, a caretaker, a watchman and visits, paid from the reserve in the first years of holding. The rates on bare land and on houses are confirmed with the council, and the book does not assume them.

Handbook 37 turns the same point into a choice between building in stages and holding the plot empty. The Kibaha family builds a core house of TZS 28,000,000 from cash, saved at TZS 550,000 a month over 51 months, and moves in before the finish is paid for. The book values the empty plot after ten years at TZS 24,190,495 at an assumed 3% a year, TZS 38,860,650 at 8% and TZS 55,905,268 at 12%, and at nothing in the loss case, while rent continues to be paid. These rates are the book's assumptions, not forecasts, and the reader should set them against the family's own alternatives.

A sale is also a rule written in advance. In Handbook 28 plot C is sold before its period ends, to a citizen builder who takes on its conditions, with the Commissioner's approval and the seller's tax certificate in hand. A headline gain of TZS 6,750,000 nets TZS 2,581,200 after costs and tax, which falls short of the investor's hurdle. The book states the seller's tax as 10% of the gain for a resident, 30% for a non-resident and 3% of value where no records exist. Those rates are stated as read on 7 October 2026 and must be confirmed with TRA before any sale.

Money inside a marriage

Handbook 33 starts from a fact most couples never state: each spouse knows their own share of the household's money, and nobody holds the whole. In the Njombe case the dairy made a profit of TZS 39,466,000 in the year to 30 September 2026, yet the business account fell by TZS 2,434,000 and TZS 7,200,000 left the household untraced. The first step is a money map and a net worth sheet, filled together from statements. The sheet in the case showed a net worth of TZS 142,066,000, untraced headings that belonged to neither spouse, and a guarantee and a savings group known to only one of them.

The book then separates three pools of money: the business's money, each spouse's pay and each spouse's own property. The dairy pays two equal salaries of TZS 1,200,000 a month into separate accounts and keeps the rest for the business, the emergency fund and the guarantee instalments. The household budget starts from those salaries, allocates every shilling, and gives each spouse an equal personal allowance of TZS 125,000 a month. The legal frame is in the Law of Marriage Act. Section 56 gives a married woman the right to hold property and contract in her own name, and section 58 provides that marriage does not merge the property of spouses. Section 60 presumes that property in one name belongs to that spouse, which is why the book asks the couple to put both names on the assets they build together.

Disclosure runs both ways. Each spouse discloses every debt, guarantee, savings group and family promise once a year on a signed sheet, read aloud and questioned for matters of fact only. Any spend, loan, sale or commitment above TZS 500,000, and any guarantee at all, waits 48 hours for both signatures. In the year of record, five of eight decisions above that threshold, worth 88.46% of their value, were taken by one spouse alone. The couple's rule applies the threshold to the plan, and each spouse keeps full control of the side of the plan allocated to them. A guarantee of TZS 6,000,000 for a cousin's loan, with TZS 4,800,000 called, shows why a guarantee belongs on the sheet on the day it is signed, not the day it is called.

Property and protection come last. The book's plan year moves the home into joint occupancy, places the dairy in a partnership deed drawn by an advocate, builds an emergency fund of TZS 7,050,000 in both names, buys family health cover and equal life cover of TZS 72,000,000 for each spouse, and ends with two signed wills. Section 59 protects the matrimonial home from sale or mortgage without the spouse's consent while the marriage subsists, and section 114 divides jointly earned assets on separation or divorce. The handbook also cites a 1983 Court of Appeal decision that treats domestic work as contribution. That reading should be confirmed with an advocate before a couple relies on it.

Protecting the breadwinner

Handbook 27 begins with a need, not a product. Its contractor earns most of a household income of TZS 3,450,000 a month. If he died, his own spending and a loan instalment would stop, but his support to his mother and the household's living costs would continue, leaving a monthly gap of TZS 1,700,000 for 14 years, to October 2040, when the youngest child reaches eighteen. Valued at an assumed 3% real return, that gap is TZS 230,439,892 in today's money. With lump needs of TZS 67,400,000, the family's need is TZS 297,839,892. At 0% real the gap alone would be TZS 55,160,108 larger. The sum of cover is settled first, and the type of policy is chosen afterwards.

The first count shows what the family already holds. A death in October 2026 would pay out TZS 37,000,000 in total, from savings, a group share, a van and tools, a funeral group and credit life cover, but the credit life cover pays the bank, leaving TZS 19,000,000 of the family's own money. That lasts 8.8 months at the household's need. The old employer's group cover ended in December 2021. NSSF's survivors' pension on the record shown is nil, subject to NSSF's written confirmation. Each source is counted by what it would pay this month, to whom, and when it ends, and a forced sale of business assets is counted at less than its estimate.

The type of cover is a separate decision. At the household's premium of TZS 2,070,000 a year, term cover buys about 20.6 times the cover that an endowment buys for the same premium, and it returns nothing if the breadwinner lives. The book's endowment example returns TZS 35,525,000 for TZS 30,450,000 paid in, a nominal gain of TZS 5,075,000 that becomes a real shortfall of TZS 3,747,369 at 4% inflation, and it loses value if surrendered early. The household chooses a ladder of term cover of TZS 150,000,000 for 20 years and TZS 120,000,000 for 10 years, covering the breadwinner first and then the spouse. All three policies cost TZS 1,246,000 a year, 3.01% of income, inside the household's own ceiling of 5%.

Disclosure is where claims are lost. The form, not the agent's promise, is what the claims officer reads, and an answer short of the truth is found years later, at the claim. In the case the contractor declares a treated medical condition and accepts a loading that adds TZS 463,500 a year to his premium. The book calls the loading the price of a policy issued in full knowledge, and sets it against a claim of TZS 270,000,000.

A policy also has to pay. A policy pays only when it is in force on the day of death and names a beneficiary able to receive the money. The household names the wife on the contractor's policies and the contractor on hers, names a guardian for the children in writing, and confirms with TIRA and the insurer how a minor's share is paid and how section 101 protects a policy from creditors. Its claim file lists every death document, where it is kept and who fetches it, the insurer's claims contact, and the clock of 48 hours, 7 days and 45 days, with the route for complaint. It fixes in advance how the payout is divided, between education, the funeral and estate, a six-month buffer and an invested balance. Each October the family tests the need against the cover.

Passing the estate on intact

Handbook 32 starts with a count, because no estate can be planned until it is known. The Mbeya husband owns TZS 1,359,600,000 in his sole name, including a home, mill shares, a shop building, a farm, vehicles, deposits, a bond and listed shares. He owes a car loan of TZS 12,000,000 and has personally guaranteed a mill loan of TZS 180,000,000. His wife's assets sit outside the estate. Only 3.53% of the gross value is cash in his name. The register records how each asset passes, whether by will, by survivorship, by a named beneficiary or under the company's articles, and it lists every claimant, including a child born outside the marriage, whose share is decided by the governing law and the will.

The governing law comes next. Section 88 of the Probate and Administration of Estates Act points to customary, Islamic or Christian law according to what the deceased professed and intended, read from declarations, acts and the way he lived. Without a will the court appoints an administrator, and a caveat can hold the estate for four months at a time. On the book's illustrative assumptions, a contest lasting about 30 months would cost the family TZS 225,954,000, 16.62% of the gross estate, against a planned cost of TZS 25,822,500. The book presents none of the three systems as better. It asks the reader to confirm succession shares with an advocate.

A valid will has to meet the form the law requires. The book sets out the requirements under sections 88 and 89 of the Probate and Administration of Estates Act and the stated requirements of RITA, including the signature, the date, the statement of a sound mind and free will, and two witnesses, one a relative and one not. Probate goes only to a named executor, and a minor's legacy goes to the Public Trustee or the court unless the will directs otherwise. The Mbeya will divides a pool of TZS 561,600,000 equally among five children, holds the youngest share in trust until 18, and gives the wife the vehicles, the deposits and the shop rent for life. A will that is set aside returns the family to a contest.

Cash comes before the grant. The first year after a death needs cash that an estate may not hold in a usable form. In the case the only cash in the husband's name, TZS 48,000,000, sits in a sole account that releases money only to the holder of a grant, and the first year needs TZS 116,394,000. The book's answer is ten-year term cover of TZS 200,000,000 naming the wife, at an assumed premium of TZS 4,800,000 a year, together with a joint household account. Those two measures change a shortfall of TZS 68,394,000 into a surplus of TZS 98,606,000, before any grant.

The business has its own test. The mill runs on one signature, and the book's model shows its earnings falling sharply in the months after a founder's death without a plan. The plan adds a second and third signatory, an emergency mandate to keep wages and maize purchases going, and a board that chooses a successor against written criteria. The will divides shares by value, while the shareholders' agreement decides how the company is run and who can be replaced.

The executor's year is the last test. The book runs it from the death, through the grant, the inventory, the distribution and the final account, here about sixteen months. It names the executor for the work rather than by birth order, and a reserve from the other side of the family. A family meeting held in the founder's lifetime tells the children the plan from the founder directly, so that they do not first hear it at a funeral. A caveat that runs four months costs money every month it runs, and a meeting held early answers most of the questions that cause one.

Figures to confirm before acting

Every book in the series marks figures that depend on an issuer and a date. Four groups need the most care. The first is the statutory position. Each book relies on Mainland Acts read at source on 7 or 8 October 2026, and the land books note that the current revised text of the Land Act is confirmed with the issuer before any step is taken. The second is fees and charges. The search fee budget of TZS 30,000 a search is a model figure. Registry fees, stamp duty bands, RITA's custody fee for a will and land rent per plot are all to be confirmed with the issuer named beside them.

The third group is tax. The seller's tax on a land gain is stated as read on the date of the book and is confirmed with TRA before a sale. The fourth is insurance and pensions. The contractor's NSSF survivors' pension is shown as nil on the record used, and the book marks any benefit not confirmed in writing. The Insurance Ombudsman's current limits, the grace and free-look periods, and the way a minor's share and section 101 creditor protection apply must be confirmed with TIRA and the insurer. Pension rules for public servants are confirmed with PSSSF. Where the books use an assumed premium, rate or return, the figure is a model input and not a quotation.

The series does not cover Zanzibar, and it does not replace advice. Where a matter turns on a dispute, a contested will or a disputed fee, the reader should take the position to an advocate, to the issuer named in the book, or to the agency concerned. The books apply one test to every figure: a figure without an issuer and a date is not printed. The reader should apply the same test to any figure carried forward into a plan.

Know the Ground is a series of four handbooks for an owner who has money to place, stock to protect or a town to choose. Each title is written for a decision that is costly to reverse: where to put a large sum, how to stop losses that never appear as a single line, how to guard a site that holds goods, and which place to live and trade in. The four books share one structure, seven steps, a closing decision and twelve tools to fill in on paper. Every case in the series is illustrative, built from inputs the books state, and the figures describe those cases rather than any real business. The setting is the Mainland throughout. Zanzibar has its own investment, revenue and licensing rules, and the books say so wherever a Zanzibar point could change a result. This summary explains what each book covers, the method they share, the decisions they support and the figures most likely to change. It is general information, not investment, legal or tax advice. Every rate, fee and list should be confirmed with the issuer named beside it before anyone acts.

What the four books cover

The series has four titles. Tanzania Business Opportunities by Sector 2026 (Handbook 44) asks where a large sum should go. Doing Business Safely in Tanzania (Handbook 45) asks how a business stops the losses that do not show in its accounts until they are large. Site Security in Tanzania (Handbook 46) asks how a place that holds stock is guarded and counted. Where to Live and Trade in Tanzania (Handbook 47) asks which town a family or business should base itself in. The first two concern money and goods at risk. The last two concern a place and the people and customers who depend on it.

Each title runs to between 159 and 170 pages and follows the same build. A short opening states the case and the question it must answer. Seven chapters each carry one step, and each ends with a task for the same evening, tied to one tool. Twelve tools are one-page forms, with the case's figures in one column and an empty column for the reader's own. The books close with a dated account of the case in practice, a kit of tools, questions readers ask, a glossary and a Kiswahili summary.

The four cases are these. Handbook 44 follows a family company in Mwanza with TZS 3,000,000,000 to place. Handbook 45 follows a solar equipment distributor in Arusha with TZS 3,600,000,000 of annual sales. Handbook 46 follows a building materials warehouse in Mwanza that lost stock while it paid for guards. Handbook 47 follows a returning family with TZS 600,000,000 that had assumed Dar es Salaam without comparing it to any other place. Where a book quotes a shilling figure for its case, the figure is calculated from the stated inputs and the working is shown in the text.

The sequence below is a suggestion, not a requirement. A reader with capital to place can start with Handbook 44. A reader who already trades can start with Handbook 45 for the losses and Handbook 46 for the site. A reader planning a move can read Handbook 47 before signing any lease. Each book stands on its own, though the method in Handbook 44 is the one the others use.

One method: proof before money moves

The books share one rule. A decision about money, stock, land or a town is tested on written evidence before any payment, deposit or lease. Handbook 44 states the point plainly: the best sector is not the best project, because a sector can grow fast and still offer no project the investor can run, finance and survive. Handbook 45 applies the same discipline to losses, where a control is judged by the record it leaves behind. Handbook 46 applies it to guards and cameras, which are bought only after the site has been rated.

Each book prints a stress or loss case beside the base case. Handbook 44 runs its three projects at an 18% hurdle in both cases. Handbook 47 prices a loss case for each finalist beside its plan. Handbook 46 sets a plan gain against its cost and prints the shortfall that arises if only half the gain arrives. Handbook 45 writes each risk into a loss register and prices it before and after the controls. A single case on its own is a forecast, not a test, so the two columns should be read together.

Figures are dated and carry their source and base year. Handbook 44 separates share from real growth, and it warns that a sector's share can fall while its output rises. Handbook 47 uses the 2022 Census as its count of people and households and treats it as a count made in that year. Where a rule, fee or rate could not be read at its issuer, the books say to confirm with the issuer and print no figure. Handbook 44 marks an approval as read only where the instrument, section and date are stated, and every other approval as still to be confirmed.

The method leaves a filled page rather than a list of ideas. Each chapter's task is tied to one tool, so that the reader ends the evening with a number, a name or a dated decision on paper. The tools are designed to be reused: the same loss register, the same callback rule and the same monthly page can serve a business for several years.

Information is rarely shared evenly. Handbook 44 notes that where sellers and introducers know more than the buyer, the buyer pays either to verify or for the gap in knowledge, and the checks are the cheaper of the two. Its checks cover title and survey, valuation, a seed supply study of 40 farmer groups and six traders, a laboratory test, a lawyer, an environmental screening and an engineer, costing TZS 41,500,000 in all, or 1.65% of the mill's capital need. Each check has a written walk-away finding, signed before the result returns, so that a result cannot be argued into a purchase after it arrives. The introducer and the seller are checked as carefully as the plot, and the book asks who pays the introducer.

Choosing a sector and a project

The case in Handbook 44 is a Mwanza family company with twenty years in fish processing and haulage. It sold two trucks and a cold room in 2026 and holds TZS 3,000,000,000 for one new business on the Mainland. At the Bank of Tanzania average rate for August 2026 that is about USD 1,129,089, an illustration only. The book runs seven steps from the sector map to a staged decision: reading the map, scoring the sectors and counting the demand, mapping the entry route, pricing the capital and the return, stressing every number, running the checks and choosing.

The family fixed six weights before scoring any sector: demand evidence 25, fit with skills 20, capital 15, entry route 10, risk 15 and speed of cash return 15. Agro-processing scored 400 of 500, construction materials 380 and transport 350, while hospitality scored 250 and ranked ninth. Transport was set aside, not because its score was weak but because the family's remaining money was already in haulage. The shortlist became three projects: a sunflower oil mill in Singida, a 24-room hotel in Dodoma and a concrete block plant in Dodoma. Demand is counted by named buyers. The mill's 3,000 tonnes of seed is 1.09% of the 276,202 tonnes of sunflower counted for 2023/24, and the hotel needs 27.2% occupancy to cover its fixed costs alone.

In the base case at the 18% hurdle the mill shows a net present value of TZS 231,805,294, an IRR of 21.17% and a payback of 4.94 years. The block plant has an NPV of TZS 76,352,933, and the hotel falls short by TZS 1,090,610,168. The hurdle is the family's choice, set above the August 2026 average lending rate of 15.05%, and the book notes that no borrower is promised that rate. In the stress case all three projects fall short, the mill by TZS 1,841,192,361. The squeeze that matters combines a seed price up 15%, an oil price down 5% and a plant cost 28% over plan. That cuts full-year EBITDA from TZS 825,000,000 to TZS 132,000,000.

The family chooses the mill. Its approvals follow a common spine of BRELA, then the Tax Identification Number, then the licence, and TRA asks for a lease or title before it issues the TIN, so the land settles first. The mill needs eight approvals before its first sale, six read at the issuer and two still to be confirmed. The plan commits TZS 1,992,000,000 in April 2027, holds TZS 528,000,000 for the second stage, spends TZS 41,500,000 on seven checks and keeps TZS 438,500,000 in reserve. Those four sums add to TZS 3,000,000,000. The reserve pays only for what went wrong, never for what was planned.

Two points need confirmation before any step is taken. The list of activities reserved to citizens sits in GN No. 487A of 28 July 2025, and the book read it in a law firm's summary, so it should be confirmed in the Gazette. Seed buying from farmer groups falls under the reserved activity of on-farm crop buying, so a lawyer must confirm the position for any company with a non-citizen shareholder. The TISEZA thresholds were set under the 2022 Act and should be confirmed with TISEZA, and the hotel's tourism levy under section 59 of the Tourism Act, 2008 should be confirmed with the Board.

The reserve is sized against the overrun, not the hope. The book uses the average plant overrun of 28% from its transport study: on a plant budget of TZS 1,200,000,000 that is TZS 336,000,000, which the reserve of TZS 438,500,000 can absorb once and not twice. One point of oil yield is worth TZS 118,500,000 of EBITDA a year, and two points below plan turn the mill's NPV of TZS 231,805,294 into a shortfall of TZS 320,460,486. The three triggers the family signed, on seed, oil and the plant cost, tell the board when to slow or stop. A seasonal plant that misses one season loses a year, and the book prices that loss at TZS 218,410,977 off the mill's value.

Stopping the losses that do not show

The distributor in Handbook 45 sells TZS 3,600,000,000 a year and should keep TZS 195,000,000 of profit. In the year to 30 September 2026 it lost TZS 247,654,000, which is 6.88% of sales and 127.0% of the planned profit, leaving a shortfall of TZS 52,654,000. No single event explains that. Payment fraud, inside losses, a land deposit, cyber incidents and a dealer who stopped paying each took a share. The first step writes every loss into a register that estimates the expected loss for the year ahead. That register puts the expected loss at TZS 163,074,000 and identifies the two largest risks: a plot bought from a seller who did not own it, at TZS 37,500,000, and a supplier payment sent to a changed account, at TZS 27,000,000. Together they hold 39.55% of the expected loss.

Payment fraud cost TZS 83,150,000 in the year. Of that, TZS 66,500,000 net was an inverter invoice paid on 12 March 2026 to an account named in a changed-account email, and TZS 16,650,000 went on nine forged confirmations at the counter. Settlement is final under the National Payment Systems Act, so the business can ask only for a recall of what remains. On the book's illustrative trail the recall held TZS 21,000,000 at hour 26, against up to TZS 80,000,000 at hour 2. The control the book sets is a callback on a number already on file for every change of bank details, two approvers above TZS 5,000,000 and for every new payee, and no counter goods released until the money shows in the statement.

Inside losses cost TZS 44,424,000: stock at cost of TZS 25,272,000, refunds on 31 transactions with no goods returned of TZS 7,812,000, and 3,600 litres of fuel that the kilometres do not explain, worth TZS 11,340,000. The controls proposed are split duties, a monthly count with two signatures, refunds only against returned goods and a weekly fuel log. They cost TZS 12,600,000 a year and are expected to remove TZS 32,835,600 of loss. The land rule is simple: no deposit before the official search, the survey and the advocate's written opinion, and any deposit held by the advocate. The distributor's TZS 45,000,000 deposit was paid on 20 May 2026 before any search, and the search later named another registered owner and a mortgage.

Cyber incidents cost TZS 15,380,000 in the year, of which the ransomware week, from 15 to 19 June 2026, accounted for TZS 13,230,000 in lost gross profit, recovery and overtime. Two-step sign-in, two offline drives rotated weekly, an off-site encrypted copy and a quarterly restore test cost TZS 7,800,000 a year and cut the worst case from TZS 40,000,000 to TZS 10,000,000. Counterparty losses of TZS 57,300,000 included a dealer whose agreement was signed by a manager who was not a director. The fix is a counterparty file opened before the first credit sale, a check of who may sign, and written limits. The First-Hour Card, the book's closing tool, starts on suspicion rather than certainty. It names the bank's fraud line first, then the Police and the other bodies in order, and it notes that a complaint about a bank goes to the bank first and then to the Bank of Tanzania.

The distributor's year was turned into a monthly page of seven measures, each counted from a record the business already keeps, from the callback log to the dealer ledger. Each measure has a baseline from the year to September 2026, a monthly figure and a target. The book's reason is that controls decay: the callback is skipped for a supplier everyone knows, the count is moved for a delivery and the backup drive stays plugged in. The page takes sixty minutes to read on the first Tuesday of each month. On the book's plan the controls remove an expected TZS 112,165,600 a year net of their cost, and the bad-year case of TZS 197,908,400 is printed beside that figure so that the reader sees both.

Guarding a site that holds stock

The warehouse in Handbook 46 is a building materials store in Mwanza. It paid TZS 17,280,000 a year for three guards and an old camera recorder, and in its year of record it lost TZS 65,675,000. Against revenue of TZS 4,256,000,000 that is 1.54%, but against operating profit before losses of TZS 191,320,000 it is 34.33%. The book starts with a site rating, a walk by day and by night, and a register that scores eight threats for likelihood and impact before any guard, camera or contract is bought. The gate scored 20 and the back fence 16, the two highest.

The security company is chosen on evidence rather than price alone. Three companies priced the same package of 36 post hours a day, with monthly bids from TZS 1,950,000 to TZS 2,900,000. The bids were scored 70% on quality across six criteria and 30% on price. The winner scored 78.79, against 74.42 for the next bid and 55.20 for the cheapest. The quality criteria cover vetting, training, supervision, response and guard pay, and each bidder had to state the guard's monthly wage and show a sample payslip with the name removed. The contract sets out posts, replacement within two hours and reporting. A guard on contract has no power to arrest, detain or search beyond what the law allows any person. Any suspect is handed to the Police, and any bag check follows a written policy agreed in advance.

The camera plan places twelve cameras against the eight threats, five of them on the gate and loading bay, and keeps 7,200 GB of recordings for 30 days. Cameras that record faces, plates and movements collect personal data, which the Personal Data Protection Act, 2022 (Act No. 11 of 2022) governs on the Mainland, and section 14(1) of that Act is the basis for the registration the book requires with the PDPC. Cash that sat in a drawer until 18:00 had left the counter short on 46 days, a loss of TZS 4,830,000. From 1 December 2026 the drawer holds under TZS 1,000,000, and drops go to a safe that two keys open. Stock is counted from the gate to the count. The count on 30 September 2026 found TZS 49,073,000 short, including 1,240 bags of cement worth TZS 20,336,000.

The investigation of the stock loss is run under a fair process. A trace of the year's documents found four routes: 640 bags on 32 delivery trips at 20 bags over the invoice, 300 through the back fence at night, 180 receipted at the counter but short-delivered, and 120 torn bags discarded with no record. The book's rule is that a loss is a control failure until a fair hearing finds otherwise, and that a document trail does not by itself show who intended a loss. Reporting a loss is protected, and a confirmed find earns a fixed reward. The plan gains TZS 51,775,000 in a year at a cost of TZS 26,499,000, a net TZS 25,276,000. If only half of the gain arrives, the year ends with a shortfall of TZS 611,500, and the book prints that figure beside the plan.

Controls are checked where a known date cannot reach them. The warehouse's monthly audit scores every control out of 100 from its records on the first Monday of the month, and surprise counts and a quarterly night visit test what a timetable would hide. The first surprise count, on Thursday 17 December 2026 at 14:10, found a drawer of TZS 640,000 with three drops logged and signed and no difference. Its value was that everyone at the counter knew the count could come at any hour. Every failure, whether a missed post, a lost item or a damaged cabinet, is written on one incident card that records the people involved, the Police reference, the scene, the recordings and the insurer's notice, so that the failure can be audited or defended at a hearing.

Choosing the town and the base

The family in Handbook 47 brings TZS 600,000,000 home to open a solar distribution business and had assumed Dar es Salaam because a relative named it at a wedding. The base brief replaces that assumption with five business needs, five household needs and four deal-breakers, and with a long list of ten places: Dar es Salaam, Dodoma, Arusha, Mwanza, Mbeya, Morogoro, Tanga, Mtwara, Kigoma and Zanzibar. The ten places hold 28,677,622 people, which is 46.45% of Tanzania's 61,741,120 at the 2022 Census. The family set five weights before any score was seen: market 30, cost 20, access 20, services 15 and safety 15, adding to 100.

Market is counted from the 2022 Census. Target households, meaning those without grid lighting and grid households that would buy a backup system, total 4,058,877 across the ten places. Morogoro has the most at 575,233, followed by Dodoma at 558,352. Dar es Salaam is the most populous place, with 5,383,728 people, but it holds 347,428 target households and ranks eighth, because 86.0% of its households already have grid lighting. The book is explicit about the limits of a regional count from 2022. It cannot show the far districts or whether a household can pay for a system, so the family checks each town with visits and calls.

Cost is priced on the same five lines in each place: house, shop, one worker, school fees, food and transport. A month of that basket cost TZS 13,650,000 in Dar es Salaam and TZS 6,560,000 in Morogoro, 2.08 times as much, which is TZS 85,080,000 a year on the basket and TZS 101,040,000 with eight staff. Access is priced on the road from the Dar es Salaam port, where every unit starts. At an illustrative TZS 60 a unit kilometre, the haulage is TZS 11,700 a unit in Morogoro and TZS 68,400 in Mwanza. Services are checked from the Census for power and water, and from the family's own calls and visits for schools and clinics. Grid lighting runs from 86.0% of households in Dar es Salaam to 16.6% in Mtwara, and piped water from 79.7% in Arusha to 32.8% in Kigoma.

Morogoro wins with 419 of 500 points and stays first on the family-first weights. Its model leaves TZS 294,921,000 after family costs over three years. At 60% of planned sales it runs short by TZS 31,559,400, which is covered by TZS 200,400,000 of working capital. Dar es Salaam ranks fifth with a base-case shortfall of TZS 316,494,000. It needs 429 sales a year to break even and plans 278 in year one. The move is tested first by a sixteen-night trial stay in Morogoro from 18 December 2026, in a short-let house, living as the family would live there. Zanzibar is one of the ten places, but its VAT, levy and licensing follow Zanzibar's own rules, which the Mainland figures do not show.

Scoring is done on points out of 500, not on averages, so that close places do not appear tied. Morogoro's services score of 2.3 costs it 40.5 points against Dar es Salaam's 5.0. Its market score of 5.0 against Dar es Salaam's 2.9 gains it 30 times the difference of 2.1, which is 63 points, and on market and services together Morogoro is ahead by 22.5 points before cost, access and safety are counted. The book adds a second set of weights chosen by the person most likely to disagree with the first, and it prints the findings that would reverse the choice before the decision is made. The same logic does not make the lowest-service place the best: Mtwara and Kigoma have large off-grid markets of 410,170 and 371,969 households, and their weak services are priced as a cost rather than set aside.

Figures to confirm before acting

The most useful habit across the four books is to check the figure before the decision. Each book dates its positions, most to October 2026, and asks the reader to confirm each rate, fee and list with the issuer before any payment, deposit or lease. The first items to check are the list of activities reserved to citizens under GN No. 487A of 28 July 2025, which should be confirmed in the Gazette, the TISEZA thresholds and fees, which should be confirmed with TISEZA, and the tourism levy under section 59 of the Tourism Act, 2008, which should be confirmed with the Board. The land search, the cost of each approval and the registration under the Personal Data Protection Act should also be confirmed with the issuer named in each book.

The lending rate of 15.05% for August 2026 is a Bank of Tanzania average, and Handbook 44 uses it only as the base for its hurdle. The hurdle itself is the family's choice, and a reader should set a hurdle for the reader's own case. Prices for seed, oil, rooms, blocks and rents are the authors' assumptions, not market quotes, and each must be replaced with a current quote before it is used. The census counts date from 2022, and the other market figures carry the years stated beside them, so a reader acting later should refresh them.

The books name no bank, fund, consultant, equipment maker or listed company, and they make no call to buy or sell. Handbook 47 says it recommends no provider. A provider's figure should be treated as a quote to test against the same case, and any claim about demand should be tested against a named buyer, a volume, a price and a source. The method is the same whether the sum is TZS 600,000,000 or TZS 3,000,000,000: write the case, price the loss beside the gain, and decide only when the evidence is on paper.

Minerals and Gemstones is a series of four Zatra Handbooks for people who must decide whether a mineral or gemstone business can be run lawfully and at a margin in Tanzania. Gold Trading in Tanzania (Handbook 48) covers the dealer's side of the gold sale. Small-Scale Gold Mining in Tanzania (Handbook 49) takes one reef from a sample to a working pit. Tanzanite and Gemstone Business (Handbook 50) follows a dealer and cutter through the coloured stone chain. Mineral Tourism in Tanzania (Handbook 51) asks how a tour company, lodge or gem shop can sell visits to mines and volcanoes. Each book stands alone and each covers Mainland Tanzania only. Zanzibar has its own law for minerals and tourism, and none of the four books applies Mainland rules there. The rules were read in October 2026, and each book states the date of its reading beside the instrument it relies on. The case businesses in the books are illustrative. Rates, fees and dates move, so the reader should confirm each one with the issuer named beside it before acting. This summary is general information, not legal, tax or investment advice.

What the series covers

The four books share one method. A mineral earns its value only when it passes through a licence, a lawful market, a weighed and tested lot, the State's share paid and a lawful buyer, with a record that reconciles every unit. Each book takes that chain in seven chapters, and each chapter ends with a task tied to one tool in the book's kit. Every gain is printed beside a loss case, meaning the price fall, delay or lost yield that would reverse it. None of the books forecasts prices, recommends a buyer, a laboratory, a bank or an equipment supplier, or calls for any purchase or sale of a mineral.

The four titles are written for different readers. Handbook 48 serves brokers, dealers and the investors who finance them. Handbook 49 serves Tanzanian citizens who hold or seek a primary mining licence for gold, with their partners. Handbook 50 serves gemstone dealers, cutters and jewellers. Handbook 51 serves tour operators, lodges, gem shops and village or miners' groups that want visitors on fair terms. The figures of one book should not be carried into another, because the prices, yields and costs differ by mineral and by role.

The Mainland legal base is shared across the books. The Mining Act, Cap. 123 is read in the TanzLII consolidation dated from 30 November 2019, which TanzLII marks as having outstanding amendments. The subsidiary instruments named in the books include GN No. 2 of 2018 (dealer and broker licence fees and the dealer's register), GN No. 84 of 2026 (royalty and fees paid before sale on gold), GN No. 539C of 2024 (the 20% domestic set-aside for gold) and GN No. 135 of 2019 (the Mirerani Controlled Area). Each book lists its rules in a local appendix with the issuer, the section and the date read. Where a rule or fee could not be read at its issuer, the book says so and asks the reader to confirm it.

The gold price used in the illustrations is a fixed model assumption, not a quotation. The Mining Commission publishes the daily gold price, and that published figure is the only one a business should use in practice. The books name no refinery, bank, insurer, laboratory, exporter, buyer or equipment maker, and they make no call to buy or sell any mineral.

Who may hold each licence

The first test is whether the reader may hold the licence at all, and Mainland law draws that line by citizenship and by role. Under section 18(1) of the Mining Act, only mineral right holders, licensed dealers and brokers, import permit holders, and their employees, agents or contractors may possess or dispose of minerals. Section 27C confines buying and selling to the Mineral and Gem Houses and recognised markets, or to buying stations where no market exists. Each book reads these two sections together before it reads anything else.

Citizenship is decisive for the primary and small-scale routes. A primary mining licence for gold may be held only by a Tanzanian citizen, or by a partnership or body corporate whose members are exclusively citizens, under section 8(2) and (3). A foreigner may be contracted for technical support that is not available in Tanzania, with the Commission's leave. A gemstone mining licence is granted only to Tanzanians under section 8(5). The small lapidary licence, which allows cutting, polishing and selling within Tanzania on three machines, is open to citizens only. The large lapidary licence is open to foreign holders as well, but it allows thirty machines for a foreign holder against ten for a Tanzanian, as read in sections 86C to 86M of the Act.

The dealer and broker licences are the commercial routes. A dealer may buy at the market, hold, sell and export, subject to the royalty, fee and set-aside rules, and it must have the strongroom and the register ready before the first gram. A broker may buy at named buying stations and sell to a licensed dealer, but it may not export. The gold book reads a broker licence as valid for twelve months and flags its eligibility conditions for confirmation with the Commission.

An investor who funds a licensed business is a separate question from the licence itself. The small-scale mining book keeps the citizens-only rule on the licence and on any transfer of it, and it records the partners' own agreement as a second layer of protection. A reader who is not a citizen will find that most mining and gemstone routes are closed to a personal licence and must be approached through a licensed Tanzanian holder on written terms, with eligibility confirmed with the Commission before any money moves.

Gold dealing: the margin lives in the gaps

Gold is a high-price, thin-margin trade, and Handbook 48 is built on that fact. Its case is a licensed dealer in Geita that buys fine gold at TZS 360,000 a gram, which is 90% of an assumed world price of TZS 400,000 a gram. That assumed price is fixed for all twelve months of the model. After royalty and assay shrink, the dealer keeps TZS 22,848 a gram when it sells through the Bank of Tanzania programme, and TZS 4,902 a gram on export.

The overheads set the threshold. At the case's overheads of TZS 237,249,000 a year, the dealer covers its costs at 12,331 fine grams a year, or about 1,028 a month, against an illustrative plan of 31,000 fine grams in 2027. The first month runs a small shortfall, the full illustrative year returns TZS 359,173,800 before tax, and a half-volume year still earns TZS 74,912,400. The capital plan is TZS 689,923,076, mostly working capital tied up for six to fourteen days between purchase and payment.

Licensing is a sequence, not a form. A dealer licence costs TZS 550,000, made up of a TZS 50,000 application fee and a TZS 500,000 licence fee under the First Schedule of GN No. 2 of 2018, with renewal at TZS 450,000 under the same schedule. The book reads that schedule at 8 October 2026 and asks the reader to confirm the current schedule with the Mining Commission. The licensed place also needs a strongroom, a register, an AML policy and a tax number, and the case holds no gold before the licence arrives. A dealer in precious metals is a reporting person under the anti-money laundering rules. It must identify its customers and owners of 20% or more, keep records for ten years and report suspicion to the Financial Intelligence Unit within 24 hours.

Buying is confined to the market. The case buys only at its counter in the Geita mineral market, and only from sellers that section 18(1) allows: primary licence holders, their agents with written authority, and licensed brokers with a licence in date. Every seller has a file with identity, licence, expiry, market receipt, source and a bank account in the seller's own name. On lots bought from miners, a 2% withholding applies. The book treats that withholding as the seller's tax, remitted to the Tanzania Revenue Authority, and warns that it must never become the dealer's cost. One lot bought outside the market can cost up to 3.25 years of the case's profit after tax, and a conviction bars the licence.

Royalty and fees are paid before the gold is sold. On the Mainland the Act sets royalty at 6% of gross value under section 87(1)(c) and a clearance fee of 1% under section 90A. GN No. 84 of 2026 requires both to be paid before sale on Form MTF 23. Through the Bank of Tanzania programme, the book reads royalty at 4% and the fee at 0%, on terms that apply from 1 October 2024. The book also records that a law firm guide lists 4% as the general royalty rate. It follows the Act and the Bank's notice, and it flags the difference for confirmation with the Mining Commission, because the answer changes the margin.

Export requires the 20% set-aside. Under GN No. 539C of 2024, at least 20% of a dealer's gold must be sold for domestic use, and no export permit is granted without it. In the case, 600 fine grams sold through the Bank nets TZS 229,708,800, against TZS 218,891,200 on export, because the Bank's royalty is lower and the Bank pays for refining. Holding gold for a better price adds price and currency risk. The book shows that a fall of 5.08% in the price before sale erases a peak week's margin, and it treats the days a lot is held as a cost, not an opportunity.

Small-scale gold mining: prove each step before paying for the next

Handbook 49 follows the Chunya mine, a company owned only by three Tanzanian citizens, from a reef on family land in Mbeya Region to a working pit. Its method is that each step must be proved before the money for the next one is spent. The grade comes before the licence, the licence and the landowners before the shaft, the permits before the first blast, the plant before the payroll, the cash before the rains and the records before the sale.

The grade is proved first. The partners rejected a grab sample of 18.0 g/t shown to them by a seller, and ran their own programme of six trenches and 36 channel samples, with TZS 1,980,000 of assays from an independent laboratory. After a top cut at 8.0 g/t and 10% dilution, the results gave a planning grade of 2.95 g/t, against a break-even of 1.55 g/t after the loan. The book's point is that a grab sample anchors every later number unless it is set aside, and that the top cut and the dilution must be written down before the results arrive. The planning grade belongs to one illustrative reef and is not typical of Tanzania.

The licence and the land come next. The company checked the cadastre, applied for 10 hectares with a local content plan and an environmental protection plan, and was granted a seven-year primary mining licence on 15 June 2026. Renewal must be applied for not later than three months before expiry. Entry onto occupied land requires consultation with the village and the lawful occupier's written consent, under section 95(1) of the Mining Act as enacted, and the case signed an agreement covering crops, one house and a village contribution of TZS 25,400,000 before entering occupied land. The application fee is not printed and is to be confirmed with the Commission.

Permits come before the first blast. Five lines stood between the licence and the first blast: the environmental certificate from the National Environment Management Council, a water use permit, a licensed blasting contractor, workplace registration with OSHA, and the business and tax file. The first calendar allowed 21 line-weeks for these, and the lines took 35. The book shows that a month of build halted for a missing permit would have cost TZS 37,897,080 with no gold produced. Blasting is done by a licensed contractor, and the book does not go further on it.

The plant and the safety system come next. The plant hoists 240 tonnes of ore a month from a 45 m shaft, crushes and mills it, concentrates the gold on a sluice and a shaking table, and smelts it with borax. No mercury is used on site. The plant and pit cost TZS 233,700,000, and with TZS 55,780,000 spent before the pit the set-up reaches TZS 289,480,000. The book costs the mercury route for comparison: the same gold won with mercury would use 460.2 to 1,062 grams of mercury a month. The safety system lists eight hazards, each with a control and a record, signs a one-page checklist at each shift start and gives every worker the right to stop work. In the case, one wall failure that stopped 18 of 26 working days would cost TZS 67,475,231, more than four times the year's safety line.

Finance and sales close the chain. The first cash plan started with TZS 294,220,000 on 1 July 2026, ran out of cash in September and reached a shortfall of TZS 64,500,240 by the end of October. A one-month operating buffer set further equity at TZS 125,000,000, paid in during June, and the funded plan's lowest month held TZS 849,760 above the buffer. Gold is sold only to licensed dealers at the mineral market, and the sale document, the royalty receipt and the inspection fee receipt are brought back the same day. The book reads the royalty at 6% of gross value and the inspection fee at 1%. Tailings sold each quarter under a written contract to a licensed processing plant brought TZS 204,435,000 in the first year.

Tanzanite and coloured gemstones: licences, grading and export

Handbook 50 treats tanzanite as a stone that earns its value in steps the law licenses one at a time. The blue-violet zoisite has one commercial source, in the Merelani Hills, where the Mirerani market and the controlled area lie. The book follows an Arusha dealer and cutter, a private company owned and run by a married couple, that buys rough each month, cuts it in its own workshop and sells to jewellers at home and buyers abroad.

The licence stack depends on the role. The Arusha company holds a dealer licence for coloured gemstones, which allows buying at the market, holding and exporting, and a small lapidary licence for cutting, polishing and selling within Tanzania on three machines. Both expire on 30 June. A broker licence allows buying at a buying station and selling to a licensed dealer, but not export. The large lapidary licence runs for five years and allows ten machines for Tanzanian holders and thirty for foreign holders. The dealer and broker fees were not read at the issuer and are shown as to be confirmed.

The Mirerani Controlled Area adds an entry rule. Under GN No. 135 of 2019, entry is by identification card or visitor's pass, between 7:00am and 7:00pm, through the main gate only, with vehicles by permission and a search on entry. Entry through any other gate carries a fine of not less than TZS 3,000,000 or up to six months. The case also needs cash before its first export receipt. It requires TZS 121,900,000 against TZS 150,000,000 available, and a two-month delay leaves a shortfall of TZS 35,500,000.

Rough is bought only at the market. A dealer buys rough only at a Mineral and Gem House or the Mirerani market, from a seller whose licence it records, against a receipt. In the case, a parcel asked at TZS 16,000,000 as all top was sorted by the buyer into 22 carats top, 12 good, 4 commercial and 2 reject. At the house's illustrative prices that was worth TZS 10,360,000, and the dealer offered TZS 9,842,000 and bought at that figure two days later. At the ask the parcel would have left TZS 2,748,000, against TZS 8,906,000 at the offer. Parcels outside the market are declined, and every lot enters the purchase register the same day.

Grading and disclosure follow. Colour leads the value of tanzanite, and the stone shows different colours in different directions, so the dealer grades colour first, face up, on a written card. The house bands of top, good and commercial are Zatra's teaching devices, not a regulator's grading system, and they never appear on an invoice as a grade. Heating is disclosed on every card and invoice, with no claim about detection. Reports from an accredited laboratory go to every top stone of 1.0 carat or more, which the case prices at TZS 150,000 a report for about 60 stones a year.

Cutting in Tanzania keeps more of the value, but only if the yield holds. Cutting in Arusha adds TZS 179,225,568 to the TZS 302,400,000 of rough. After TZS 92,660,752 of extra costs, the gain over selling rough is TZS 86,564,816. Three points lower yield cuts the profit to TZS 30,232,966, and six points lower makes a loss of TZS 24,658,884.

A lawful export of cut stones needs a dealer licence current on the day of shipment, a valuation, payment in lieu of royalty at 1% and the 1% clearance fee as read, an export permit, an invoice that states the true value and discloses treatment, laboratory reports and a packing list. On the case's TZS 368,037,600 of exports, the State's share is TZS 7,360,752. The same stones exported as rough would have carried TZS 17,040,240. The valuation procedure and the rough export limits were not confirmed at the Commission and are to be checked there.

The annual plan is priced with its loss case. In the licence year from 1 July 2026 to 30 June 2027, the case buys 2,760 carats of rough for TZS 302,400,000, sells TZS 526,985,568 of cut stones and makes TZS 85,124,816 before tax. A 20% fall in prices ends the year in a shortfall of TZS 18,800,148, and the breakeven fall is 16.38%. The book prices each band from the rough cost of a cut carat, at about 1.8 times the rough inside it, and limits memos to known buyers, never above TZS 20,000,000 each. These are illustrative prices, not market quotations.

Mineral tourism: a visit that is lawful, safe and paid for by the community

Handbook 51 starts from a plain observation. Tanzania received 2,294,495 international visitors in 2025, and very few of them stood at a mine, a carbonate volcano or a goldfield. The book follows an Arusha tour company that adds a mineral line for 2027. The plan has 74 departures carrying 370 guests, gross trip sales of TZS 437,200,000 and a line profit of TZS 51,700,000. The start-up outlay of TZS 46,000,000 is repaid in 10.68 months on the plan, and if only half the departures run, the line makes a loss of TZS 8,722,500.

The product set is three mineral trips and a counter. The tanzanite day at Mirerani is priced at TZS 380,000 a guest. The three-day Geopark trip, covering Lake Natron, the Lengai viewpoint, Engaresero and Karatu, is priced at TZS 2,350,000. The four-day gold heritage trip to Mwanza and Geita is priced at TZS 3,700,000. A gem and craft counter sits at the office. These are the case's illustrative retail prices.

Permission comes first. Each product needs permission to trade, to guide and to enter, and the book counts 17 permissions in all: five for the tanzanite day and four each for the Geopark trip, the gold heritage trip and the counter. The company confirms its tourism and business licences, registers its four guides under the Tourism (Tour Guides) Regulations, 2015, GN No. 270, and checks each mineral right with the Mining Commission before it signs written terms with the named holder. A guide must be a citizen aged 21 or over, enlisted with a licensed company, with a valid first aid certificate. Guide identity cards run to 31 December, and fines for breach are TZS 50,000 to 500,000 as read.

Surface only is the first safety rule. No guest goes underground or climbs a volcano. At Mirerani, guests enter only through the main gate with a visitor's pass, between 7:00am and 7:00pm, in vehicles with prior permission, and only to sites whose holders have signed written terms. Guests are briefed, equipped and counted in and out, and they are told never to buy stones at a site. The safety register lists 24 risks, 7 of them rated high before controls and 1 after, with the Natron road after dark held by a rule the owner signs. The book uses one serious injury, costing TZS 13,000,000 uninsured against TZS 4,500,000 insured, to argue for cover before the first departure.

Community consent is written down. A mineral trip holds the community's consent only when the village, the miners or right holders and the landholders have each signed an agreement. That agreement fixes the fund per guest, the jobs, the conduct and photograph rules, the accounts, the grievance route and the review date. The company's agreements leave TZS 35,200,000 a year at its sites: TZS 15,000,000 of mine visit fees, TZS 10,000,000 of community funds, TZS 7,200,000 for the cutting workshop and TZS 3,000,000 for walking guides. Every payment is per guest, by transfer, against a quarterly statement read at the village meeting.

Pricing starts from the pocket price. The company prices from the amount it keeps, not from the rate on the brochure. At a 25% agent commission, 3% bank and card charges and a 1% currency spread, a Geopark sale at TZS 2,350,000 leaves TZS 1,692,000, or 72.0% of retail. Agents bring 187 of the 370 guests and earn TZS 69,670,000. Government park, conservation, museum and permit fees are passed to the guest at cost, as the issuer publishes them on the travel date, and they sit outside the price. The book reads VAT at 18%, with registration from TZS 200,000,000 of turnover in twelve months, and asks the reader to confirm the treatment of tourism services with TRA.

The counter and the camp are different businesses. The counter is the dealer's: the stones, the licence, the receipts and any export papers stay with the licensed dealer, and the tour company earns rent, a commission and its own craft margin. In the case that is TZS 18,000,000 of rent, TZS 6,660,000 of commission and TZS 19,710,000 of craft margin. A camp of eight tents costs TZS 540,000,000 and is a separate and much riskier decision. The book runs the camp at four occupancy levels and shows a loss of TZS 163,880,000 at 15% occupancy, so the reader should treat the camp as a later step, after the trips have run through a full season.

Figures and authorities to confirm

Several figures in the series move, and the reader should confirm each one with the authority named before pricing a business or a trip. The gold royalty is the most contested. Section 87(1)(c) of the Mining Act sets 6% of gross value, the Bank of Tanzania's programme reads 4%, and a law firm guide lists 4% as the general rate. The book follows the Act and the Bank's notice, and the Mining Commission should confirm the position before a dealer prices a sale.

Licence fees are the second group. The gold book prints the schedule in the First Schedule of GN No. 2 of 2018, namely a dealer licence of TZS 550,000, a dealer renewal of TZS 450,000 and a broker licence of TZS 115,000 in the first year, and it flags the current schedule for confirmation. The gemstone book prints no dealer or broker fee, because the schedule was not read at the issuer. The Mining Commission is the authority for both.

Bank terms and set-aside rules come third. The Bank of Tanzania terms for gold, including the 4% royalty, the 0% fee, zero-rated VAT and payment within 24 hours of the assay report, apply as read on the date stated, and the Bank's pre-emption right must be confirmed before each export. The 20% domestic set-aside under GN No. 539C of 2024 and the pay-before-sale rule under GN No. 84 of 2026 should be confirmed with the Commission and with the Tanzania Revenue Authority before any shipment.

Site permits are open items. The small-scale mining book flags the fees for the primary licence application, the environmental certificate, the water use permit, the blasting contractor and the workplace registration as open, and it names the bodies to confirm each one: the Commission, the National Environment Management Council, the basin water board, OSHA, the Tanzania Revenue Authority and the local council. The council business licence and any service levy are also open in the gold case, which holds TZS 2,550,000 for levies still unresolved.

The gemstone and tourism figures need the same care. The gemstone book states the Mirerani entry fine and the royalty on gemstones as read, but it states that the Mirerani rough export limits are not confirmed and that the dealer and broker fees are to be confirmed. The tourism book leaves the Tourism Development Levy, the tour operator licence class and the Mirerani entry fee schedule unread, and it confirms park and museum fees with each issuer on the travel date. For the tour company, class A business licences come from BRELA and class B from local government authorities, and the tourism licence comes from the Ministry responsible for tourism. The fees for each were not read and are to be confirmed with the issuer.

Several instruments carry outstanding amendments. The Mining Act consolidation dated from 30 November 2019 is marked by TanzLII as having outstanding amendments, and the Mirerani regulations, GN No. 135 of 2019, list amendments of 19 April 2019, 17 January 2020 and 2 July 2021 that the book has not applied. The reader should check the current consolidation and the Commission's position before relying on any section number in these books.

Mainland and Zanzibar are separate

Every book in the series applies Mainland Tanzania only, and none of them reads or applies Zanzibar law. The gold book states that Zanzibar has its own regime and is not covered. The tanzanite book states that Zanzibar regulates minerals under its own law. The mineral tourism book names the Zanzibar Commission for Tourism as the regulator on that island, acting under its own law, which the book did not read. No Zanzibar rule is printed in any of the four.

The separation matters in practice. A Mainland licence class, royalty rate, entry gate or guide registration has no force in Zanzibar, and a reader planning a business there should start from Zanzibar's own authorities rather than adapt a Mainland figure. The Mirerani Controlled Area rules, the sections of the Mining Act and the tour guide regulations quoted here are Mainland instruments. Applying them to a Zanzibar business would produce a wrong cost or an invalid licence, and the books give no basis for that transfer.

The case figures are Mainland examples too. The Geita, Chunya, Arusha and Mirerani businesses are all placed on the Mainland, and their prices, yields and costs are illustrative. A reader who carries one of those figures into a Zanzibar model is working from neither the Zanzibar law nor the Zanzibar market.

Choosing the book and the reading order

The four books are not a sequence, although they share a method. A reader who has not yet chosen a role should begin with Handbook 48 if the business is buying or selling gold, and with Handbook 49 if it is digging the gold. Gemstone dealing and cutting belong to Handbook 50, and visits to mines and volcanoes belong to Handbook 51. The gold pair covers both sides of the same chain: the mine's sale to a licensed dealer and the dealer's purchase from the mine. The gemstone and tourism pair shares Mirerani, because the tourism book's counter sits on the gem trade that Handbook 50 describes.

Within each book, the reader begins with a short check. The reading check in each book scores the reader on twenty statements and names the chapters and tools to start with. Each book then closes with twelve tools, each on one A4 page, with the case's figures in one column and a blank column for the reader's own figures. These are the Gold Dealer Kit, the Pit to Market Kit, the Gem Dealer's Kit and the Geotour Kit. Each book's kit can also be read online at the guides pages on zatra.co.

Cross references matter. Handbook 35, Scams in African Mineral Deals, covers fake sellers of licences and gold, and the small-scale mining and gemstone books point to it as the buyer's defence against fraud in mineral deals. The tourism book points to Handbook 50 for the tanzanite trade, to Handbook 57 for starting and licensing a tour company, and to Handbook 56 for hotels and lodges. A reader who plans a camp should read Handbook 56 before the camp chapter of Handbook 51.

Where the margin is tested

Each book places the margin at the centre of its argument, and each test sets a loss case beside the gain. The gold dealer keeps TZS 22,848 a gram through the Bank and TZS 4,902 on export, and a fall of 5.08% in the price before sale erases a peak week's margin. The small-scale mine shows a shortfall of TZS 846,780 a month when a lower grade and a lower price arrive together, and a month of build stopped for a missing permit costs TZS 37,897,080 with no gold produced. The tanzanite dealer breaks even at a 16.38% fall in prices. The tour line makes TZS 51,700,000 on the plan and loses TZS 8,722,500 if half its departures run.

The loss cases share three features. First, each is priced in the same currency and period as the gain, so the reader sees a net figure and not a best case. Second, each names the step that fails, whether that is a permit, a yield, a price or a departure. Third, each sets a rule that protects the margin, such as a walk-away grade, a buying ceiling, a permit that must precede the lot, a minimum group size or a cut-off date. Several of these rules are written as tools that the owner signs.

None of the books forecasts a price or recommends an investment. Each is an education in process, and each states that its numbers are an example. The reader should test the same tools against the reader's own business, rather than take the case's figures as a benchmark.

Property and Construction is the Zatra Handbooks shelf for a reader who wants to earn from land, buildings and the materials that go into them. It has four titles, each written for one decision: whether to develop a plot, whether to buy or build a rental block, whether to become a registered contractor, and whether to open a building materials yard. All four apply Mainland Tanzania law and practice. The worked cases are set in Dar es Salaam and Dodoma, are illustrative, and are built from stated inputs rather than from records of real businesses. Several figures refer to 2027 to 2029 and are labelled as projections. The series does not cover Zanzibar, which has its own land, planning and contractor rules. Every rate, fee and date should be confirmed with the issuer named beside it before anyone commits money.

What the series covers

The series has four handbooks, each a book of about 170 pages. Real Estate Development in Tanzania (Handbook 52) takes a family or a small developer from a titled plot to a registered unit title. Buy, Build and Rent Property in Tanzania (Handbook 53) tests whether a rental block should be bought or built, then runs its tenancy and tax. Start a Construction Company in Tanzania (Handbook 54) covers registration with the Contractors Registration Board and winning public work. Building Materials Business in Tanzania (Handbook 55) compares four ways to run a yard and sets out how its stock and credit should be managed.

The four books share a method. Each is built around one decision, and each runs the sums twice: once on the base case and once on a loss case. Figures are computed from inputs printed beside them. Cases are described by type, not by name, and each worked example says that it is an example. Each book also carries twelve tools, one page each, for use at a desk, such as a feasibility sheet and a monthly cash pack.

Because the cases are illustrative, the reader should not read the figures as benchmarks. The books give a return on a particular plot, block or yard, on particular assumptions. A reader should replace those assumptions with his or her own quotations, rents and payment records before relying on any result. The books state that the rules they rely on were read on 7 and 8 October 2026, and that rates and dates may have moved since then.

Test the plot before the drawings

Real Estate Development in Tanzania opens with a brochure. A contractor's sales sheet for a six-floor apartment block on a family plot in Mbezi Beach, Dar es Salaam, shows sales of TZS 10,560,000,000 and a profit of TZS 6,600,000,000, with the note "Land: already owned". The book's first chapter takes the brochure apart line by line. The sellable area is 80% of the gross floor area, which gives 2,640 square metres and 24 units, and the units sell for less per square metre than the brochure assumed. On the stated inputs the sales come to TZS 8,298,000,000.

The cost side is corrected in the same way. The car park, services and external works add to the building cost, and a set of soft costs that the brochure omitted is added: contingency, professional fees, permits and legal work, marketing and sales, and interest and facility fees. The family land is counted at its market value of TZS 675,000,000, not at the sum already paid for it, because putting the land into the project gives up the money it would fetch. With these corrections the base case shows a total cost of TZS 6,709,738,282 and a profit of TZS 1,588,261,718, which is 23.67% on cost, as stated at 10 October 2026.

The book then applies what it calls the outside view. Its source is a study of 258 transport projects, in which costs were underestimated in almost nine cases out of ten and actual costs averaged 27.6% above the estimate. The book uses that average as a stress test for the building works, not as a forecast for housing. With a 27.6% overrun and the contingency in place, TZS 836,000,000 is uncovered and the profit falls to TZS 84,929,463. The loss case adds a six-month delay, and that combination produces a loss of TZS 287,534,593. The loss case sits beside the base profit on every page.

The chapter also reverses the test. At a builder's hurdle of 20% on cost, the plot is worth TZS 880,261,718 inside the scheme, which is TZS 205,261,718 more than its market value. The project can absorb a fall in prices of 17.75% or a rise in building works of 29.13% before the profit reaches nil. The reader is advised to set those limits in writing before the design begins, and to sign a go decision only if the loss case is acceptable to every owner.

Clear the title and search again

The second chapter asks whether the land is fit to build on. A certificate kept at home shows what was registered at some point in the past; an official search shows what is registered today. In the case, the official search was lodged on 19 October 2026 and showed an old mortgage that had been repaid but never discharged. The discharge was registered, and a second search confirmed that the register was clear. The book's lesson is simple: discharge old charges and search again, because no bank or buyer will accept a title that still carries another party's mortgage.

Use and density are the next checks. A title described as residential does not settle how many floors may be built. The council's written advice on density and height does that, and the book asks for it in writing before design. The chapter also explains the development clock under the Land Act, which is said to run from the date of the grant. The book reads the clocks under sections 34 and 37 of the Act as closed in this case. This summary flags that reading for confirmation with the Ministry of Lands and an advocate before any reliance is placed on it.

Moving family land into a company is the last step in the chapter. The book recommends that the land go into the company at its market value, with TRA's certificate before registration, and that an advocate settle the transfer. Boundaries, access and services are surveyed before the site is hoarded, and any encroachment is settled in writing. The chapter ends with the question that opens the permit chapter: which approvals the scheme needs, from which office, in what order, and what happens if an office stays silent.

Permits follow a sequence

The third chapter treats permits as a chain rather than a pile. The land file comes first, followed by planning consent and the building permit from the council, then environmental screening by the National Environment Management Council (NEMC) and its certificate, and finally the approvals for fire, water and power. The sequence matters because each step depends on the one before it. A building permit cannot be issued before the land file and planning consent are in place, and a contractor cannot be engaged before the permit is granted.

In the worked example, the scheme design was lodged with the council on 4 December 2026. The council's sixty days ran while the drawings and tender were completed, and consent was projected for 28 January 2027. NEMC registration followed, and the certificate was projected for 26 March 2027, before the contract and the site start. The book then measures what a delay costs. Six months of delay takes TZS 247,674,728 from the base profit, and combined with the average overrun it produces the loss shown in the summary.

The reader should note that none of the fees for consent, NEMC screening, survey, unit plans or titles is printed in the book from an official schedule. Those items are listed in a verification table and carried in the case as one illustrative lump sum of TZS 95,000,000 for statutory and legal costs. Each fee must be confirmed with the issuing office before the budget is set. Whether VAT applies to the units is also to be confirmed with TRA before pricing.

Build, fund and sell off-plan without betting the land

The fourth chapter buys the building in three steps: a design brief with an 80% efficiency target, a cost plan that totals TZS 4,750,000,000 for the building works, and a lump-sum contract let after a tender to four registered contractors. The contract carries an advance against a guarantee, retention of 5% and certificates signed by an independent quantity surveyor. The book's point is that the build is priced three times, in the brief, the plan and the contract, and each pricing fixes a share of the risk that the owners carry.

The fifth chapter funds the project. The family's cash of TZS 900,000,000 goes in first and is used up by January 2028. The loan that follows is sized for timing only, since the sales cover the cost. In the base case the loan peaks at TZS 685,776,654 in October 2028, costs TZS 31,440,515 in interest and a facility fee of TZS 6,857,767, and is open at only seven month-ends out of thirty. If no unit sells before completion, the peak rises and the profit falls to TZS 752,369,798. The book also shows that in the loss case the family recovers only 0.82 times its money.

The sixth chapter sells off-plan. The projected launch is 1 March 2027. Four units are sold before the build starts and twelve before the foundations are certified. Each buyer first receives the documents that section 29(1) of the Unit Titles Act 2008 requires, and may rescind within ten days and be refunded within ten days. The payment plan takes 20% at signing, 10% at each of four certified stages and 40% at handover. The book notes that the Act, as read, does not itself require an escrow, so the escrow that holds buyers' money is written into the sale agreement and into the lender's terms, and money is released only against a stage certificate. If prices include 18% VAT, profit falls to TZS 226,039,272.

Register the units and hand over

The seventh chapter treats the developer's work as ending a year after the keys are handed over. Units exist in law only when a unit plan meeting section 14 of the Unit Titles Act, with the surveyor's and local authority certificates required by section 15(1), is registered. Each unit then receives a title showing its unit factor, which in the case is 95 divided by 2,640 for a two-bedroom unit and 125 divided by 2,640 for a three-bedroom unit. TRA's certificate is needed before each transfer.

Registration creates an association of unit owners. It needs a committee, a managing agent, its founding documents within six months and a budget, which the case sets at TZS 104,400,000 split by unit factor. Half the retention, TZS 118,750,000 in the case, waits on a closed defects log. The retention therefore reaches the developer after handover, and the cash plan should carry that gap.

Buy or build: the yield decides

Buy, Build and Rent Property in Tanzania begins with a choice. A Dodoma couple had two routes to a rental income. Route A bought three finished flats near the centre of the city for TZS 432,600,000. Route B bought a plot seven kilometres out and built a block on it for a budget of TZS 430,280,000 in total. The book compares the two on net yield. Route A nets 4.62% and starts rent at once. Route B nets 7.61% on budget, and 7.68% on actual cost after fourteen months of building. A twelve-month bank deposit was set at 9.36% before tax in the August 2026 rate, and 8.42% after a 10% withholding.

The couple chose to build, for an extra TZS 12,771,866 a year, the engineer's skills and two salaries that could carry the loss case. The book is explicit that another family could rightly choose to buy. Its method is to set the two routes side by side on paper, with the same rent, the same vacancy and the same costs, before any plot is walked.

The second chapter reads the location before the plot. Between March and June 2024 the couple walked three plots in Dodoma with the same block and budget on each. Plot A, seven kilometres out and near tarmac, cost TZS 430,280,000 in total and yielded 11.71% gross. Plot B, fourteen kilometres out on an earth road, was TZS 28,840,000 cheaper but yielded 7.53% with 30% of nearby units empty. Plot C, three kilometres out, yielded 12.99% but cost TZS 74,160,000 more than the couple could fund. The book scores each plot on a sheet weighted before the scoring begins, and recommends that the title be searched and beacon-checked before sale. Plot A passed both and was bought on 14 June 2024.

Cost the build and keep to budget

The third chapter costs the build. The block's quotation was TZS 324,000,000, or TZS 540,000 a square metre. Fees of TZS 19,440,000, a permits allowance of TZS 5,000,000 and a contingency of TZS 32,400,000 held by the owner brought the build budget to TZS 380,840,000. Construction ran 8.8% over, or TZS 28,512,000, on the frame, roof and external works. The contingency absorbed the overrun and left TZS 3,888,000. Two months of delay cost a further TZS 10,800,000 outside the budget.

The book uses a 28% overrun as a test, not a forecast. At that level the block's yearly cash would turn into a shortfall of TZS 8,146,114. The lesson is that the contingency is the first line of defence, and that the owner should record every variation in writing and decide it before the work is done.

Lease, screen and collect

The fourth chapter sets the return. The block's potential rent is TZS 50,400,000 a year. It collected TZS 46,200,000 after seven empty unit-months and one unpaid month, or 91.67% of potential. Operating costs of TZS 13,448,134 left TZS 32,751,866, a net yield of 7.68% on TZS 426,392,000. After tax at 10% of the rent collected and loan payments of TZS 21,451,032, the cash left was TZS 6,680,834, or 1.99% on the couple's own funds. The block breaks even at 69.24% occupancy before tax and 76.94% after it. In the loss case it costs the couple TZS 19,345,090 a year.

The fifth chapter replaces the handshake with a lease. Eleven days before the first tenants moved in, the couple signed a four-page lease in English and Swahili built from twelve clauses: parties, premises, term, rent, advance and deposit, utilities, repairs, use, subletting, rent review, notice and inventory. Flats are let six months ahead with a one-month deposit. Shops are let twelve months ahead, with 10% withheld for TRA. Every lease is stamped at 1% of the annual rent within 30 days. The book notes that rent control was repealed in 2005, so the terms are largely those the parties write.

The sixth chapter screens tenants on evidence that bears on paying and keeping the lease. Flat 4 was re-let in October 2026 after 31 enquiries led to 12 viewings, 6 applications and 3 full checks at about TZS 60,000 each. The chosen applicant showed proof of income, a rent at 27.8% of that income, and a previous landlord who confirmed three years of on-time payment. Late rent meets a dated ladder with contact on day 1, day 5 and day 15, then advice and, if the matter is contested, the tribunal. The book warns against forcing a tenant out, and applies the same ladder to every tenant.

Keep the block and the tax in order

The seventh chapter covers maintenance and tax. Six months in, the block showed its first wear and its first tax receipt. The couple keep a calendar and a reserve. Repairs of TZS 2,520,000 a year are set aside before the rains, and a reserve of TZS 3,525,120 a year builds to TZS 35,251,200 against TZS 24,000,000 of roof and repainting work. Upkeep takes nearly half of the running cost.

The tax set-aside is counted at 10% of rent collected, which in the first year was TZS 4,620,000. The shops' share of TZS 1,320,000 was withheld by the tenants, and the flats' share of TZS 3,300,000 was set aside until TRA confirms who pays the tax and whether the payment is final. The book's withholding rates are stated as 10% for a resident landlord and 20% for a non-resident, and the TRA page on withholding gives 10% for commercial rental. The reader should confirm the rate that applies to his or her own position with TRA before setting aside the sum.

Enter construction: company, CRB class and safety

Start a Construction Company in Tanzania follows a contractor who has run other people's sites for years and now wants a company of its own. The case begins with the choice of class. A Dodoma company with TZS 150,000,000 of capital applies to the Contractors Registration Board (CRB) as a Building contractor at Class VI. The book's reading of the By-Laws gives a limit of TZS 100,000,000 a contract for that class. The company holds all five contracts it wins in its first year, worth TZS 331,614,296, and a Class VII registration would have held only TZS 88,216,128 of that work.

The second chapter sets up the company. The founders use a private company limited by shares. The book gives BRELA's fees for the name reservation at TZS 50,000 and for incorporation at TZS 450,000, as read on 8 October 2026, and each must be confirmed with BRELA. The company takes its own TIN, confirms its VAT and withholding position in writing with TRA, and registers with the Workers Compensation Fund within 30 days of its first employee. Head office costs of TZS 2,950,000 a month and fixed costs of TZS 52,230,000 set the break-even. A founders' agreement settles roles, reserved matters, drawings, deadlock and exit.

The third chapter covers CRB registration. The class limit, the liquidity test of TZS 2,000,000 and the fixed asset test of TZS 10,000,000 are taken from the CRB By-Laws. The book's case clears both, with cash of TZS 71,650,000 and equipment of TZS 58,500,000. Its registration file scores 68 out of 100 against a pass mark of 60, with experience at 5 of 10, which is the minimum that each criterion must reach. The certificate follows payment of fees within 60 days of approval.

The fourth chapter treats site safety and insurance as part of the price. On the case's first contract, the contractor budgets TZS 2,870,000 for safety gear, first aid, fire equipment, signboard, hoarding, and contractor's all-risks and third-party cover, or 2.89% of the tender. The book states a penalty of TZS 1,611,678 on that contract if the signboard, hoarding, gear and accident register were left out. The CRB penalty figures must be confirmed with CRB, as the book itself requires.

Win public work and price the bill of quantities

The fifth chapter deals with NeST, the electronic tendering system through which public work is advertised. The book cites section 73 of the procurement law for NeST, and the reference should be confirmed with the Public Procurement Regulatory Authority. NeST lets a new contractor see and bid for public work in its class and type, and it brings competitors from other regions to every tender. The case wins by bidding on fewer tenders. A twice-weekly watch, a written go or no-go score, a second reader for the responsiveness checklist before every upload, and a diary entry for the fifth working day after each notice of intention to award keep the bids disciplined. In 2027 the case submitted 18 bids at TZS 285,000 each, won 3 and lost 3 on documents. It spent TZS 5,130,000 to win TZS 203,670,066 of public work.

The sixth chapter prices the bill of quantities. Each main rate is built from materials, waste, labour output and plant. Preliminaries are priced line by line, and head office at 12%, risk at 2% and profit at 8% are added as stated lines. Contract A came to TZS 99,167,883, which is TZS 832,117 inside the Class VI limit and TZS 1,632,117 below the next responsive bid. A rival bid of TZS 78,900,000 sat below the case's site cost, and the book treats such a bid as a warning. The walk-away price of TZS 91,822,114 is written down before the tender is closed.

Manage the site cash

The seventh chapter is about cash, and the sixth shows the price behind it. A contractor spends first and is paid last, so profit and cash part company on every contract. In 2027 the Dodoma contractor earned TZS 331,614,296 of work and a profit of TZS 9,100,004, yet reached an overdraft of TZS 37,307,189 in June, because public clients paid three months after the work instead of one. The site cash plan showed the overdraft before it arrived.

The remedies are practical. Certificates are claimed on the last day of the period with agreed measurements. Variations are carried out only on written instructions. A register of retention release dates is kept. The book shows that a late payment on a single certificate, IPC 3 paid 60 days late, cost TZS 780,519. Beside the base case, one 10% overrun and one more month of payment lag produce a loss of TZS 1,465,735. The reader should note that the book's public-client payment periods are case assumptions, and should be checked against the contract and the client's record.

Choose the model before the plot

Building Materials Business in Tanzania begins with the model. The owner in the case has twelve years' experience as a contractor's storekeeper and has about TZS 430 million of his own money and a possible loan to deploy. He compares four models: a hardware shop, a block yard, a quarry and a distributorship. The hardware shop returns 18.46% in the plan. The block yard returns 25.58%. The quarry returns 39.14%, but sits behind the hardest licence gate. The distributorship returns 11.67% on the largest capital and turns into a loss of TZS 2,000,000 in its loss case.

The case chooses a mixed yard on a main road, with cement, steel, roofing sheets, hardware, sand and its own concrete blocks. It returns 23.38% on a budget of TZS 576,061,214. In a slow year it still earns TZS 16,968,642 before tax, and with one default from a contractor the worst case is a loss of TZS 41,231,358. The book says plainly that another investor could rightly choose a different model, and it asks the reader to compare the four on capital, margin and return before choosing the site.

Licences, standards and the quarry line

The second chapter sets the licences and standards. The owner fills an eight-line register covering the company, the TIN, the NEMC category, the block test against the compulsory standard TZS 283:2025, the business licence for the plot, the VAT and EFD position, the employer registrations, and the mineral right of each sand supplier. The business licence is required for each place of business under section 3(2) of the Business Licensing Act, as read in the book. Each rule is stated with its Act and section as read on 8 October 2026.

The quarry line carries the hardest test. Sand and aggregate must come from a holder of a mineral right, and the royalty is 3% of gross value. The owner in the case turns down a cousin's riverbed, because it would need a mineral right, an environmental assessment and a royalty on every load. The book's position is that a cheaper supply that cannot show a royalty receipt is a liability for the yard. Each fee, royalty and standard is to be confirmed with its issuer: the authority under the Mining Act, NEMC and the Tanzania Bureau of Standards.

Stock, credit and the cash cycle

The third and fourth chapters cover the budget and the first margins. The owner rebuilt the budget from TZS 317,000,000 to TZS 576,061,214. Site and plant lines come to TZS 257,000,000, and opening stock of TZS 174,203,022, a credit fund of TZS 106,402,192 and a reserve of TZS 38,456,000 bring the total to TZS 576,061,214. He funds it with TZS 426,061,214 of his own money and a loan of TZS 150,000,000, which the case assumes at 17%, repayable at TZS 5,347,909 a month for 36 months. Supplier credit is counted only when it arrives.

The plan year's gross profit of TZS 404,078,351 comes mainly from blocks, which earn TZS 203,678,351, and from hardware, which earns TZS 93,600,000. Cement contributes TZS 24,000,000. The yard's counter view, that cement is the business, does not survive the first month's breakdown. A bag of cement earns TZS 1,000 when it is sold as a bag and TZS 14,144 when it is moulded into blocks. A good block costs TZS 928.52 and earns TZS 471.48. The owner prices each line on its own margin, charges delivery as a separate line and refuses general discounts.

The fifth chapter controls the stock. The yard opens with 36.26 days of cost of sales in stock, held from three days for sand to ninety days for hardware. Cement is used at 128 bags a day and is reordered at 640 bags in orders of 1,800, using a square-root rule rounded to whole lorry loads. Forty hardware lines sell 80% of hardware, while 640 lines hold half the hardware stock for 5% of its sales. The first full count, on 5 March 2027, found TZS 1,920,300 short, which at that rate would be TZS 23,043,600 a year, or 5.70% of gross profit. Pairs, gate notes, chalked cement and a variance rule close the gap.

The sixth chapter manages credit. The yard sells TZS 970,920,000 a year on account. If contractors drift to 75 days and 2.5% bad debts, credit ties up TZS 199,504,110. Under a policy of 40 days and 0.5% bad debts it ties up TZS 106,402,192, releasing TZS 93,101,918 and saving TZS 35,245,726 a year in bad debts and carrying cost. The book notes that a stricter policy could lose 5% of credit sales, and in that case the saving falls to TZS 26,153,963. At 31 March 2027, six accounts owed TZS 168,400,000 on the projection, and 60.87% of that sum was on invoices more than 60 days old. The book shows that one default of TZS 58,200,000 needs TZS 310,762,306 of extra sales to replace.

The seventh chapter runs the cash cycle. The yard sells most in August and, on the drift pattern, holds the least cash in August: TZS 16,747,996, against a reserve floor of TZS 38,456,000. The cash conversion cycle explains it. Stock days of 36.26, plus 33.75 credit days, less 9.17 supplier days, make 60.84 days on drift. Collecting in 40 days cuts the cycle to 45.09 and releases TZS 93,101,918. The book sets a reserve floor, a thirteen-week cash plan and a six-page monthly pack to keep cash ahead of growth.

Figures to confirm before acting

Several rules in the series rest on instruments that the books read at a point in time, and some of those readings carry notes that the reader must check. The CRB class limits and the other tests in Handbook 54 are taken from the 2002 consolidation of the CRB By-Laws. The book states that the amendments of 2005, 2017 and 2022 were not applied in the text read. Any class limit, test or penalty must be confirmed with CRB before a bid is priced, and the reader should ask CRB for the current consolidated text.

In Handbook 52, the Land Act sections on the clocks, and the Urban Planning Act, are read from consolidations to 2019 and 2023. The TanzLII notes on the Unit Titles Act record an amendment of 26 March 2010 that was not applied in the text read. The TRA stamp duty page lists the lease rate at 1% of the annual reserved rent, but the rate for a transfer of land or property is not listed, and the handbook says so. The VAT registration threshold of TZS 200 million in twelve months is taken from the TRA VAT page, as read on 8 October 2026. Each of these points should be confirmed with the issuer before any transaction is signed.

In Handbook 53, the rental withholding rate and the treatment of rent for income tax are the points most likely to change. The handbook gives 10% for a resident landlord and 20% for a non-resident, and the TRA withholding page gives 10% for rental income for commercial purposes. The reader should confirm the rate and the filing dates with TRA, and should not treat the monthly set-aside as the final tax until TRA has confirmed it. The deposit rate of 8.42% after tax is taken from the Bank of Tanzania's monthly review for August 2026 and is an illustration, not a market quotation.

In Handbooks 54 and 55, the registration fees, the BRELA schedule, the licence fees and the royalty rate should be confirmed with the issuer named in each case. The book's illustrative figures for the contractor and the yard are built from stated inputs, and the reader should replace them with his or her own quotations and records.

Reading order and where to go next

A reader who is considering a plot should begin with Handbook 52, and should run the one-page feasibility and the loss case before any drawings are commissioned. A reader who already owns a block or is weighing a purchase should go to Handbook 53 and compare the two routes on net yield. A reader who wants to become a contractor should start with Handbook 54, and should first confirm the CRB class with CRB before a company is formed. A reader who is planning a yard or a shop for building materials should begin with Handbook 55 and compare the four models before choosing a site.

Across all four books, the reader should keep three habits. First, put every figure next to its issuer and date. Second, write the loss case beside the base case, and decide in advance what outcome would stop the project. Third, confirm each rule with the issuer named in the handbook before acting. These books explain the sums and the sequence; they do not replace the advice of a qualified adviser, an advocate for any dispute, or the authority that issues the rule.

Hospitality and Transport is a set of four Mainland books in the Zatra Handbooks. Each takes a reader through one licensed, asset-heavy business: a lodge, a tour company, a truck business and a customs clearing and forwarding agency. Each book is Book 1 to 4 of the series and stands alone, but the four share one method. Every book works through seven chapters, each a step with a task for the same evening, and closes with a kit of twelve tools to complete at a desk. Every book also runs one illustrative case through all seven chapters, and every gain it prints sits beside a loss case. The cases are composite, described by type rather than by name, and their figures are worked examples built from stated inputs. The books state their positions in October 2026. This summary explains what the series settles, which licence or permit each business must clear first, how the Mainland and Zanzibar are kept apart, and where the figures are most likely to move. It is general business information, not legal, tax or investment advice, and every fee and rule should be confirmed with the issuer named beside it before anyone acts.

What the series covers

Hospitality and Transport holds four books. Start a Hotel or Lodge in Tanzania (Handbook 56) takes a lodge from a plot of land to a loan that it can pay. Start a Tour Company in Tanzania (Handbook 57) takes a guide's trip menu to a licensed company in Arusha. Truck Transport Business in Tanzania (Handbook 58) chooses a lane and a fleet of one to ten trucks. Clearing and Forwarding in Tanzania (Handbook 59) licenses a customs agency and clears port cargo. Each book has the same structure: a decision in each chapter, a closing task for the same evening, and a kit of twelve tools that can be copied into a notebook and used inside the reader's own business with the source named.

The cases are composite. People and businesses are described by type, not by name, and each money figure is an illustration in Tanzanian shillings, built from inputs the book states. The figures are useful for seeing how the parts of a business interact, such as how a fixed cost changes the break-even of a truck or how a deposit changes a tour company's cash in its first season. They are not benchmarks, and the books do not recommend any lender, platform, dealer, contractor or supplier.

The reading order follows the decision. A reader with land near a park road and a plan for rooms starts with Book 1. A driver-guide who wants to sell safaris and climbs starts with Book 2, because the trip menu decides ownership before any vehicle is bought. A trader or former driver who is choosing between lanes, or deciding whether to buy a fifth truck, starts with Book 3. A reader who wants to open an agency, or to buy into one, starts with Book 4. The hotel book also sets out reading paths by stage, so that a reader who already holds a dated market count or a registered title can move straight to the relevant chapter.

Test the licence before the spend

The four books share one rule: no money is committed to a business until the licence or certificate that governs it has been read at the issuer. The rule takes a different form in each business, and in each the first working document is a file that lists every paper, its issuer and its date.

For the lodge the order is clean title in the company's name, planning consent from the planning authority under the Urban Planning Act, and the EIA certificate, all before construction starts. After construction, registration of the facility under the Tourism Act, 2008 (sections 8 to 11), the TTLB licence under section 21, a fire safety certificate under GN No. 106 of 2008 and environmental compliance with NEMC form the opening file. The fire certificate lasts one year, and the book asks that its renewal be diarised two months early.

For the tour company, the Board licence under section 21(1) of the Tourism Act, 2008 is required for every tourism activity and cannot be transferred under section 21(2). A breach is fined at not less than TZS 10,000,000 under section 46. The company takes no deposit until the licence is issued, and every quote carries the licence number. The ownership test comes first: the trip menu determines whether section 58(2) applies. In the case the Board's decision takes 63 of the 98 days between the founders' decision and the licence, and the book treats that timing as a planning assumption, not a promise by the Board.

For the truck business, carrying goods for hire needs a LATRA licence under Cap. 317, section 11. Registration, inspection, insurance and goods-in-transit cover must be in date before any paid load. The own-goods exemption does not cover other people's cargo. A haulier who carries for others therefore needs the licence even when the haulier owns the truck outright.

For the clearing agency, only a licensed agent may clear goods for others under sections 145 to 148 of the East African Community Customs Management Act, and the importer answers for the agent's acts. The TRA file is built in a fixed order: company and tax papers, an office, two clearance staff, TASAC, the application, the interview and the TZS 100,000,000 customs bond. The book's advice is to refuse a borrowed licence, a clearance before licensing and any promise to speed up the licence, and to finish the file in order.

The lodge: count the rooms before the site

The Karatu case is a 24-room lodge on four acres of former coffee land, between Lake Manyara and the Ngorongoro Conservation Area. Its guests are safari travellers on one or two nights between parks, and most are sent by tour operators. The project costs TZS 4,726,906,975 at constant 2026 prices, excluding VAT, which is TZS 196,954,457 a room. Construction is 63.94% of the cost, and the rest covers land, studies, fees, equipment, pre-opening costs, a 10% contingency and a three-month reserve.

The market count is the first test. Within 30 km of the plot, 38 properties with 1,150 rooms sell 218,270 room nights a year at 52% occupancy. The lodge's planned 4,968 room nights equal 2.28% of that total. Nine of 22 tour operators promise 3,600 room nights, which is 72.46% of the stabilised year, but the book notes that operator nights at contract rates do not alone pay the loan. Twelve rooms break even with debt only at 56.54% occupancy, and 36 rooms need guests the count has not found. The case therefore stays at 24 rooms and tests a smaller and a larger count against the same break-even.

The budget is tested for overrun rather than trusted. The book takes an average overrun of about 28% from research on large transport works and tests the budget at 10%, 20% and 28%. The contingency of TZS 302,260,000 meets the first test. At 28% the unfunded amount is TZS 544,068,000. Each month of delay costs TZS 83,260,325, so three months cost TZS 249,780,975. The loan is 45% of the cost, TZS 2,127,000,000, at a 16% rate that the book labels a case assumption rather than a market quote. Shareholders commit TZS 2,798,426,975, including TZS 198,520,000 of interest during construction.

Operating break-even is 22.89% occupancy. With debt service of TZS 472,928,736 the line is 39.32%, or 3,445 room nights a year. The stabilised year leaves TZS 500,428,839 in cash after the loan. The loss case, with 15 points less occupancy each month and agent rates 10% lower, leaves TZS 12,556,545 in cash and an accounting loss of TZS 27,580,819. A shock year like 2020 ends with a shortfall of TZS 666,442,080. The reserve of TZS 164,700,975 and the first two years' cash are the book's answer to such a year, and the monthly owner's pack shows the break-even line every month.

Licensing and grading come after construction. The book reads the TTLB citizen fee outside protected areas at TZS 766,500 for an approved property and TZS 1,277,500 at three stars, under GN No. 352B of 2025 as read at the issuer on 8 October 2026, and adds a TZS 20,000,000 allowance for charges still to be confirmed. Grading waits for a full season: the application goes in after the first season, and the decision is due within 45 days. Staffing is 33 people, with gross pay of TZS 25,850,000 a month, on-costs of 14% and meals and uniforms of TZS 3,300,000, a total of TZS 32,769,000 a month. Eight weeks of training and a soft opening come before the first guest on 1 June 2028, and the general manager starts at least twelve weeks before opening.

The tour company: price from the costing sheet

The Arusha case is a former driver-guide and a citizen investor who hold TZS 600,000,000 of capital at 25% and 75% under a written shareholders' agreement. The first season plans 150 safari guests, 80 climbers and 40 Zanzibar extensions, the last bought from a licensed Zanzibar ground handler. Planned revenue is TZS 1,732,050,000, and the plan shows a loss of TZS 7,574,200. The loss case shows TZS 76,824,304. The book sets the number of guests needed to break even, 237 against the 230 planned trip guests, and shows that two more direct parties would close the plan's loss.

Price begins with cost. A six-day safari for two guests costs TZS 12,414,615, or TZS 6,207,308 each, and 81.03% of that passes to parks, entries and lodges. The book prices at cost divided by 0.74 rather than cost plus 26%, so the direct price is TZS 8,400,000 a guest and the net rate to overseas agents is TZS 7,140,000. The seven-day Machame climb costs TZS 10,541,000 for two climbers, with crew wages at 24.90% of the cost, set at or above the wage order and the park's crew rules, which the book asks to be confirmed with the issuer. A single climber with the same crew costs TZS 7,326,000, which at the direct climb price is a loss of TZS 426,000. The book prints that case rather than hide it.

Vehicles follow the trip plan. The first season needs about 450 vehicle-days, or 150 a vehicle. An owned vehicle costs TZS 37,944,000 a year before fuel and wear, against a hire rate of TZS 380,000 a day. Owning wins only above about 146 days a year, and in the loss case each vehicle drives 113 days, so owning costs TZS 25,692,000 more than hiring. Driver-guides are placed on written retainers, and each vehicle has a file with its registration in the company's name, insurance, inspection and the name on both sides of the papers.

Terms decide who carries the risk. Guests pay 30% at booking and the balance 60 days before arrival. A written cancellation scale runs from 30% to 100% of the price, so no charge has to be chased after the event. A fee-change clause protects TZS 19,980,000 of contribution against a 15% rise in park fees. Overseas agents sign one agreement with one net rate, payment before arrival and no exclusivity. In the plan, nine of fourteen agents send 160 guests, and the company gives up TZS 190,350,000 of margin for them. The direct channel turns 600 enquiries into 360 quotes and 35 parties, and the company does not buy or fake reviews.

Park fees are planning allowances, because the season's TANAPA or NCAA tariff governs. The Board's licence fee is not printed. The book notes that fee schedules were reported in the press in 2018 and 2021 and does not treat any of them as the current fee. Each figure that the reader needs from the Board or from the parks must be confirmed on the day of the quote.

The truck business: choose the lane and count the border days

The Dar es Salaam case runs four trucks and billed TZS 1,075,200,000 in the year to 30 September 2026, keeping TZS 60,083,108 before tax, a margin of 5.59%. The book compares three lanes on one truck's year rather than on one trip's rate. The Lusaka lane earns a truck profit of TZS 42,463,280, Kigali TZS 20,556,480 and Mwanza TZS 17,440,900. Because the fixed costs are the same on every lane, the lane decides the margin. In the loss case every lane loses money, Mwanza the most at TZS 28,428,850. Brokering loads to match one truck would need 56 loads a year.

Papers come before the truck is bought. The case allows TZS 15,900,000 a truck a year for papers and cover, or TZS 63,600,000 for the fleet, and a stopped truck costs TZS 581,099 a day. The book sets out the LATRA licence, the registration, inspection, insurance and goods-in-transit cover, the driver's licence class, and the employer's record under section 79 with written particulars. The book notes that the 2012 ten-year age rule is no longer the test, and asks the reader to confirm the current rule with LATRA in writing before buying a truck.

Trucks are compared on cost per kilometre and cover. On the Lusaka lane a five-year-old combination costs TZS 2,886.95 a km and earns TZS 42,463,280 a year, ahead of a new one at TZS 3,003.49 a km and TZS 33,326,280, and a ten-year-old one at TZS 3,022.59 a km and TZS 29,926,320. On a 48-month loan at 16%, the new truck does not cover its payments, and in the loss case none of them does. A loaded round trip to Lusaka costs TZS 12,989,625 and earns TZS 16,000,000. At 60% backloads the average full cost is TZS 12,688,961, and a 12% margin needs a rate of about TZS 10,371,636 a load, against TZS 9,600,000 paid today.

Terms carry as much weight as the rate. Sixty days of credit keep TZS 176,745,205 with the cargo owners on the case's billing, and thirty days would free TZS 88,372,602, worth TZS 14,139,616 a year at 16%. A fuel advance cuts the cash out before payment. Detention after three free days at TZS 350,000 recovers part of every waiting day, and a liability limit of TZS 1,500,000 a tonne keeps a TZS 180,000,000 claim within the cover. Each cargo contract fixes payment days, detention and the liability limit in writing.

Border days are the largest single leak. At Tunduma and Nakonde the round trip of about 15 days carries 3.0 days of waiting, or 240 truck days a year across four trucks, worth TZS 139,463,760 at TZS 581,099 a day. Most holds began in an office: papers that did not match the cargo, declarations lodged on arrival, far-side agents not instructed and missing insurance documents. A border pack checked by its holder before departure, declarations lodged ahead and a written held-truck routine cut the border days to 1.5 and the round trip to 13.0 days. That gives 23 trips a truck instead of 20 and adds TZS 62,283,840 of contribution to the fleet.

Controls on fuel, tyres and drivers come next. The fleet runs at 0.45 litres a km loaded against a standard of 0.42, and tyres last 80,000 km against 100,000. Together the leaks come to TZS 43,762,880 a year, of which one truck running at 47.6 litres per 100 km accounts for TZS 10,361,344 of the fuel gap. The book holds that closing these leaks needs no debt and should come before any expansion. It also records driver fatigue as both a cost and a danger, and asks that tracking data be read monthly with the driver.

The clearing agency: client money and the five-day clock

The Dar es Salaam case plans an agency with two licensed declarants and seven staff in its first trading year, handling 474 jobs in 2027. The agency keeps one line of each importer's bill. For a 20-foot container the importer pays TZS 18,850,000, of which the customs value is TZS 35,000,000 and the agency's fee is TZS 550,000, or 2.92%. The rest of the bill is client money, recorded as such. Start-up capital is TZS 159,270,000, of which TZS 125,100,000 is float, reserve and bond cover. Break-even is 320 jobs a year, or 27 a month, and the plan earns an operating profit before tax of TZS 67,033,258.

Licensing sets the pace. As read at TRA on 8 October 2026, the application fee is USD 50 and the licence fee USD 400, with a bond security of TZS 100,000,000. The book instructs the reader to confirm each fee before applying. The importer answers for the agent's acts, so the agency's file must be right before its first declaration. Cargo reserved to TASAC is screened out of the agency's ordinary work and handled only under TASAC's own steps.

The port clock starts at discharge. The six steps take 11 days when documents are chased after arrival, and 5 days when they are checked at loading and the importer's deposit is in hand. No declaration is lodged without that deposit, so payment takes hours rather than days. Invoice, packing list and bill of lading are checked against each other at loading, because a mismatch found at examination costs days. A 40-foot container four days late costs TZS 416,000 in charges, which is 59.43% of its fee. The routine cut the agency's own share of line charges from TZS 9,100,000 a year to TZS 1,560,000.

Money is the main risk. Under a loose policy of advancing 35% for nine days, the leaks reach TZS 35,915,493: TZS 8,243,093 of float cost, TZS 18,572,400 of bad debt and TZS 9,100,000 of container charges absorbed. Under the tight policy, with deposits before lodging, approved accounts on written limits and a weekly days review, the leaks fall to TZS 2,606,742. The plan's operating profit before tax falls to TZS 33,724,507 under the loose policy, and the book sets that figure beside the loss case. Client money sits in its own account and ledger, reconciled weekly, with a receipt for every payment.

Transit is a promise to TRA that goods will leave Tanzania, and the agency answers for that promise until the file is acquitted. In the case a container of goods worth TZS 60,000,000 carries TZS 27,000,000 of duty at an illustrative rate of 45%, against a fee of TZS 900,000. The TZS 100,000,000 bond covers 3.7 such containers at once. The agency controls the risk with a checked, subcontracted truck, a recorded seal, a twice-daily position report and a ten-day acquittal rule. An open transit also blocks the licence renewal.

Growth follows relationships. The five largest importers bring TZS 63,430,000 of the plan's contribution, or 30.90%, so the book asks for a quarterly review by importer, with problems raised the same day. A 1,000 m2 duty-paid warehouse costs TZS 16,200,000 a month, breaks even at 51.43% occupancy, earns TZS 89,100,000 a year at 75% and loses TZS 43,200,000 at 40%. A lease is signed only once 360 committed pallet places are in view.

The kit and the mistakes the books name

Each book ends its chapters with a kit. The hotel kit holds twelve tools, each on one A4 page, covering the market count, the permit path, the project budget and overrun test, the opening compliance file, the grading readiness check, the staffing plan and pay grid, the rate card and channel mix, the tour operator contract checklist and the monthly owner's pack. Each page carries the case's figures in one column and blank columns for the reader's own project. The tour kit is built the same way around the trip menu and ownership check, the licence file, the safari and climb costing sheets, the rate sheet and booking terms, and the season cash calendar. The clearing kit includes a lane and capital planner, a break-even job counter, a licence and bond file, an exclusive items screen, a pre-arrival document checklist, a container days clock, a job quote builder and a price hold test. The truck book follows the same structure for the lane, the truck year, the trip and the border.

The kits are meant to be filled in, not read once. The tools ask for dated figures with their issuer, and the reader is expected to replace each case figure with a real one before relying on the result. Each book sets its loss case beside its plan, so a reader sees the effect of a poor year without having to build that comparison alone.

Each of the four books closes with a list of common errors. The hotel book lists sixteen hotel project mistakes and the tour book lists sixteen tour company mistakes; the truck book lists sixteen haulage mistakes. Several errors in the tour and truck books are typical of new operators. A tour company that adds a margin to cost, rather than dividing cost by the share it keeps, underprices its trips. A haulier that prices a load from the cost of fuel alone leaves out the fixed share of the truck and the overhead. A truck held at a border that is left to clear by itself costs days that the trip sheet did not allow for. A fleet bought to meet a cargo owner's offer before the existing trucks are under control adds cost to a business that already leaks. The clearing book names the parallel errors in its own trade: a fee set below cost or below the TASAC minimum, a declaration lodged before the importer's deposit is held, and importers' money mixed with the agency's own.

Mainland and Zanzibar are separate

All four books apply Mainland law and state Zanzibar separately, and none treats the two as a single regime. The lodge book says that licensing of accommodation in Zanzibar is by the Zanzibar Commission for Tourism, whose rules the book did not read at source, and that the Mainland TTLB fees, the NEMC regulations and the fire and rescue rules set out in the book do not apply there as written. It also states that the Zanzibar Revenue Board collects an infrastructure tax from hotel guests at USD 1 to USD 5 a person a day by star rating. That figure is given as the book reads it and is to be confirmed with the Zanzibar authority.

The tour book sells Zanzibar extensions bought from a licensed Zanzibar ground handler and does not carry the Mainland licence across to them. The truck book does not cover Zanzibar at all, and says that a haulier working there must confirm every road-transport document with the Zanzibar authorities. The clearing book says that cargo moving between the Mainland and Zanzibar follows the procedures of the Zanzibar Revenue Board and the Zanzibar ports. Those procedures were not read for the book, and the reader is asked to confirm them with the Zanzibar authorities.

For a reader whose plan is in Zanzibar, the Mainland figures in these books are a method, not a budget. The licence, the fees, the tax and the port procedures must be read at the Zanzibar issuer before any Mainland number is reused.

Where the figures move and what to confirm

The books rest on one practice. Each rule is named with its issuer and date, and where a fee or step could not be read at the issuer, the book prints no figure and says confirm with the issuer. A reader should keep that practice and not fill the gap with an estimate. The points most likely to change are set out here.

Licence and fee schedules move most often. The lodge's TTLB fee was read under GN No. 352B of 2025. The Board's licence fee for tour operators is not printed in the tour book. LATRA's regulations were revised from 1 July 2026, so the truck book's licence rules must be confirmed in writing. TRA's application fee, licence fee and bond were read on 8 October 2026. The clearing book states that each fee sits above the TASAC minimum as read, and asks the reader to confirm the order in force with TASAC.

Park and crew rules are set by the parks and the wage order. The TANAPA and NCAA tariffs govern park fees, and the crew rules for climbs must be confirmed with the issuer on the day of the quote. The lodge book applies VAT at 18% and treats VAT registration as due once turnover passes the threshold, which in its case happens in the first month of trading. The threshold and the registration step are to be confirmed with TRA.

Case assumptions are not forecasts. The 16% loan rate, the 18% online commission, the 28% overrun test and the 2020 shock year are assumptions and test cases chosen by the books. They are not market quotes or predictions. The results should be rerun with the reader's own inputs before any decision, and the loss case should be read beside every plan figure.

The books give general business information, not legal, tax, accounting or investment advice. Where a decision carries real money or legal risk, the reader consults a qualified adviser before acting, and an advocate for any dispute or demand. The books are a method for asking the right questions in the right order, and the answers to those questions belong to the issuers named in each chapter.

The Lending series covers how a company lends money to others in Tanzania, and what the Bank of Tanzania expects of it. It has two books, numbered 60 and 61 in the Zatra Handbooks. Handbook 60 explains the licensed microfinance route: the Tier 2 licence, the capital, the board and its written policies, one loan product, the cash test that sizes each loan, lawful collection and the reports that keep a licence in force. Handbook 61 applies the same licence to a lender that wants to lend by phone, and adds the platform clearance, the limits on phone data, a scoring ladder and a complaints desk. Both books are written for Mainland Tanzania. Zanzibar has its own regime, which neither book reads, so a lender planning there must confirm every rule with the issuer before acting. The worked cases are illustrative: a credit company in Morogoro and a digital lender in Mwanza, each built from stated inputs, with a loss case set beside every gain. The series recommends no lender, bank, app or provider. Read together, the books make one point. A lending business is first a licence question, then a pricing and cash question, then a collection question, and its three-year result depends on a poor season as much as on a good one. All figures are before tax and should be confirmed with the issuer named beside them before anyone relies on them. This summary is general information, not legal, tax or investment advice.

What the two books cover

Handbook 60, Start a Microfinance Company in Tanzania, is written for shareholders, accountants, bank officers and directors who plan a licensed company that lends but does not take deposits. Its case is a company in Morogoro with TZS 300,000,000 of paid-up capital, a licence assumed on 1 March 2027 and a first loan on 5 April 2027. The book runs in seven steps, from the shareholders' decision on the numbers to the reports that keep the licence, and each step ends with a working tool.

Handbook 61, Digital Lending in Tanzania, is written for Tier 2 companies adding a phone channel, fintech founders and investors who are asked to fund a loan app. Its case is a licensed lender in Mwanza that wants to lend to fish and grain traders by phone. It follows the same seven steps, with the platform clearance and the digital rules added at each one, and it opens with the Bank of Tanzania's list of 21 November 2024.

Both books use the same method. Each figure is computed from inputs printed beside it, each rule carries the issuer and the date on which it was read, and each illustrative case is labelled as such. Both end with a kit of twelve tools that a reader may copy into a notebook and use inside a business, notes by chapter, and an appendix of the rules as read on 8 October 2026. A Kiswahili edition of each book is listed in the catalogue under the same number.

The choice between the two books turns on the channel. A lender that will serve borrowers at a branch, through officers who visit them, starts with Handbook 60. A lender that intends to disburse and collect through a phone app starts with Handbook 61, but reads Handbook 60 first for the licence, the board and the capital, because Handbook 61 assumes that foundation. Each book points to the other at the places where the channel changes the rule.

The licence comes before the first loan

The licence is the first gate, and the law treats lending without it as a serious offence. Under section 16(1) of the Microfinance Act, 2018, no loan, promise of a loan or collection may be made before the Tier 2 licence is issued. An unlicensed business faces a fine of TZS 20,000,000 to TZS 100,000,000, imprisonment, or both, under section 16(2)(a) of the same Act. Handbook 60 notes that the Act's commencement date was not read at source and asks the reader to confirm it with the Bank.

The application pack has nineteen documents. It includes proof and source of the capital, the lending policy, bureau reports and tax clearances for every holder of 5% or more of the shares, every director and the chief executive, and a company name that carries a permitted word such as microfinance, finance, financial services, credit or microcredit. The fee under regulation 6(3) of GN 679 of 2019 is TZS 500,000 for an entity and TZS 300,000 for an individual, as read on 8 October 2026. A foreign owner must also add approval from its home regulator, training and succession plans under regulation 6(4).

Capital is tested in two places. The Bank requires TZS 20,000,000 at all times under regulation 46(1), and at least 5% of total assets held as liquid assets under regulation 47(1). The Morogoro company holds TZS 300,000,000, fifteen times the minimum. After TZS 29,500,000 of set-up costs and a TZS 10,000,000 floor, TZS 260,500,000 is left to lend. The book's blind spot is the belief that a licence is a formality. Its six sample applications show otherwise: a name without a permitted word, a loan shown as a partner's savings, a 6% shareholder left out of the bureau checks, a policy copied from another lender, a foreign shareholder with no approval from its home regulator, and a trader who kept lending while the application was pending. Each failure cost time at a pre-opening cost of TZS 3,000,000 to TZS 5,000,000 a month, and lending before the licence carried a fine of TZS 20,000,000 to TZS 100,000,000.

For a lender that plans to lend by phone, the licence is only the first of three gates. Handbook 61 requires the Tier 2 licence, a no-objection letter for the platform under paragraphs 4.1 to 4.3 of the Bank's Guidance Note on Digital Lenders of August 2024, and the Bank's clearance of the platform before it operates under paragraph 5.1(h). PDPC registration under the Personal Data Protection Act runs alongside. The book's plan of 140 days from decision to launch is built from five stretches. The longest, from the no-objection file to clearance, is 42 days that the book itself calls an assumption, because the Bank publishes no timeline for clearance. The lender's answer is to make its file complete on the first day, not to launch early.

The digital book's blind spot is that a licence is enough. It is not. The platform needs the Bank's no-objection letter and clearance, and every later change to price, product, screen or platform needs approval again. On 21 November 2024 the Bank had listed 69 platforms and apps barred from digital lending, six months after warning that lending without a licence is an offence under section 16 of the Act. That list is why the digital book starts where it does.

Governance and the written policy

The board of the Morogoro company has three directors, one more than the two Tanzanian directors that regulation 18 requires. Each of the four people who control the company holds a fit-and-proper file under regulation 10 and the Third Schedule: certificates, a bureau report and a tax clearance. The handbook separates the chair from the manager. A retired bank manager chairs the board, a hired manager runs the office, and delegated authority is written down so that neither can act beyond it.

The lending policy is the working rulebook. In the book it sets the size and term of loans, a single-borrower limit, a rule that no instalment may exceed half of verified surplus, a limit of one loan at a time, a cash floor, and an insider rule under which shareholders and directors borrow on the same terms as any other borrower. In one sample application a policy copied from another lender described twelve-month group loans, while the business plan described six-month individual loans, so the board rewrote the policy around its own product before filing.

Governance has a price, and the book shows it. Board fees of TZS 4,800,000 a year and an external audit of TZS 4,800,000 a year together come to TZS 9,600,000, which is 8.21% of the first-year running cost of TZS 117,000,000. The board pays for itself if its rules prevent about seven failed loans a year, since an average failed loan leaves TZS 1,384,891 unpaid after two instalments, and 9,600,000 divided by 1,384,891 is 6.93.

In the poor season the board cannot stop the losses. The loss case adds TZS 68,953,787 to year-two loan loss expense. The board reads arrears each month by month of disbursement and, in October 2028, cuts new lending by 30%. The company still ends the three years with equity of TZS 212,719,129, 10.64 times the minimum. The book's point is that governance is judged on profit and portfolio at risk, not on repayment alone.

Price from cost and show the full cost

Handbook 60 sells one product first: a business loan of about TZS 2,000,000 over six months at 4% a month on the reducing balance, with a 2% fee. A borrower receives TZS 1,960,000, pays six instalments of TZS 381,524, and the loan costs TZS 329,144 in all, interest and fee together. That is an effective annual rate of 72.02%. A flat 4% on the same loan would cost TZS 190,856 more, at an effective annual rate of 129.50%. The book insists that a lender quote the method and the effective annual rate, never the monthly rate alone.

The rate is built from cost, not copied from another lender. The book's parts are the cost of funds, running cost, expected loss, fixed cost and the margin. In the third year of the base case, running cost takes 24.01% of the average portfolio of TZS 558,927,374 and loss takes 6.11%, leaving a profit of 15.61%. In the loss case the same price leaves a shortfall of 7.33%. A higher rate is not a cure. At 5% a month with 7.5% of loans failing, 200 loans earn TZS 998,420 less than at 4%, because a higher price draws weaker borrowers. Term matters more than rate, and the six-month cap is treated as a risk rule.

The digital book reaches the same method by another route. Its lender prices an average loan of TZS 182,500 at 9% for 30 days. The 9% is built from 1.42% for funds, 1.32% running cost, 3.00% expected loss, 2.63% fixed cost at 5,000 loans a month, and a margin of 0.63%. The nominal annual rate is 109.50% and the effective annual rate 185.34%. The first design, which took a fee up front and allowed rollovers, showed an effective annual rate of 8,359.78% on a 14-day loan and was rejected by the board on 12 October 2026. Under the Bank's guidance all costs are shown before the application, in Kiswahili by default, with the items of regulation 39(2) written into the agreement.

Price cannot rescue careless lending, and the digital book says so plainly. A loan earns nothing once default reaches 7.05%, and at 9% default each average loan loses TZS 3,878 before fixed costs. On paper, a price of 16.19% for 30 days would keep the contribution at 9% default, an effective annual rate of 520.50%. The book rejects that route: it needs the Bank's approval, fails the lender's own test of fairness, and draws the borrowers least able to repay. The fix lies in who is lent to and how fast, which the next sections cover.

The second half of the method is the borrower's understanding. The Morogoro agreement carries every item of regulation 39(2), forbids any change of terms by one side, and allows early repayment without penalty. The book shows that one failed loan takes the earnings of 3.64 good ones, and that interest is front-loaded while principal is back-loaded, so the early months of a loan are the ones in which a borrower most needs a clear statement of what remains owing.

Lend to verified cash

The cash test decides how much a borrower may take. In the Morogoro book the instalment may not exceed 50% of verified monthly surplus, and the largest loan is that instalment multiplied by 5.242137, the annuity factor for six monthly instalments at 4%, rounded down. Surplus is verified from three months of statements, a visit, and the household's drawings written down. A tailor who asked for TZS 2,500,000 on a surplus she believed was TZS 1,000,000 was found to have TZS 600,000. The rule supports a loan of TZS 1,550,000 with an instalment of TZS 295,681, which is 49.28% of her verified surplus.

Two principles recur. Guarantors and collateral are protection, not the reason to lend, because the verified cash either supports the loan or does not. And the lender should reduce before it refuses: a smaller first loan keeps a good borrower and limits the loss on a weak one. The bureau report must be taken with the borrower's written consent under regulation 36 before each enquiry, and other lenders' instalments count as costs in the cash test.

The digital book applies the same discipline to phone loans, using a scorecard built from identity, a bureau report taken with written consent, the stated purpose and the lender's own record. The ladder starts small and climbs only on repayment: TZS 50,000, then TZS 100,000, TZS 200,000 and TZS 350,000. A first loan waits 24 hours and is never paid out at night, and a refusal carries reasons in Kiswahili within seven days. The loss case shows the cost of skipping the ladder. A push that skipped the first step added 660 unpaid loans at the same price and volume, and cost the year TZS 143,135,368 against the base case.

Data limits sit beside the scorecard. The first build of the app asked for eleven phone permissions, including contacts and messages. The cleared build asks for three: network access, notifications and the camera, the last used only to photograph an identity card, a point the book asks the reader to confirm with the issuer. The Bank's guidance bars access to contact lists, call logs, messages, photos, storage, social media, email and files. Financial transaction data is sensitive data that needs prior written consent. The lender registers with the Personal Data Protection Commission before any borrower record exists, assumed for 7 December 2026 in the model, and the maximum PDPC fine of TZS 100,000,000 equals 2.29 years of the base-case result.

Collect lawfully and count the loss

When instalments stop, the Morogoro company follows a written ladder with a date, a person and a record for each step. A missed instalment moves the whole balance of the loan into the past-due classes under regulation 44(2). Provisions follow regulation 45: current loans at 1%, especially mentioned loans (6 to 30 days late) at 5%, substandard loans (31 to 60 days) at 25%, doubtful loans (61 to 90 days) at 50%, and loss loans (over 90 days) at 100%. The ladder runs from a reminder and a workplace visit to a guarantor notice, a fourteen-day written notice under regulation 56, and a committee decision. Loans are written off after 120 days past due.

The poor season of 2028 shows why the ladder matters. In the loss case, 84 of the 563 loans made from April to December 2028 stop paying after two instalments. The September sheet shows a PAR 30 of 6.35%, against 1.72% in the base case, and a provision of TZS 18,202,188 against TZS 9,712,290. A loan is restructured only on a new cash test, through the committee, and is reported as restructured. The book warns against restructuring a late loan and calling it current, since a loan 70 days late given a new plan keeps the class of its first missed date until it is paid in full.

The phone-lending ladder is written in the same spirit, with consumer-protection limits added. A late loan is collected by reminders, a receipt for every payment, a call with a plan, fourteen days' written notice under regulation 56(1), and assignment only to a licensed collector after the borrower has been told. The agreement carries no late penalty and no recovery fee, so a collector may charge nothing. Threats, shaming, contact with third parties, unagreed charges and sending data abroad are all ruled out. In the loss case, the August loans written off come to TZS 73,912,500, three times the base case of TZS 24,637,500, and both are provided for in full at 31 December 2027.

The routine is monthly. Every loan is sorted by days past due on the last day of each month, the accountant applies the regulation 45 rates and books the change in the same month, and the board reads arrears by month of disbursement. The Morogoro book's warning is that PAR 30 must be read by batch, because a season shows first in the newer batches and only later in the totals. A written-off loan is still owed. The bureau report under regulation 35 still shows it as unpaid, and any later payment is recorded as a recovery and reported at once.

What the evidence says about phone credit

The digital book sets its commercial case beside the research, and the evidence is narrower than the claims in both directions. A study of a fully digital phone loan in Kenya by Bharadwaj, Jack and Suri (2019) found that about 34% of those eligible took one. The loans did not replace other credit, and borrowers were 6.3 percentage points less likely to forgo expenses after a shock. The authors caution that digital loans are not a cure for wider failures in credit markets.

Surveys summarised in the book by Kaffenberger, Totolo and Soursourian (2018) covered more than 3,000 Kenyans and 4,500 Tanzanians. About a quarter of phone owners in Tanzania had used digital credit, which is 1,132 of 4,574. About half of digital borrowers in each country had repaid late, 31% in Tanzania had defaulted, and 9% in Tanzania had cut food purchases to repay. The lender therefore reads its complaints register and its arrears count together each month, so that a rise in late payment is traced to its cause before it becomes hardship.

A study in Malawi by Brailovskaya, Dupas and Robinson (2024) found strong demand for a digital loan at 10% for 15 days. Most borrowers did not know the terms, often repaid late and paid substantial late fees. A short phone-based literacy lesson raised knowledge of the terms but did not improve timely repayment, and it raised the chance of ever defaulting. The book's design answer is a product with no late fee and one repayment date, so that a borrower who misreads the screen is not trapped by a charge the agreement does not state.

The practical conclusion is modest. A digital loan can soften a bad month for a trader, and the book's lender describes it in exactly those terms, making no wider promise in its advertising or its reports. A lender that wants the gain without the Tanzanian cost runs the steps of the book: a clear price, a small first loan, a waiting period, lawful collection and a desk that answers complaints.

Reporting, complaints and the three-year test

A Tier 2 licence does not expire, but it is kept by a record built every month. Bureau data is sent each month under regulation 35(2). Reports go to the Bank in its format under regulation 50, whose frequency the book leaves to confirm with the Bank. Audited accounts reach the Bank within four months of the year end under regulation 25, which for a 31 March year end means by 31 July. The company also keeps a complaints register, holds TZS 20,000,000 of capital and keeps 5% of total assets in liquid form. The Morogoro company's routine begins on the tenth of each month, the company's own bureau date.

Over three years the Morogoro base case earns TZS 124,281,269 before tax and leaves equity of TZS 424,281,269 at 31 March 2030. The loss case turns that into a shortfall of TZS 87,280,871, with equity of TZS 212,719,129, still 10.64 times the minimum. Of the gap of TZS 211,562,140 between the two cases, loss expense explains 48.64%. The rest is income the company did not earn because it lent less, which the book calls the second cost of a poor season. The book's decision rule is that a lender grows below 3% portfolio at risk with a profit, holds between 3% and 10% or in a loss year, and stops and takes advice above 10%.

The monthly pack puts capital first: equity, the TZS 20,000,000 floor and the headroom in the loss case. Cash is checked against the 5% liquidity line before every disbursement run. The digital lender reports on the same cycle and adds a complaints desk. A credit complaint must be closed within 14 days. Where a complaint goes to the Bank, the lender answers within 10 days and the Bank decides within 60 days under regulation 11, although the day on which that period starts is to be confirmed with the Bank. Every change of price, product, screen or platform goes to the Bank before it goes live, and a store removal is reported within 14 days. The desk received 444 complaints in the first year.

The book states that a decision of the Bank which is not carried out is liable to a fine of TZS 1,000,000 a day, and that thirty days of defiance costs more than the best month's result. Advertising matters as much as price. In the digital case, the July 2027 push changed who borrowed without changing the price or the number of loans, which is why the book treats marketing messages as part of the lending decision. The year-one review, held on 1 March 2028 in the model, is written with its stop conditions and is made from the loss case as well as the plan.

What to confirm before acting

Both books state their limits, and a reader should take them as seriously as the figures. Neither book reads the Zanzibar regime. The Mainland rules were read on 8 October 2026 from the Bank of Tanzania's requirements and guidance, the Microfinance Act, 2018, GN 679 of 2019 and the Guidance Note on Digital Lenders of August 2024. The Tanzania Legal Information Institute lists no amendment to GN 679 as read. However, the Bank's feedback folder holds a text of Tier 2 regulations dated 2024 whose status was not confirmed, and the book cites GN 679 of 2019 throughout. The shareholders confirm with the Bank which regulations apply on the day they file.

Several points are left open by the issuer and are marked for confirmation. These include the commencement date of the Act; the frequency of periodic reports under regulation 50; the length of the Bank's decision period for a licence, which Handbook 60 leaves to confirm while Handbook 61 cites 60 days under regulation 11, a difference the reader should resolve with the Bank before filing; the instrument that requires the reducing balance method; the day from which the 60 days run; the length of the platform clearance wait; the cooling-off period for a phone loan under regulation 14 of the consumer-protection rules; whether camera access for the identity photograph sits within paragraphs 3.1(b) and 5.1(b) of the Guidance Note; and how the 2025 amendment requiring the Bank's written approval for a new or higher fee applies to Tier 2 digital lenders.

Tax is not computed in either book. Every profit is shown before tax, and TRA and the company's accountant set the position. The fees for BRELA incorporation and name reservation, any business licence required by the council, and the rules on foreign investment outside the microfinance regulations carry no figure in either book and must be confirmed with each issuer. The Morogoro model carries TZS 3,500,000 for incorporation, legal drafting and policies as an estimate, not as a fee.

The practical sequence is the one the series sets itself. Confirm the licence route and the regulations in force, size the capital to the break-even portfolio, hold the licence before the first loan, price from cost and show the full cost, lend to verified cash, collect within a written ladder, and report on time. A reader whose plan lies in Zanzibar should not rely on these figures until an adviser there has confirmed the position with the issuer.

Farms, Herds and Fisheries is the Zatra Handbooks shelf for a reader who must decide whether to grow, buy, process, trade, export or farm a Tanzanian food, animal or fish product. Nine handbooks carry one decision each from the census map to a written choice, with a set of tools and a loss case beside every base case. The books apply Mainland law and Mainland census figures. Zanzibar is kept apart and is named wherever its rule or figure differs. The facts come from the 2019/20 agricultural and livestock censuses published by the National Bureau of Statistics and the Office of the Chief Government Statistician, and from fees, prices and permits that the handbooks state as at 10 October 2026. This summary sets out what the nine books cover, how they are built, where their numbers come from, which book fits which decision, and what must be confirmed with the issuer before anyone acts. It is general information, not legal, tax, financial or investment advice.

What the nine books cover

The shelf follows the Tanzanian farm, herd and fishery from the census map to a single business decision. Two books help a reader choose a place. Handbook 62 chooses among five crop regions, and Handbook 63 among five livestock regions. The other seven take one product through its chain: cashew (Handbook 64), sesame (Handbook 65), coffee (Handbook 66), fish (Handbook 67), meat (Handbook 68), horticulture export (Handbook 69) and dairy (Handbook 70). The numbering is not a sequence to be read in turn. A reader comes to the book that matches the decision in hand, and the reading path inside each book shows which chapter to open when some of the work is already done.

The books are written for people who must commit money before a season starts: a farmers' group leader choosing a buyer, a licensed trader setting a bid, a processor placing a loan, an agro-vet owner opening a depot, or a family company deciding whether to enter a sector at all. The books assume that the reader knows the business but not the herd, the crop, the cage or the cold chain. Each therefore explains the biology or the market in one short section, then moves to money, permits and risk.

The structure is the same across the nine. Each book has seven chapters, each one a step, and each chapter closes with a worked example of a named business type, labelled as an illustration. Recurring boxes set out what most people get wrong and the fix. Each book also carries a kit of tools to complete at a desk, among them a land or season sheet, a licence map, a lot quality sheet, a loss case, a scorecard and a monthly pack. The tools are the part of each book the reader is expected to use inside a real business, not only to read.

The series is Mainland-based, and the books keep Mainland and Zanzibar apart. Where a national total is used, the Zanzibar part is shown separately, and where a Zanzibar rule was not read, the book says so and leaves the check to the Zanzibar authorities.

The census base and its limits

Most figures in the series come from the 2019/20 agricultural census published by the National Bureau of Statistics and the Office of the Chief Government Statistician in 2021. The livestock count was taken on one day, 1 August 2020. It recorded 33,928,391 cattle in Tanzania, made up of 33,657,393 on the Mainland and 270,998 in Zanzibar. The same count found 24,568,396 goats and 8,516,990 sheep. The regional analysis in Handbook 63 uses the Mainland figure, and the Zanzibar figure appears only in the national total.

Handbook 63 finds that five regions hold 12,731,048 cattle between them, about 37.5% of the national count. They are Tabora, Arusha, Simiyu, Manyara and Mara. Dodoma is sixth, at about 2,197,000 cattle read from the census chart, and the book labels that figure as an approximate reading rather than a table value. Both livestock and crop books use the same census but rank regions by different measures, which is why the same region can lead one table and trail another.

For crops, Handbook 62 fixes five leading regions by a written rule that uses tonnes harvested rather than area planted. Ruvuma leads maize with 498,685 tonnes, or 7.7% of the Mainland crop. Morogoro leads paddy with 572,884 tonnes, or 17.2%. Mtwara leads cassava, and it also holds 201,411 tonnes of cashewnut, 51.6% of the Mainland crop. Kagera leads beans, and Dodoma leads sunflower with 40.3% of the Mainland crop. The book gives Tanga as the reason for the tonnage rule. Tanga has the largest maize area, 397,028 hectares, yet harvests less maize than Ruvuma or Manyara.

The census has limits, and the books say where. Several readings are taken from charts and marked as approximate. The 2019/20 season is a past year, and the handbooks pair census values with working assumptions, such as the rains for a region, which the reader must confirm with the Tanzania Meteorological Authority before planting. Regional tonnes also say nothing about what one plant can buy, so the books convert totals into farm counts, catchments and buying points before they size a plant.

Losses after harvest need the same care. In its storage chapter, Handbook 62 cites Christiaensen and Kaminski (2014), whose household surveys in Malawi, Tanzania and Uganda find farmers reporting on-farm post-harvest losses of 1.4% to 5.9% of the national maize harvest, against an FAO figure of 8%. The book's advice is to write each loss into the reader's own weighed record rather than adopt a borrowed national average.

The method every book shares

The books share one method, set out before the first chapter. Each figure is computed from inputs printed beside it. Each rule carries its issuer and its date. Each illustrative case is labelled as one. Each book closes with a decision that can be written down, and with the conditions that would reverse it. The method is what makes the nine books comparable, even where the products differ.

The loss case is the most consistent feature. Every book sets a base case beside a loss case, and the loss case is not a worst-day guess. It is a named set of shocks: a price fall, a poor season, a rejected lot, a failed chiller or a disease hold. Handbook 62 shows the effect on an oil press, whose base return of 23.87% falls to 5.59% in the loss case. Handbook 65 shows a 12% fall in the buyer's price, with all stock unsold, turning a planned profit of TZS 231,387,500 into a loss of TZS 298,482,100.

Buying discipline runs through the books as a ceiling. Handbook 64 sets a cashew ceiling of TZS 2,880,730 a tonne at a 23% outturn, and states that the ceiling moves with the outturn and the kernel prices, so the owner recomputes it before raising a bid. Handbook 65 sets a sesame ceiling before each auction round, at today's export price and at a price 12% lower. The rule is the same in each case: the highest bid that still pays, not the highest bid the market will accept.

Scorecards show how each choice is reached. Handbook 62 fixes its weights before scoring and then ranks five regions: Dodoma scores 87.0 out of 100, Morogoro 81.0, Ruvuma 71.0, Mtwara 70.0 and Kagera 39.0. Handbook 63 writes its decision rule before the scorecard: the highest score wins, provided its base-case operating result, with capture cut for competition, is a profit. Manyara scores 450 of 500 and passes, while Arusha scores second on paper and fails the test. Handbook 64 scores three cashew routes out of 500, with the kernel route at 380, the second at 350 and the third at 245.

The books do not recommend. Handbooks 66 to 70 say so in their closing notes, and the figures in every book are the case's own figures, not advice to any other business. A reader who copies a case without its inputs copies a number without its meaning.

Choosing a farming region: Handbook 62

The starting question is which of five crop regions should hold a fixed sum of capital. The book's case is a family company that holds TZS 1,500,000,000 of its own capital and requires any new line to earn 18% a year, or TZS 270,000,000. Its maize mill buys most of its grain through traders it has never met, which is why the board wants a business of its own nearer the farms.

The land and season chapter tests each region's land route, buying window and working capital. Each plant site is classed as a lease, general land or village land, and the book asks whether section 20 of the Land Act, Cap. 113 applies, since that section treats a company with a non-citizen majority as a non-citizen. The model shows Ruvuma with one buying turn at TZS 1,120,000,000, Morogoro three turns at TZS 850,000,000, Mtwara six turns at TZS 165,000,000, Kagera two turns at TZS 1,125,000,000 and Dodoma one turn at TZS 1,045,000,000.

The Dodoma plan carries the sharpest warning. Its oil press has the highest base return of the five, 23.87%, and the thinnest working capital headroom, TZS 5,000,000, because seed is bought from June to October for TZS 1,045,000,000 against working capital of TZS 1,050,000,000. The book shows that a buying price above TZS 954,545 a tonne would exhaust the working capital at full volume, and that at the loss-case price of TZS 1,045,000 a tonne the full 1,100 tonnes would need TZS 1,149,500,000. The title to the Dodoma plot had not been seen at the date of the case.

The permit chapter finds that each of the five plans needs five or six registrations, licences and certificates from at least five issuers. The example of an agro-dealer in Songea shows why the order matters. It sold fertiliser from a second store that was not registered as premises, which the book sets against the Fertilizers Act, 2009, and the TFRA registration must be confirmed with the Authority before the planting season.

The scorecard gives Dodoma 435 points out of 500, or 87.0 out of 100, on weights fixed before scoring, against 81.0 for Morogoro. The book does not stop at the score. It asks the reader to write the loss case beside the choice and the conditions that would reverse it. For Dodoma the reversal lies in the headroom, which the scorecard does not show and which the book therefore carries as the case's real weakness.

Cashew: buying against a ceiling

Handbook 64 follows a cashew nut from the grower through the buying season, the Cashewnut Board registration and the auction, to a plant that shells, grades and exports. The Mtwara belt counted 201,411 tonnes of cashewnut in the 2019/20 census. The model's grower sells 600 kg at TZS 2,600 a kilogram and keeps TZS 1,140,000 after an assumed TZS 420,000 of labour. In a low-price year, at an assumed TZS 1,800 a kilogram, the same harvest leaves TZS 660,000, because the costs were paid before the price was known. The first lesson of the book is that a grower carries the longest wait and the smallest buffer.

Chapter 2 is the auction ceiling. The model starts from an auction price of TZS 2,600,000 a tonne, takes the room left after the costs that do not change with the price, and carries that room through the interest. At a 23% outturn it reaches a ceiling of TZS 2,880,730 a tonne. The book calls this a model figure. The ceiling moves with the outturn, and each lot is priced by its own test result rather than the society's average. The Lot Quality Sheet records the moisture, the count and the outturn for each lot.

Chapter 4 finances the season. The loan climbs for three months and takes six to clear. It peaks at TZS 1,550,747,534 at the end of December, with a cover of 67.75% against 850 tonnes at cost and interest of TZS 75,040,943 for the season. A 15% fall in the value of the stock would need TZS 188,762,784 repaid in December. The book states that none of these figures is a forecast or a statement of a lender's terms, and that the interest rate is the model's assumption.

Chapter 5 covers the raw route. The assumed levy is the largest cost a raw tonne carries after the nut itself. The book also says that a lower auction price is a lower price to growers, and it does not treat a fall in the auction price as the company's good fortune alone. That point matters to any reader who plans a buying margin on a falling market.

Chapter 6 covers the plant. Shelling and grading turn a raw nut into a packed kernel, so the plant needs its own council licence under the Business Licensing Act, and the book cites the Act in the form discussed in the licensing section below. The Cashewnut Board registers processors, buyers and exporters. The book states that the registration class, its conditions and its fees were not read at source and must be confirmed with the Board. Chapter 7 scores three routes on six weighted criteria out of 500. The route that shells all 1,000 tonnes and sells half the kernels forward scores 380, the second route 350 and the third 245. A signed forward costing more than the assumed 3% reverses the choice between the two processing routes.

Export trades: sesame and avocado

Sesame and avocado are the two export chains in the series. Both depend on a licence on the export side and on a buyer's standard or an exchange rule at the point of sale. Handbook 65 is set in Lindi. Its case company buys 1,200 tonnes of raw seed in six auction rounds on the Tanzania Mercantile Exchange, cleans it to 1,170 tonnes of export grade, and ships 60 containers of 19.5 tonnes at a model price of USD 1,480 a tonne. The book lays the exchange calendar beside the farm calendar, from October 2026 to November 2027, and states that auction dates are confirmed with TMX each season.

Two lessons run through the sesame case. The first is that the auction purchase is 83% of the cost of a cleaned tonne, so a bid above the ceiling costs the season more than any other line. The second is that a failed container costs far more at its destination than in the warehouse. A container rejected at the destination carries TZS 29,393,000 of cost, the lost contract value and the return. The same lot found out of specification in the Lindi warehouse costs only the work to correct it or a local sale, and the company buys a replacement lot at the next round.

The loss case is a 12% fall in the buyer's price between the first round and shipment, with nothing sold forward. The planned profit of TZS 231,387,500 becomes a loss of TZS 298,482,100. The book ranks this as the only single risk that turns the season into a deep loss. It then tests the controls. A forward sale closes part of the exposure, and a letter of credit costs least on one container and is the only payment term with nothing exposed.

Handbook 69 is set in the Southern Highlands and follows Hass avocado from the company's own 20 hectares, planted in 2020, together with 269 outgrowers under group certification. The fruit is shipped as 36 containers to an importer at Rotterdam, through Dar es Salaam. The sea trip takes 36 days door to door, of which 26 are at sea. The book treats the ten days on land as the company's to manage, and the four days at the port as the most likely to grow.

The regulatory question comes first. Under the Cereals and Other Produce Act, 2009 (COPRA), a licence is needed to export cereals or other produce commercially, and the Act's term 'other produce' covers only produce the Minister has gazetted. The book does not establish that fresh avocado is on that list, and it states that the answer must be confirmed with the Authority before a container is booked. The financial case then runs in two forms. The base season earns TZS 581.7 million, and one rejected container cuts it to TZS 449.6 million, since a rejection costs TZS 132,080,000 under the contract. The loss case, with a carton price of TZS 22,000 instead of TZS 25,000, only 30 containers sailing and one rejection, loses TZS 168,547,600.

Livestock: the depot and the meat chain

Handbook 63 opens an agro-vet and livestock depot in Babati, Manyara. Its census region is one of the five that hold about 37.5% of the national cattle count, but the scorecard decides the region, not the size of the herd. Manyara scores 450 of 500 points and earns an operating profit of TZS 16,925,593 in the depot's first year, which opens on Monday 1 February 2027. Arusha, the best-known name, would have been chosen on reputation, yet it turns into a loss of TZS 67,585,444 once competition is counted. If capture halves and the margin slips to 21%, Manyara's own result falls to a loss of TZS 47,983,303. The decision memo carries an exit trigger beside the budget.

The depot's costs are carried line by line. The veterinary officer is budgeted at TZS 1,400,000 a month, or TZS 16,800,000 a year, and the book treats the post as a cost to keep, because the officer is the reason the depot may lawfully sell prescription medicines. Vaccines need a cold chain, and a broken chain spoils the batch without any visible sign. Hides and skins add a second margin, which is lost to knife cuts, scars and slow drying. The trade is regulated under the Hides, Skins and Leather Trade Act, No. 18 of 2008, with registration, grading and defects regulations of 2010, and the sections and any fee are to be confirmed with the Ministry of Livestock and Fisheries.

Handbook 68 follows the same animals into the meat chain. Its case is a trader near Dodoma who adds fattening, an abattoir and two butcheries. The book prices each link per head, from auction trader to exporter, and costs 100 steers over 90 days. Feed is the largest cost after the animal itself, at TZS 297,000 of a TZS 1,283,000 steer cost. Movement of animals is controlled under the Animal Diseases Act, No. 17 of 2003 and the Animal Diseases (Movement Control) Regulations, G.N. No. 28 of 2007, and a 45-day foot-and-mouth hold is costed as an extra case rather than left out.

The abattoir is the least forgiving link. Its break-even is 258.1 slaughter days, and its after-interest figure is 281.1 days, so each day the line stands idle is paid for in full. The book values each lost day at TZS 1,776,000 of contribution. The butcheries earn most in December, at TZS 79.2 million, and their loss case is a failed display chiller. Sales fall 10%, from 72,000 kg to 64,800 kg, and trim, shrink and spoilage rise from 6% to 10%.

The disease evidence is cited with care. The book refers to a study of 878 outbreaks in 605 villages from 2001 to 2006 (Picado and others, 2011), concentrated along the Kenya and Zambia borders and in the Kagera basin. Dodoma lies away from those borders, which lowers the risk without removing it. Zanzibar's own rules on animal movement and markets were not read for the book and must be confirmed with the Zanzibar authorities.

Fish and dairy: cages, cold chains and drought

Handbook 67 covers a tilapia cage farm and cold store near Musoma, with the trade options for the lake and the coast. Its governing rule is to find the buyers first. A farm's harvest date is fixed six months ahead, and fish lose value from the hour they leave the water, so the book requires three buyers to be filled before any build, with written confirmation from two. The owner's licence map has ten items in the order they are needed. Three sit at the farm stage and must be lodged before a cage goes into the water.

The farm stocks two of its twelve cages a month from 1 March 2027. The first harvests come in September 2027, after which two cages are harvested and two restocked each month, a yield of 81,600 kg a year. Feed is nearly half the cost of every harvest in the second year. Frozen and export channels net TZS 1,680 to TZS 1,850 more per whole kilogram than the beach price, which is why the book costs a cold chain from landing to buyer. Cash falls for six months before the first harvest and bottoms at TZS 55.4 million in August 2027.

The exit trigger is the book's most practical device, and it is signed in advance. If the three-month average survival on harvested cages falls below 70%, no new cages are stocked until a fisheries officer or a veterinarian has advised on the cause. If forecast cash falls below TZS 50,000,000 at any month end in the next six, stocking is cut from two cages a month to one and the bank is told in the same week. A third trigger applies after two weak quarters in succession.

Fish leaving for Kenya needs papers on both sides of the border. The book names an export permit, a health certificate and proof of origin as the documents usually discussed, and it cites the Fisheries Act, 2003, under which the Minister may make rules on fish imports. The procedures, the issuers and the fees were not read for the book. The owner confirms them with the Ministry responsible for fisheries and with the importing country's authorities before the first load, and agrees the grading rule in writing.

Handbook 70 follows a twenty-cow dairy near Muheza that buys milk from about 45 neighbours at TZS 900 a litre and makes pasteurised milk, yoghurt and cheese. The national census counted 836,056 improved dairy cattle, and the book notes that Tanga region held the most improved dairy cattle. The first test is cost. The farm's own milk must cost less per litre than the gate price, so the book builds the cost from concentrates, hay, water, breeding and labour month by month.

The base year makes an operating profit of TZS 38.5 million. Disease alone, with lost milk sales as the largest cost of a bad spring, cuts that by TZS 7.9 million. Drought from June to October cuts it by TZS 47.6 million, and the two together by TZS 55.6 million, leaving an operating loss of TZS 17,125,322. After loan payments the same year shows a cash shortfall of TZS 29,575,322, which is more than the TZS 15,000,000 reserve. The fodder reserve is the first defence. A trigger written in advance moves the herd to nine bales a day from July if the fodder plot is short.

The plant is priced against the raw route. Its fixed costs of TZS 10,861,600 a month run whether the tank is full or not, and the drought case still ends in an operating loss for the year. The registration path has sixteen approvals in seven layers, and every Dairy Board and animal health line is still to be confirmed with the issuer, under the Dairy Industry Act, No. 8 of 2004. Zanzibar is stated apart in the book's own Part H.

Licences, and the Mainland and Zanzibar line

Each book ends its licence work in a map. A map lists each permit, its issuer, its instrument as read, the date applied for, its status and a 'confirm with the issuer' note wherever the text was not read. The maps are ordered by the season, not by the office. Handbook 62 gives the warning in concrete form: a press can be built, staffed and registered and still be unable to buy seed, because its COPRA licence is pending. The standard across the series is to order the map backwards from the first purchase and the first sale, and to apply first for the item with the longest lead time.

The citation of the core licensing statute needs one check before a reader relies on it. Handbook 65 cites the Business Licensing Act, Cap. 101 R.E. 2023, for the export of agricultural goods. Handbook 64 cites the same Act in its 2002 consolidation as read on TanzLII, and notes that the section numbers in the current revision must be confirmed with the council. Handbook 70 cites the Business Licensing Act, No. 25 of 1972. The three forms do not agree, and the reader should confirm the current citation with the issuing council and the Ministry before drafting an application from any of them.

Zanzibar appears in three ways. The first is the census: Zanzibar's 270,998 cattle sit in the national total and nowhere in the regional analysis. The second is tax: Handbook 65 prints the VAT rate as 18% on the Mainland and 15% in Zanzibar, and it prints no Zanzibar regime. The third is regulation. Handbooks 64 and 68 state that the Zanzibar regime was not read for the book, Handbook 65 says its Zanzibar regime is not printed in that part, and Handbooks 67, 69 and 70 state Zanzibar in their own appendices.

Handbook 66 confirms the census point for coffee: the crop is grown in 15 Mainland regions and in none in Zanzibar, so the chain is set on the Mainland alone. A reader who plans a business in Zanzibar should begin again from the Zanzibar regime, because licensing, fisheries, animal movement, cashew and dairy rules there are not the Mainland rules these books describe.

Mistakes the series names most often

Several errors recur across the nine books, and each book names them in its own box. The most common is to build a budget on a fee taken from a neighbour, a supplier or an old receipt. Handbooks 63 and 70 both say that fee schedules change, that councils set their own, and that a figure from another district or year is a guess printed as fact. The series prescribes one fix: record each fee with its issuer and the date it was read, and label any unread line as an assumed allowance until the receipt replaces it.

The second is to treat a contract as a guaranteed supply. Handbook 62 warns that an outgrower contract fails quietly when a better price appears at the roadside, and that the plant's volume must not rest on contracted tonnes alone. Handbook 64 warns against blaming a whole society for one poor lot, and Handbook 66 warns against answering a rival's higher price after the cherry has gone. In each case the fix is to count contracted volume as part of the catchment, not all of it, and to price each lot on its own result.

The third is to build before the buyers are found. Handbook 67 requires three buyers filled before the first build. Handbook 69 puts the buyer's farm standard, the audit and the residue tests into the budget before the first carton leaves, because the standard is a year-round obligation rather than a one-day event. Handbook 63 shows the cost of the opposite habit: the veterinary post is cut when sales are slow, although it is the reason the depot may sell medicines lawfully.

The fourth is to read the licence map as a list. Each book orders its permits backwards from the first purchase and the first sale, applies first for the item with the longest lead, and records the date each item must be in hand. A permit that arrives after the season is as useless as one that was never applied for.

Reading order: which book fits which decision

The books can be read in any order, but the paths differ. A reader with capital and no region should begin with Handbook 62, which builds the region choice from the census and ends in a signed 90-day entry plan. A reader choosing where to hold a herd should begin with Handbook 63, and then turn to Handbook 68 for fattening, butchery or export meat, or to Handbook 70 for dairy.

A buyer in a crop chain should begin with the buying ceiling. Handbook 64 is the book for cashew, where the ceiling is written before the first bid, and Handbook 65 for sesame, where each auction round has its own ceiling. A coffee business should read Handbook 66 on the step that makes the margin, from grower to pulpery to roastery. A fisher moving into trade, or a family company planning cages, should begin with Handbook 67 and its buyers' table.

A company with export ambitions in fresh produce should read Handbook 69 before any container is booked, because it sets out the COPRA question, the rejection clauses and the eighteen export documents. Its counterpart in the milk trade is Handbook 70, which applies the same logic to the herd: size it, hold the drought reserve, and test the price against the alternative.

Each book's reading panel names the condition under which a reader may skip a chapter. Handbook 62, for example, tells a reader who already holds a written selection rule and a shortlist of five regions, each with its census figure and the date it was read, to begin at Chapter 2. For a reader who holds some of the work already, the panels are the faster route than the contents page.

Figures to confirm before acting

The books state their figures as at 10 October 2026, and each says so on its title page. Rates, fees, deadlines and auction calendars move. The series asks every reader to confirm each figure with the issuer named beside it, and this summary keeps the same rule. Nothing here is a quotation of an official fee, and the figures from the case models are illustrations of method, not forecasts.

Several items were expressly not read at source and must be checked. Handbook 62 treats the season and the rains for each region as working assumptions to be confirmed with TMA, and it records that the Dodoma plot's title had not been seen. Handbook 63 reads several cattle figures from a census chart and marks them as approximate. Handbook 64 leaves the Cashewnut Board registration class and its fees unconfirmed. Handbook 67 leaves the Kenya export procedures and fees unconfirmed.

Handbook 69 leaves open whether fresh avocado is gazetted produce under COPRA, which makes the export licence path a question for the Authority. Handbook 68 leaves Zanzibar's rules unread. Handbook 70 assumes a statutory employer cost of 14% on its staff, which the reader must replace with the rate in force, and budgets its fee lines as assumptions until each receipt is held.

A reader who changes any input should rerun the loss case before reading the answer. Each book's conclusions depend on the inputs printed beside them, and the books say so in every chapter.

Money at Home is the seventh shelf of the Zatra Handbooks, a set of seven books for a Mainland Tanzanian household that earns, spends, saves, borrows, pays for school and plans for retirement. The books are working tools. Each asks the reader to fill in a sheet, a card or a table with the household's own figures, and each works the method through a few worked households. Zanzibar is not covered by any title, and each book's local appendix says so. The series rests on three rules stated at its front: every figure has an issuer and a date, every worked case is labelled as illustrative, and every chapter ends with a tool for use that week. The 2026 editions state their appendix rules at 10 October 2026, and several rules are marked for confirmation with the issuer. This summary explains what each book covers, how the books connect, and which figures change most often. It is general information, not financial, legal or tax advice, and no figure in it should be acted on without the check the books recommend.

What the series covers

The seven books move from the month to the lifetime. Personal Finance for Tanzanians (Handbook 71) and The Family Budget That Works (Handbook 72) set the monthly base. Personal Goals You Can Measure (Handbook 73) turns goals into dated weekly actions. Money-Eating Monsters (Handbook 74) finds the small habits that take money without notice. Get Out of Debt in Tanzania (Handbook 75) clears loans in an order the household can check. School Fees Without Debt (Handbook 76) prices schooling and the student loan. Retire on Your Own Terms (Handbook 77) tests what a pension will pay from age 60 and how to fill any gap.

All seven share one method, which the first book names as map, count, split, fund, protect, limit, learn and review. The method is the same whatever the income, but the figures differ. Each book works its steps through a few households of different sizes, from a household on a few hundred dollars a month to one with a higher income. Worked households are described by type, such as a driver, a nurse or a teacher, and every worked example says that it is an example.

Each book runs to roughly 170 to 185 pages. Each carries a set of tools for use at a desk, a Kiswahili summary, notes by chapter and a list of sources. Where a source could not be read at the time of writing, the appendix says so and marks the rule for confirmation. That marking keeps a household from relying on a figure the writers could not check at source.

The month comes first: Handbooks 71 and 72

The first step in Personal Finance for Tanzanians is to find the income that can be counted. The book's test is the floor, the average of the weakest three months in a row, rounded down. A payslip alone is a poor base for many households, because allowances, sales, tuition and irregular amounts arrive outside it. In the book's driver household the twelve months must add up to TZS 7,200,000 for the year, and a total that falls short shows a month that was left out. A savings group's payout is the household's own money coming back, so it is kept off the income line.

The second step counts what the month really costs. The driver household believed it spent TZS 520,000 a month. Its count came to TZS 640,000, a gap of TZS 120,000, of which TZS 70,000 was leaks that nobody had chosen. Over a year, the book says, a leak of that size comes to TZS 840,000. The third step sets the split before the month begins. The book tests the 50/30/20 rule and sets it aside, because essentials take between 70% and 85% of planning income in the households it models. The sheet is built from the bottom up: essentials, relatives, loan repayments, cover, a limited wants line and the saving line, which is paid first on the day money arrives. The driver household shares TZS 580,000 of planning income and sets a saving line of TZS 40,000. Half of every raise goes to saving.

The emergency fund is built in two stages, one month of essentials and then three months, or three and six months for an owner. The driver household's first stage is TZS 490,000. It starts from a seed of TZS 300,000 from a share-out and a deposit of TZS 40,000 a month, and it passes stage one in Month 8 of the worked year. The fund sits apart from the spending wallet, and a written test decides what counts as an emergency. The book also prices cover. In the worked case a premium of TZS 15,000 a month, or TZS 180,000 a year, turns a hospital bill of TZS 450,000 into a household payment of TZS 90,000. Borrowing is then limited by three ceilings set before any offer arrives: repayments of at most 20% of planning income, balances of at most three months of income and a price of at most 5% a month.

The second book, The Family Budget That Works, takes the same logic down to the week. The household's month fits on one page of eleven lines that add up to an income it can count on, and the month is cut into four weeks, because money arrives and leaves in days. Fixed bills are placed in the week after the pay that carries them. Food and transport are kept in weekly envelopes, checked on Sunday, and a week that runs over is carried and recovered rather than repaired from a later bill. A term's school cost is divided by the paydays before it falls due and paid into a named pot. Giving is treated as a bill with no date, and a line is set for it. The worked households are on USD 300, USD 1,200 and USD 6,000 a month. The USD 6,000 household's count of its month came to USD 6,270, and a 30-day tally found seven leaks worth USD 870 that closed the gap once they were trimmed.

Goals and the habits that drain money: Handbooks 73 and 74

Personal Goals You Can Measure starts from a plain observation: most January goals die by March, not for want of effort. The teacher in Arusha in the worked case wrote nine wishes on 2 January 2026, two with a date, one with a number and none with a weekly action. Her days of action fell from 22 in January to 14, 6 and 1 in the three months that followed. The book's answer is a target with the four parts named above, and a goal is sorted into one of four kinds, each of which counts something different. The four kinds count different things, such as net worth in dollars, hours of study, steps finished and fee money paid. The weekly action is sized by arithmetic: the quarter's gain divided by thirteen and rounded up. In the teacher's case the figure came to USD 17 a Monday.

The working unit is a thirteen-week cycle of 91 days. Each week a fifteen-minute review takes four minutes to read five numbers, nine minutes to name the cause and two minutes to write one decision: hold, adjust or come back. A missed month is a gap with a number, not a verdict. The book warns against the make-up trap, in which a four-week gap closed in one month would double the weekly action and break the cap. The comeback rule instead resets the size of the action and restarts on a named Monday. A goal is dropped on purpose only when it fails a four-question test, and the people it touches are told in writing. The shared goal method for couples gives each person six wants, ranked with points from six for first place to one for last. Ties are settled by a rule written in advance. The result is a plan that shows how many months each order of saving takes. A family or team uses one page showing each person's number, read in two short meetings of 30 minutes for a family and 15 minutes for a team.

Money-Eating Monsters names twelve habits that drain a household, each priced over a year and stopped by one rule. The list runs from betting and impulse buying through borrowing apps, unplanned family requests, status spending, idle subscriptions and fake investments, then the tax on ease, cover that duplicates other cover, pressure to match friends and late penalties. The household audit scores each monster from 0 to 3 according to how often it occurs. The worked driver household scores 23 out of 36 and spends USD 1,148 a year on twelve small habits, 22.8% of its pay. The worked contractor stakes USD 3,120 across 26 bets and has USD 1,820 returned, a loss of USD 1,300. The book then ranks the monsters by the money a rule would keep, which is the cost minus the household's chosen line. The rules fit on one fridge card, and the ninety-day watch counts the money kept against a Day 1 baseline, reviewed on Monday 1 February 2027 for the worked households.

Clearing debt: Handbook 75

Get Out of Debt in Tanzania begins with a list. In the worked case a nurse in Arusha wrote six debts on 18 October 2026, totalling TZS 6,800,000, or 2.7 months of her net pay. Debt hides in six places: the payroll, the savings society, the phone, the credit book, other people and papers signed for others. The list is built from balances, not from instalments, because instalments of TZS 380,000 a month, or 15.2% of pay, can hide a debt more than seventeen times larger. Each loan is then converted into one unit, the percentage of cash received, charged for one month. A fee taken first raises the price: 20% on a loan of TZS 300,000 is 25.0% a month on the TZS 240,000 actually received. A flat rate understates the true one, so a flat 3% a month is 4.94% on a reducing balance.

The first step to stop the cycle is one signed sentence that freezes new borrowing. It is supported by a month that adds up, with a fixed debt payment, a cap on other spending and a buffer line. Two orders then clear the list. The snowball pays the smallest balance first and the avalanche pays the highest monthly rate first, and in both the freed payments roll onto the next debt. On the nurse's six debts and a fixed payment of TZS 408,000, the snowball's interest totals TZS 2,144,070 and the avalanche's totals TZS 2,007,020. The difference of TZS 137,050 is 2.0% of the debt, and it is modest because the two orders are identical until month nine. The Choice Test uses three numbers: the saving of the cheaper order, the month of the first win under each order and the household's history of dropped plans. Where the saving is small, history decides. The payoff tracker shows the nurse's plan reaching zero in 22 months, on Sunday 6 August 2028.

Talking to the lender comes before a payment is missed. A call made before the due date can win a new date, a lower instalment, one waived fee and a written record. The same call after a miss can only be noted. The appendix sets out the complaint route as it was read: the lender first, with a reply within 14 days for credit, and a Bank of Tanzania direction to reply within 10 days and decide within 60 days. A review may be asked within 7 days and decided within 21 days. These timings are marked for confirmation with the issuer. In one worked case an app lender demanded TZS 680,000 on a loan whose table showed TZS 640,280 after three payments of TZS 80,000. The gap of TZS 39,720 is the first point to put to the lender in writing.

The book's eight written household rules keep the balance at zero. They are a price of at most 5% a month for any new loan, debt payments of at most 10% of net pay, balances of at most three months of pay, a 48-hour wait before a new loan, a buffer of one month of essentials, a monthly Sunday check, one person told the total and a yearly credit report. These limits are set by the household and are not requirements of the law, and the book says so. The owner in Mwanza, with TZS 78,000,000 on four loans, shows that a refinancing offer can cost more in total. Staying at TZS 7,000,000 a month clears the debts in 13 months at interest of TZS 7,898,772. The 36-month offer in the case adds TZS 15,617,153 to the total paid.

School fees without borrowing: Handbook 76

School Fees Without Debt prices a school year on seven cost lines in every stage: fees, uniform and shoes, books and stationery, meals, transport, contributions and exams, and extras. A public primary year in the worked case costs USD 117, and 70.9% of that appears on no school notice. A private primary year costs USD 1,065, of which the fee is 58.2%. Half of the public year falls due in January, which is why the book asks for a monthly line before the term. Over seven primary years the public route costs USD 819 and the private route USD 7,455. The gap of USD 6,636 is USD 79 a month, or 22.6% of the worked driver's monthly pay.

Secondary school adds a step. On the public route the year rises from USD 187 at age 12 to USD 438 at age 13. A day year is USD 248 and a boarding year is USD 1,900, so boarding adds USD 9,912 over six years. The book also prices what sits outside the fee notice. Classes, exam fees and a device total USD 1,230 in the worked case, or 10.3% of schooling, and only the USD 1,060 of classes is a choice. Five tests then turn each offer into a decision before the first lesson. For college, the public degree costs USD 2,370 a year plus a laptop of USD 400 once, or USD 7,510 over three years. Rooms and meals are 55% of a public year and tuition is 25%, so a student loan sized to tuition and books reaches only part of the bill.

The student loan is treated as a dated process rather than an assumption. The Higher Education Students' Loans Board guideline for bachelor's degree students in 2025/26, read on 10 October 2026, sets an applicant aged 35 or under at application and an application window from 15 June to 31 August 2025. It caps tuition at TZS 3,100,000 and books at TZS 200,000 a year, and it requires repayment of at least 15% of basic salary, deducted by the employer, with a one-time 1% administration fee. An employer that fails to pay over a deduction is charged 10% of it. In the worked case the caps give a loan of USD 2,031 over three years, or 27.0% of the public degree, and two of the eight loan questions stay open. A plan enters the loan at the amount allowed, marked unconfirmed, with a line for no loan beside it.

Three children in the worked family cost USD 76,005 before loans and USD 69,912 after loans at the published caps, and their costs overlap from 2028 to 2036. The Family School Fund Sheet sets a monthly line for each household: USD 43, USD 118 and USD 276 in the three worked cases, which are 12.3%, 6.6% and 4.6% of pay. Pay rises 17.1 times from the lowest case to the highest, while the first-year bill rises 7.7 times, so the share of pay devoted to schooling falls as income rises. The sheet is reviewed each January, and the line is raised by any gap found in that review.

Retirement and the long view: Handbook 77

Retire on Your Own Terms starts from what retirement will cost, not from the pension. The reader lists seven lines of spending in today's money before any pension figure is seen. In the worked case a primary teacher in Dodoma, aged 30 on TZS 1,300,000 a month, needs TZS 890,000 a month at 60, which is 68.5% of pay. Some costs end when work ends, and others, health above all, grow with age, so the list is built from the bottom rather than by taking a share off today's spending.

The book reads the pension formula as months of contribution divided by 580, times annual pay, paid 67% monthly and 33% as a lump sum at a factor of 12.5. The appendix gives this as read from the regulation effective 1 July 2022, and it marks which regulation governs a given fund for confirmation. A monthly pension needs 180 months and is paid from age 60. An early option from age 55 carries a cut of 0.3% a month. The worked teacher's 360 months give a monthly pension of TZS 540,621 and a lump sum of TZS 39,941,381, leaving a gap of TZS 349,379 a month. Closing that gap needs capital of TZS 73,005,258. A saving of TZS 58,000 a month, or 4.5% of pay, reaches it if the saving earns 3% a year above inflation. With no real return the monthly saving rises to TZS 181,000, or 13.9% of pay.

A person with no employer builds pension rights through a voluntary route, paid as a percentage of a declared income with no deduction at source. The workshop owner in Arusha starts at 40 on a profit of TZS 4,500,000 a month and needs TZS 2,250,000 a month. The worked plan combines a voluntary pension line of TZS 415,862, two rented rooms netting TZS 178,000 a month, a manager's draw and a saving line for the rest. The shop owner in Mwanza starts at 50 with a record of 144 months, 36 short of the 180 required. Five moves in order bring the load from 100% of profit to 14.7%, which shows that a late start is workable. The book also compares where money is placed. A savings account of TZS 10,000,000 at 3% is worth TZS 9,079,018 in today's money after ten years, while a bond coupon of 12.25% pays TZS 1,225,000 a year on a TZS 10,000,000 holding but buys only TZS 827,566 of today's goods by year ten. A pooled fund that falls 30% at month 216 takes 86 months of deposits to refill, so the book tests a written glide to cash five years before the draw.

Insurance belongs to the same plan. A hospital bill of TZS 9,000,000 is told three ways in the book. With cover paying 75%, the household pays TZS 2,250,000 and passes. With a limit of TZS 5,000,000 and an excess of TZS 500,000, it pays TZS 4,000,000 and fails. With a 12-month waiting period and an illness in month 8, it pays the full bill and fails. A plan is kept alive by a yearly check on a birthday with six numbers. Flags are raised when the pot is 5% below plan, the need is 10% above plan or the load exceeds 30% of income.

Figures that change, and where to check

The series is careful about which figures it prints. Where an issuer's rule was read at source, the book prints it with its date. Where only a title, a notice or a summary was read, the book says confirm with the issuer and prints no figure in its place. The following are the first items a reader should check, because they change most often or because the books disagree on them.

The deposit insurance limit is TZS 1,500,000 per depositor under a Deposit Insurance Board notice of 1 November 2023. A rise to TZS 7,500,000 was reported on 20 September 2025, and the books do not rely on the higher figure until the Board confirms it. Government securities are offered in bill tenors of 35, 91, 182 and 364 days and in bonds of 2 to 20 years, with a 10% withholding tax on coupons that is marked for confirmation. The rates shown on the Bank of Tanzania page in 2026, 10.25% for a five-year bond and 12.25% for a 15-year bond, were not confirmed as coupon or yield, so they must be checked with the Bank. The two books that quote the 15-year bond give different auction dates, 30 September 2026 in one and 19 August 2026 in the other, so the date should be confirmed as well.

Income tax bands, health scheme contributions and the NSSF contribution shares are marked for confirmation with the Tanzania Revenue Authority, the health scheme or the fund concerned. The PSSSF contribution is given as 20% of pay, with 5% from the employee and 15% from the employer, as read at source. The NSSF figure comes from a secondary summary and must be checked with the fund. Tenants are said to withhold 10% of rent under a TRA statement of 16 January 2025, and landlords must confirm that rule with TRA before relying on it.

Fee-free basic education is the item with the least certainty. The circulars that set its coverage and the list of permitted parental contributions were not opened, so the books give no figure for them. A parent should ask the Ministry of Education, Science and Technology what the policy covers before budgeting. The student loan window, caps and deduction rules belong to the Higher Education Students' Loans Board and should be confirmed with it for each cycle.

Zanzibar is outside the series. Each book states that its rules are for Mainland Tanzania and that Zanzibar's arrangements, including its own fees, pension and tax rules, are not covered. A reader with a household or business in Zanzibar needs that island's own rules and should confirm them with the Zanzibar authorities concerned. Nothing in this series should be carried across the two sides.

A reader on a tight month starts with Personal Finance for Tanzanians and then The Family Budget That Works. A reader carrying several loans starts with Get Out of Debt in Tanzania and returns to the first book for the monthly base. A parent facing January bills starts with School Fees Without Debt, and a reader aged 30 to 55 turns to Retire on Your Own Terms. Money-Eating Monsters is useful at any stage, and Personal Goals You Can Measure supplies the weekly routine that keeps the other books in use.

Find the Opening is the second shelf of the Zatra Handbooks, written for the reader who has not yet settled what to sell, to whom, or under which licence. Six books take that reader from a list of ideas to a tested problem, then to what is already owned, a Mainland licence route, a foreign firm's route to Tanzanian buyers and a separate Zanzibar market. The first three books are general and draw their cases from Tanzania, Kenya, Uganda and Ghana. The licence selector and the Tanzania market-entry book apply Mainland law. The Zanzibar book applies the islands' own authorities and is kept apart throughout. This summary explains what each book covers, which one to open first, and which figures are most likely to change. It is general information, not legal, tax or investment advice. Every fee, rate and date should be confirmed with the issuer named beside it before anyone acts.

What the series covers

The series answers a question that comes before any product or premises: where the opening is, and what must be true before money is spent on it. The six books follow one order. Book 78 gives a directory of 300 ideas, sorted by start money, skill and weekly hours, with a thirteen-week route to a first sale. Book 79 tests whether a problem is paid for at all, and uses ten paying customers as the test. Book 80 counts what the reader already holds, from skills and phones to vehicles, plots and reputation, and turns the first of those into a sale before capital is sought. Book 81 maps the licence route for a business in Mainland Tanzania. Book 82 sets out how a foreign firm reaches Tanzanian buyers through five routes. Book 83 does the same for Zanzibar, where the approvals, taxes and product gatekeepers differ.

Each book uses the same discipline. It names one decision, shows its arithmetic from inputs printed beside it, and prints a loss case beside each gain. Start budgets in the early books run from USD 100 to USD 10,000. The Kiswahili edition prints the same figures in shillings at TZS 2,600 to the US dollar, which is an illustrative rate and not a market quotation. Every book closes with sixteen common mistakes, a set of questions readers ask, a local appendix and a glossary, and each names an open starter, a free tool that can be used in the same week.

The reader does not need all six books. The three questions that matter most are whether the idea is paid for, whether the licence route is correct, and whether the buyer is in Mainland Tanzania or in Zanzibar. Each of those questions has its own book, and the reader should not try to answer one with the arithmetic of another.

Choose an idea that has a route to a first sale

Book 78 holds 300 business ideas in thirty families of ten. Every family shares one thirteen-week route, which starts with the choice of an idea and ends with a stop, fix or grow verdict on day 90. The ideas are grouped into four capital bands that the book describes as its own working rule and not a market standard. Band 1 runs from USD 100 to USD 400 and holds 62 ideas. Band 2 runs from USD 401 to USD 1,200 and holds 97. Band 3 runs from USD 1,201 to USD 3,000 and holds 83. Band 4 runs from USD 3,001 to USD 10,000 and holds 58.

The first route is a filter, not a list. A start of USD 600 reaches 159 of the 300 ideas. The skill tag and weekly hours then reduce those 159 to a shortlist of six, and the reader circles three. The book argues that a reader who keeps comparing all 300 is reading about the 141 ideas that the money cannot start, and that each week spent on them is a week the route does not reach a customer. The filtering takes about twenty-five minutes with the first tool in the kit.

The route itself asks for five conversations before any money is spent, the smallest sellable version of the idea, and a weekly sales ledger that is read every Monday. Four modelled starts run it, in Dodoma, Arusha, Dar es Salaam and Kisumu in Kenya. The Dodoma case is a phone repair stand funded with USD 520 of a USD 600 budget. It reached its first sale on 18 November 2026, made 104 sales by week thirteen and recorded USD 1,400 of sales across the thirteen weeks. Those figures are the book's illustration and not a forecast.

The verdict on day 90 rests on three numbers read from the weekly ledger. The repeat or referred share must be at least 25% of sales. The gross margin must reach the floor set for the family, which the book places at 5%, 12%, 25%, 35%, 55% or 65% by family. The payback, computed as capital multiplied by four and divided by the surplus of the last four weeks, must be 26 weeks or less. The four cases show one stop, which cost USD 560 and is presented as a loss worth taking, one fix that cut a payback from 22 weeks to 6, and two growth steps. The one-page idea card records the whole decision for one idea and is carried forward into the next.

Test the problem before the product

Book 79 starts from a different place. The reader does not begin with an idea but with a problem that someone already pays to have solved. The method has twelve steps in five parts: the rule, the search, the test, six worked cases, and the climb from ten customers to one hundred. The book is a field guide, with one tool at the end of each step.

The search is a seven-day Problem Walk. The reader asks one question a day, aims for six conversations a day, and fills a ranking grid. The questions ask about the last time a problem happened rather than about the idea the reader already has, and they are drawn from a list of ten. The walk looks in four settings, the market, the office, the farm and the home. Each problem found is scored from 1 to 5 on size, urgency and pay, and the three scores combine into an order. In the model walk, lunch for office staff scored 80 of a possible 125, ahead of a school pick-up at 60 and tomatoes at 50.

The test is ten paying customers, reached within 40 asks and 14 days. The book counts only money that has moved. Replies, promises and polite interest are recorded, but they are not counted as payment. The pass mark is written before the test begins, and the closed test sheet gives one of three readings: go, change one thing, or stop. In the model's second round, 55.6% of asks became payers against a pass mark of 25.0%. The first round had produced 16.7%.

Price comes last and is set from the value of the problem, not from cost alone. A price has three anchors. The cost of serving one customer is the floor. Above that sit either what the customer already spends on the problem or one third of the value the answer gives the customer, which the book calls the one third rule. In the model, a tray that costs USD 6 to serve sells for USD 10 and is worth USD 30 to the farmer who uses it, so the price is set within that range.

Six owners follow the method. They are a stallholder in Kampala, a salaried clerk in Dodoma, the manager of a farm-input shop near Arusha, a nurse in Dar es Salaam, a lecturer in Accra and the owner of a transport firm in Eldoret. Three are in Tanzania and three are not, so the book is written for any market. Their earnings run from USD 250 to USD 2,500 a month, and their start budgets from USD 100 to USD 10,000. Each case prints its loss case beside the gain. In the first three cases, the loss from a test not run is four to twenty-five times the cost of the test. At one hundred customers the six cases earn between USD 1,800 and USD 7,500 of profit, while fixed costs rise from USD 600 to USD 3,600 as the business grows.

Start with what is already in hand

Book 80 begins from the reader's hand, meaning every skill, tool, contact, vehicle, plot, room, kitchen and reputation already owned. The question it sets is what is in the hand, and it answers with a count rather than a plan. The first step asks the reader to list the items on one page, the Hand Map, and the book states that a count of twelve lines is usually enough to show what can be sold.

Capital comes second in this method. The book holds that capital is the second purchase, because the first income can pay for it faster than saving can. The thirteen steps walk through the places where a hand hides: skills and knowledge, phone, tools and equipment, vehicle, plot and rooms, kitchen and home production, and network and reputation. A skill is worth listing once someone nearby has already paid for it. The reader tests this by putting five questions to ten neighbours. An idle phone, tool or machine earns when its idle hours are sold, and an idle vehicle, plot or room earns what it is let for, less what it costs to let it. The book prices each of these with worked exhibits.

Each item becomes an offer in three moves. The reader writes a five-line First Offer Card that names one unit, a price, a promise, a buyer and a delivery date. The price must clear a floor, sit inside the range that five neighbours name, and be paid on the day of delivery. The reader then asks ten named people within seven days, with the offer and the price said aloud and every answer written down. The book counts a payment as the only proof of a sale.

Twelve starting hands show the method at work. The through-case is a seamstress in Mwanza, followed from the count of her hand to a ninety-day plan that runs from Monday 12 October 2026 to Sunday 10 January 2027. Eleven of the twelve hands repaid their start in the first month, four needed nothing to begin, and every steady month netted between USD 316 and USD 1,200 in the model. The largest start in the twelve is USD 700. Each hand is described and not named, and each prints its loss case beside the gain.

The last two steps deal with growth. A partner who brings the missing piece is taken on one written page that names who brings what, how good and bad months are shared, what the trial is, what the record shows, and when the arrangement is reviewed. The first income is then split by a written rule, half to a next-step fund and half to the household. In the seamstress model the fund reaches USD 357 in three months, enough for a USD 180 overlocker in month two.

Decide the licence route before applying

Book 81 covers Mainland Tanzania only. Its opening claim is that a business there needs four things: a business licence, a sector permit, a tax registration and a local permit. Each comes from a different office with its own order and its own fee, and the three offices a reader visits often give three different answers. The book does not list licences in general. It sends the reader through five questions about who owns the business, what it does, where it sits, how large it is and what it touches, and each answer leads to one of twelve sector route cards.

The book has fourteen branches. The first explains how the licensing system fits together, and the second gives the master decision tree. The later branches set out route cards for food, drinks and retail, for medicines, cosmetics and health services, for energy, fuel and utilities, for transport, vehicles and aviation, for communications and online business, for construction, mining and real estate, for tourism, hotels and events, for agriculture, livestock, fisheries and education, for finance and microfinance, and for manufacturing, imports and professional services. The energy branch, for example, follows five stops that stand between a site and its first litre, cylinder or kilowatt, each with its own issuer.

The order of application is where the book earns its place, because a wrong order costs weeks and fees. In the model, a food maker in Mwanza needs 111 days when the steps are taken in the right order, and 125 days after one refused step, which is 14 extra days. The book keeps the council business licence (Class B) first and the sector permit second, as the fee register lists them. It notes that the national business portal shows the sector permit earlier, and each card flags where the two orders differ. Every route card closes with a line headed What this route prevents.

The book names the costs of a wrong route in five headings: delays, mistakes, surprises, rejections and inconveniences. The model cases give the sums involved. In one case a consultancy's licence costs 26 times the local fee. A lodge faces levies of up to TZS 1,708,200. A trader's late renewal costs TZS 62,500. An incentives path that TZS 40,000,000 of capital cannot meet is set out so that the reader can see it before choosing the route.

Fees are quoted from the Handbook 06 fee register as stated at 26 September 2026, in the currency the instrument prints, with the register ID and the issuer. The Business Licensing Act is cited as Cap. 101 R.E. 2023, formerly Cap. 208. Where a rule could not be read at its source, the book says confirm with the issuer, and the reader should treat that line as open until the issuer answers. The licence file and the renewal calendar in the final branch keep every date in one place so that a licence does not lapse without warning.

How a foreign firm reaches Tanzanian buyers

Book 82 is for a foreign manufacturer, brand or service firm that wants buyers in Mainland Tanzania, and not only a registered company. The book opens with the buyer, not with the vehicle. It holds that demand is proved by what buyers say on a call, what they pay on a shelf, and what they order in a trial, and never by how warmly a visit is received. The five plays are then compared before any company is formed: an agent, a distributor, a branch, a subsidiary and a joint venture.

The Route Scorecard rates the plays on cost, control, speed and risk, and a break-even ladder adds a view in shillings. In the model an agent takes 8 of every 100 in list value as commission, a distributor buys 20 below list, and a branch or subsidiary keeps the 40 left after the cost of the product. A joint venture keeps the maker's share of that 40. The ratings are Zatra's judgement and not evidence, and the lead times, rents, wages and rival prices are labelled as assumptions.

Making the product sellable comes before selling it. The desk that owns the product's kind of goods is named before a container is booked. The book names the Tanzania Bureau of Standards and the Tanzania Medicines and Medical Devices Authority for the products they cover. Landed cost is built line by line from the factory gate to the warehouse door, and the unit price is that total divided by the units. A shelf price survives the channel only if every rung from the warehouse to the buyer has been counted, and the book shows which rung is cut when a price must be held.

Buyers are won in three ways. Government sales run through the NeST tender system, where a registered firm lodges a complete bid before the deadline and is paid long after the goods are delivered. Partners are vetted on ten questions answered with evidence, and a partner is safe to sign at 8 of 10. A contract turns the vetting into rights through ten clauses, each answering a loss that a case in the book has already shown. The first year is four dated gates and one review, and it ends when the file shows how many buyers came back.

Some figures in the model need confirming before anyone relies on them. The register as stated at 26 September 2026 gives, for a manager from abroad, a class B-1 residence permit at USD 2,000 with a further USD 50, and a class A work permit at USD 1,000, a total of USD 3,050. A citizen of an East African Community partner state pays USD 500 in place of the USD 2,000 residence charge. A service fee paid by a subsidiary to a non-resident parent carries 15% withholding, against 5% where the payee is resident. The model also applies 10% withholding to a dividend share. Each of these should be confirmed with the issuer before it is used.

The five cases show what each route does in practice. A skincare maker sends a first order of TZS 20,000,000 FOB through an agent. A hardware maker sends one container of 3,000 units, TZS 150,000,000 FOB, to a distributor. A medical consumables firm forms a subsidiary with TZS 60,000,000 of capital and bids for public tenders. An agent paid a retainer of TZS 1,400,000 a month delivers no order in a year. A joint venture is priced once above its cost and once below it. Two of the five cases are failures. The book covers Mainland Tanzania only, and Zanzibar is left to Book 83.

Zanzibar is a separate market

Book 83 applies Zanzibar's own authorities to a seller who wants customers on Unguja or Pemba. It does not assume that a Mainland rule applies there. Where a Zanzibar rule was read, the issuer and the date are named. Where it was not read, the book says so and asks the reader to confirm it with the issuer. A seller from the Mainland should not assume that the Mainland approvals, tax authority or product gatekeepers carry over.

The buyer atlas comes first. Hotels and resorts are the largest buyer. They buy through several desks and pay about 45 days after a signed delivery. Tour operators pay their suppliers about 30 days after the invoice, and an excursion day is a chain of small purchases. Restaurants buy fresh and early and pay in about 14 days, but a restaurant can close owing a month. Government and public projects are found in a notice and won on a registered bid that survives the preference margins. Traders, wholesalers, households and shops pay in a week or three, in shillings, and follow the tourist season only at the edges. The atlas describes five buyer groups across its chapters.

The approvals form a ladder of seven rungs, and each rung carries a flag. The first is the Union question, which asks which Union duties or approvals follow a Mainland seller. It is not yet read, and it is marked for confirmation with the Tanzania Revenue Authority and the Zanzibar Revenue Authority. The second rung is the company itself, whether a subsidiary, a branch or none. The third is the investment authority, ZIPA, which decides whether a certificate is needed at all. The fourth is premises, which may be a lease of up to 99 years rather than a purchase. The fifth is the tourism licence, the sixth the product gatekeepers at the quay, and the seventh tax.

Fees are quoted as read. The ZIPA fees read on 12 October 2026 are given as USD 1,400, which is TZS 3,640,000 at the book's rate of TZS 2,600 to the dollar. The book also prices a two-stage path with an application fee of USD 200, or TZS 520,000 at the same rate, for a step that may not have been needed. Zanzibar has fifteen taxes and levies. VAT is the one that most plainly reaches a supplier's invoice, and the book works through what a 15% VAT charge does to a price. Each figure should be confirmed with the issuer named on its rung.

Product rules sit with two gatekeepers, a standards body and the food, drugs and cosmetics authority, and a seller can meet one and still find the shelf closed by the other. The local partner chapter makes a hard point. A partner who holds the papers does not thereby hold the customers. In the model, a food supplier that ships on the strength of the papers loses TZS 18,000,000. The dollar pricing chapter adds a second risk. Many buyers see a dollar price, but the seller's costs are in shillings. In the book's example, a shilling 10% stronger turns a year that earns TZS 24,480,000 into a shortfall of TZS 480,000.

The seasons shape cash. Zanzibar's year has five high months, five shoulder months and two low months, and the book warns against planning sales, staff and cash on the yearly average. For a linen supplier that signs one hotel or tour operator contract a month from November 2026, the model shows TZS 11,308,800 earned by October 2027. The same business needs TZS 35,723,200 of cash by August 2027, because profit on paper arrives long before the money does. The book's closing chapter turns that pattern into a twelve-month calendar of contracts, pricing and cash.

Fees and dates that move

Four kinds of figure change most often across this series, and each needs a date and an issuer before it is used. Mainland licence fees are quoted from the Handbook 06 fee register as stated at 26 September 2026, in the currency the instrument prints. Foreign-manager permit fees and withholding rates in the market-entry book come from the same register at the same date. Zanzibar figures were read on 12 October 2026, and several of its sources were read second hand, which the book marks in its flags. Positions across all six books are stated at 10 October 2026.

The series uses one illustrative rate of TZS 2,600 to the US dollar throughout. The rate is chosen for arithmetic and is not a market quotation, so a reader who budgets in one currency should convert at the rate on the day of payment. Where a fee is printed in dollars, the books keep it in dollars and never add dollars to shillings.

A second kind of figure is an assumption. Lead times, rents, wages, the carry rate on late payment and a rival's price are labelled as assumptions, and the scorecard ratings in Book 82 are Zatra's judgement. Figures in the worked cases are illustrative and computed from stated inputs. None of them is a forecast, a promise of return or a benchmark for a particular trade.

Laws and regulators are cited as the books read them. The Business Licensing Act, Cap. 101 R.E. 2023, formerly Cap. 208, is one example. Where a point was not read at its source, the book says confirm with the issuer. The reader should treat each such line as open until the issuer has answered in writing and should keep that answer on file with the other papers of the business. Nothing in the series is legal, tax or investment advice, and it sets out method and the questions to put to the issuer.

Reading order and where to check

The six books can be read in order or by need. A reader who has no idea yet starts with Book 78, chooses one idea by capital, skill and hours, and runs its thirteen-week route. A reader with a clear idea but no proof that people pay starts with Book 79, whose Problem Walk and ten-customer test come before any spending. A reader who holds a skill, a vehicle, a plot or a kitchen but has no capital starts with Book 80 at the Hand Map and reads on to the first offer.

A reader about to open a business in Mainland Tanzania should work through Book 81 and fix the order of applications before any form is filed. A foreign manufacturer or service firm that wants Mainland buyers reads Book 82 after the licence route, because the route into the market affects both the price and the approvals. A seller who wants customers in Zanzibar reads Book 83 on its own, since none of the Mainland books applies there. The reader who holds both a Mainland and a Zanzibar interest should keep the two sets of approvals on separate pages.

Each book is built around one-page tools. The kit for each book holds about ten tools, each on one A4 page, and each chapter ends with the tool to use that evening or that week. The work that matters is the dated record of what was sold, what was paid, what was filed and what was confirmed with the issuer.

Sell and Keep is a four book series for owners, shop managers and sales teams in Mainland Tanzania whose income depends on winning a buyer, keeping that buyer and gaining the people who send new work. The four books cover the copy that makes a buyer act, the phone shop that takes the order and the payment, the routines that bring a customer back, and the network that brings new names. Each book is a handbook in its own right, but the four follow one sequence, and this summary explains where to begin. The series applies Mainland Tanzania law only and states no Zanzibar rule. Its examples are illustrative sellers whose figures are worked from stated inputs, with positions stated at 10 October 2026. Nothing here is legal, tax, accounting or investment advice, and every rule, threshold and fee should be confirmed with the issuer named beside it before anyone acts.

What the series covers and where to begin

The four books are Sales Copy That Sells in Swahili and English (Book 1 of 4), Sell on WhatsApp, Instagram and TikTok (Book 2 of 4), Keep Every Customer (Book 3 of 4) and Build a Network That Brings Business (Book 4 of 4). Each can be read alone, but they follow a chain. The copy in Book 1 earns the reply, Book 2 turns the reply into a paid order, Book 3 brings the buyer back for a second and third purchase, and Book 4 supplies the people who send buyers in the first place.

A reader whose adverts, posts or proposals draw attention but few orders should begin with Book 1, because the other books assume its five step method: know the buyer, lead with the problem, prove the claim, state the offer and ask for the order. A reader who sells from a phone, a market stall or a small shop without a working catalogue or payment routine should go to Book 2 next. A clinic, shop or supplier that spends to win customers and then loses them should read Book 3. An owner whose growth depends on introductions reads Book 4 last, since its asks and follow ups rest on the trust built through the earlier books.

The law in this series is Mainland Tanzania law only. Each book has a local appendix that lists the rules bearing on its subject, states the date each rule was read and marks which rules were read only second hand, in a summary, by heading, or not at all. The series says nothing about Zanzibar, whose rules differ from those of the Mainland and are outside these four books.

Start with the buyer and the problem

Book 1 begins with the buyer, because an advert written for everyone is read by almost no one. The worked example, stated at 10 October 2026, is a lunch caterer in Mwanza whose ledger of 40 orders shows that events were 25% of orders but 61% of revenue, while single boxes were 20% of orders and 3% of revenue. Five calls of fifteen minutes each, 75 minutes in all, supply the buyer's own words, and the book argues that the research should come before the writing. Naming the buyer in the advert lifted the caterer's weekly count from 10 to 12 orders, which the book puts at TZS 48,000 a week.

The first line of an advert then states a problem the buyer already has, in the buyer's own words, and leaves the seller's name for later. The late office lunch was the largest reason for orders in the caterer's ledger, at 14 of 40 orders or 35%. Moving it to the first line raised a week from 12 to 15 orders, from TZS 288,000 to TZS 360,000.

Proof comes next. The caterer kept a receipt book, a delivery log and seven saved messages of thanks. The log showed 46 of 50 deliveries on time, or 92%, so the advert may say '46 of 50' or '9 in 10' but may never say 'always'. Testimonials need the customer's permission, and social proof is limited to counted facts about buyers like the reader. A seller with no proof is told to begin counting that day.

The offer is one sentence holding the price, what comes with it, the limit and the deadline. The book shows that a 5% discount on ten bags takes TZS 8,750 of a TZS 12,500 margin, or 70%, while free delivery within 10 km takes 48% of the margin. A cut off hour and a bundle cost nothing, so they come first. A guarantee is printed only if it will be honoured. Stating the offer lifted the caterer's week from 18 to 22 orders, worth TZS 96,000 more.

The ask closes the advert. It names one action, in one channel, in two steps or fewer, written as a verb, an object and a place. At an assumed 8 in 10 kept at each step, 1,000 interested buyers become 261 after six steps and 640 after two, so every extra step has a cost that the seller can count. A single word to send makes the advert countable, and the caterer's ladder rose from 10 orders a week to 26, a margin gain the book puts at TZS 115,200 a week.

Headlines, product pages, messages and scripts

Headlines are chosen by counting taps, not by taste. A shop tested five headlines of five types, each shown to 1,000 views, and they drew 8, 14, 11, 23 and 17 taps. The two strongest drew 40 of the 73 taps, or 54.8%. The Swahili headline is written for Swahili readers and tested on its own count, not translated from the English one.

Product pages close when each screen answers the next question a buyer would ask. The shop's weak page turned 1,000 visits into 20 enquiries, 8 orders and TZS 3,040,000, at a close rate of 40% and TZS 380,000 an order. The rewritten page turned the same visits into 50 enquiries, 20 orders and TZS 7,600,000. The book's test walks the page screen by screen with the buyer's question at the top of each.

Messages get replies when the first line carries the buyer's own question, one price and one proof, and the last line asks for a single word. In the caterer's broadcast, a 96 word message drew 12 replies and 3 orders from 240 contacts. A 38 word rewrite drew 48 replies and 12 orders. Eight saved replies cover the chat from price to thanks, and one follow up after silence carries a new fact rather than a reminder. Rules on unsolicited messages are marked to be confirmed with the issuer.

Radio and voice scripts are measured in words. At an assumed two words a second, a thirty second spot holds 60. The shop's weak script ran to 65 words, or 32 seconds, and lost its last two lines. The rewrite ran to 50 words, or 25 seconds. Forty spots at TZS 25,000 each, TZS 1,000,000 in all, brought 8 calls at TZS 125,000 a call with the weak script and 20 calls at TZS 50,000 with the new one. The station's rates and rules are to be confirmed with the station before recording.

Proposals, claims and the advertising rules

Proposals and price lists win when the first page states the client's problem, the proof, the price and the next step. Two proposals describing the same cleaning work produced different outcomes. From ten proposals of six hours each, wins rose from 2 to 4, and the hours for each win fell from 30 to 15. A price list of three packages brought TZS 16,500,000 a month from 20 clients, which the book puts at TZS 900,000 more than a single price at the middle level.

The advertising lesson sorts every claim into one of three piles: safe as written, qualify first, or remove. The shop's twelve banner claims fell into 5, 4 and 3 respectively. A claim file records each claim, the proof held, the form of the proof and its date. A claim with no proof is removed, and a claim that depends on a condition is qualified in the advert itself.

Only the Fair Competition Act, Cap. 285, was read in its text, and only to section 50. Its headings that bear on adverts are misleading or deceptive conduct (section 15), false or misleading representations about price, sponsorship or approval (section 16), misleading conduct about goods and services (sections 18 and 19), misleading representations about business activities (section 20), bait advertising (section 22) and accepting payment without the means to supply (section 23). The Fair Competition Commission enforces the Act. Penalties and later sections were not read and are to be confirmed with the issuer.

The practical test is stock against demand. A banner printing TZS 17,500 a bag, with 40 bags in stock, when 65 bags are asked for, leaves 25 buyers, or 38% of demand, facing an empty yard. The book values the 25 bags at TZS 437,500 of orders that the advertised price cannot meet. The qualified claim, 'TZS 17,500 a bag while the 40 bags in stock last', protects both the buyer and the shop. The book counts a reprint of three banners at TZS 255,000, or TZS 85,000 a banner, for the printing alone.

Other rules are listed as questions, not answers. The Food, Drugs and Cosmetics Act, 2003 controls the promotion of regulated products, and only its Part VII headings, covering sections 95 to 98, were read. The radio and television content regulations, GN 134 of 2018, were noted by citation. The online content regulations of 2020 were read only in a law firm summary. The Copyright and Neighbouring Rights Act, 1999, the Personal Data Protection Act, 2022 and the consumer rule on unsolicited messages were not read in full. The working habit is to message only people who have asked, and to stop when asked.

Running a phone shop in thirty days

Book 2 is a thirty day programme in fourteen stages for a seller who runs a business from a phone. It starts by reducing the offer. In the three illustrative shops, most price questions in the month before the programme ended with no order. The first stage chooses one product by a four test score, writes one buyer in a single sentence, and asks three past buyers by voice note whether they would buy again. The research on choice supports a narrow start over a wide range.

The catalogue holds seven fields for each item, among them name, price, size or contents, stock, area and delivery, and the same entry serves WhatsApp, Instagram and TikTok. The stock line is the first to fail, so each stock figure carries a date. Photographs must show what a buyer would check in person, and the first price must be set beside its true margin.

Content runs as a week of small jobs on one channel. Each post sits on a rung of a ladder of show, prove, answer and ask, and closes with a question offering a choice of two. The weekly time budgets are 105, 125 and 210 minutes for the three sellers, which the book converts to 7.5, 8.9 and 15.0 hours across thirty days. Six if then rules say what happens when a week slips, so the plan survives a bad week without being abandoned.

Price questions are answered from a reply bank. The book identifies six kinds of question that cover most chats. A first reply runs to four lines in a fixed order: a greeting by name where the name is visible, the thing asked about, the price with one useful fact, and a question the buyer can answer in a word. Messages to people who have not written first need a recorded opt in, and a stop is honoured at once, as the WhatsApp Business Messaging Policy requires on the book's reading of it.

Money that lands and fake payment messages

A seller is paid when the balance in the shop's own account changes. Everything else is a message. The book sets out five routes by which a phone shop is paid, names no provider, and gives a loss case for each. A business number or account in the trade name, kept apart from personal money, gives a clean record. Goods leave only after the money has landed, and two written rules follow: landed is not sent, and cash on delivery carries its own refusals and risks.

Fake payment messages rely on haste and on a source the other party controls. The book gives five checks that defeat them, run in order before any parcel is handed over: check the shop's own balance; check the name and the amount; release nothing before the money lands; give no refund before settlement; and seek a second pair of eyes above a set amount. A drill on day 19 and a written rule on day 20 make the order automatic.

The argument rests on the cost of one case. Six constructed cases total USD 574 in losses, from USD 14 to USD 320, and the largest is a bulk order of ten worth USD 320. The book prints no frequency statistic for fake payments. For a seller turning over USD 1,260 a month, as Seller A does in the worked example, even the smallest case is a visible part of a week's trade.

The wider context is cited from two sources and flagged for checking. The FinScope survey of 2023 reports formal financial inclusion of 76% of adults, up from 65% in 2017, and mobile money use of 72%, up from 60%. The communications regulator's figures, read second hand, report about 87.05 million mobile money accounts and 2.1 billion transactions in the quarter to 30 June 2026. Subscriptions and accounts are not people, so the book uses these figures only to show the scale of the system. Both should be confirmed at source before reuse.

Price, delivery and returns

A price set before delivery and returns leaves less than the seller expects. The book's test starts from the shelf price, counts the margin, then removes delivery and a returns reserve of USD 1 an order. The counted margins of 45.0%, 42.5% and 48.0% fall to 25.0%, 27.5% and 34.0% once those costs are taken out. The fix is a cost stack and a formula. The stack adds goods, packaging, fees, delivery and the reserve. The price is the stack multiplied by 100 and divided by 65, rounded up to the next whole dollar, which keeps at least 35% of the price after those costs.

The three sellers in the worked example, stated at 10 October 2026, have stacks of USD 15, USD 29 and USD 33, which give prices of USD 24, USD 45 and USD 51. Delivery takes 15.0%, 12.5% and 12.0% of the old price. A lost parcel costs USD 14, USD 28 and USD 32, and the reserve covers four, four and thirteen of them. A stress test of 10% fewer orders leaves the first two sellers still ahead of their old net, while the third falls behind it.

Delivery can be carried in three ways: a separate delivery fee, a minimum order, or a price with delivery inside it. The book prefers the last for a small seller. It also counts each dollar on the fee, which costs 21.0%, 10.0% and 7.0% of the net for the three sellers respectively. A pack that survives the first and last ten minutes avoids most damage, and two photographs with a received message settle most complaints.

Complaints and refunds run on a written policy of five lines, and the reply has four parts. The three remedies of refund, exchange and markdown cost the seller different amounts. Seller B's twelve returns cost USD 120 against a reserve of USD 120, and a rule that the buyer pays on fit problems saves USD 60 a month, or 5.0% of the net. The book warns that a policy too strict loses repeat buyers, and one too loose empties the reserve.

Customer lists, tax and data offices

A customer list is a record of permission. The book builds it from three sources: past chats, order notes and new buyers. A yes is written in words and dated, a stop is recorded on the day it comes, and each monthly message carries one thing to do. A message sent to every buyer risks a blocked number that puts the whole list out of reach. In the worked examples, lists of 84, 195 and 720 names carry a reorder profit at 10% of USD 72, USD 304 and USD 1,296.

The month on one page is the document that answers most questions. Seller A's worked month, stated at 10 October 2026, shows 63 orders at USD 20, or USD 1,260 of turnover, less goods and packaging of USD 693, rider fees of USD 189, a returns reserve of USD 63 and a fixed cost of USD 15, leaving a net of USD 300. A tax office, a bank or a licensing office will ask for that page first.

The book sets the licence route in order: the business name, the council licence for the selling address, the sector permit where a product needs one, the tax office, and then data registration. In Tanzania the business registry is BRELA, the tax office is the Tanzania Revenue Authority and the council issues the business licence for the address. The book gives no rate and no threshold, because a printed number is wrong the day the law moves.

A data office asks what is collected, why, who agreed, how long it is kept, who sees it and how a person asks for removal. For Tanzania the book reports that the Personal Data Protection Act, 2022 requires those who collect or process personal data to register with the Personal Data Protection Commission, with a certificate valid for five years. That point rests on a law firm summary dated 16 February 2023 and is to be confirmed with the Commission before anyone relies on it.

The same appendix records a second hand report of a change for small traders: an upper presumptive turnover threshold reported to move from TZS 100 million to TZS 200 million, with a nil band to TZS 4 million, under the Finance Act 2026, enacted on 30 June 2026. The book read it from a secondary summary only. The Tanzania Revenue Authority is to confirm it before any decision, and the book did not read the taxpayer registration, receipt or VAT rules.

Keeping the customer: value, first week and second purchase

Book 3 begins by valuing one customer over five years. A household spending USD 5 a week is worth USD 1,300 in sales if it stays. The worked model, stated at 10 October 2026, takes 24 of 320 regulars lost in a quarter, which annualises to a 30% yearly leaving rate, and the value falls to USD 721 in sales and an expected USD 216 in margin. The figures are assumptions, not findings. Their purpose is to show what a monthly outlay on keeping a customer can safely reach.

The loss is measured before any change is made. Between the April to June quarter and the July to September quarter, the shop lost 168 of its 680 buyers, or 24.7%, and 120 of the 168 were first time buyers. The clinic lost 130 of 520, or 25.0%, and the supplier lost 4 of 44, or 9.1%. The book then calls a sample of lost buyers with a set script and tallies the reasons.

The first week decides whether a first purchase becomes a customer. Four contacts, on day 1, day 7, day 30 and day 90, are written as dated plans. The day 7 call asks one question with no yes in it. At a lift of ten points in return rates, the calls return USD 36, USD 48 and USD 540 of margin an hour for the three cases, and at half that lift USD 18, USD 24 and nothing. The book treats these hours as the cheapest a business spends on keeping customers.

The second purchase is the point at which most first time buyers are lost. The shop's path to a second order has five steps, the clinic's six and the supplier's seven. Each can be cut to two by taking over the steps that need a memory or a search, at a cost of USD 50, USD 50 and USD 100. The first thirty days carry 62.5%, 60.0% and 50.0% of the second purchases in the models, so the cut pays most when it is made early.

Promises, complaints, reminders and the customer card

A buyer pays for a promise with a plan for the evening, so a promise broken is remembered at the second order. In the three cases the kept rates were 70.0%, 60.0% and 75.0%, against promises of 30 minutes, 20 minutes past the booked time and 8 a.m. Lengthening each promise to 45 minutes, 30 minutes and 10 a.m. raised the kept rate to 95.0%, 90.0% and 95.0%. The Fair Competition Act limits misleading claims, so a promise should be one the business can keep.

A complaint that waits is where buyers go. The book's rule of one day sets three marks: receive and tell within the hour; decide and reply within four hours; close and count by the end of the day. At assumed leave rates of 20% for a complaint closed in a day and 40% for one left to wait, the shop loses 5 buyers a month instead of 7, the clinic 4 instead of 5 and the supplier 2 a quarter instead of 3. The rule is worth USD 1,872, USD 1,536 and USD 4,320 of margin a year on those assumptions.

A reminder sent two days before a buyer runs out brings back orders that would otherwise lapse. Under assumed lapse rates, 120 households on a 28 day bundle, 160 on a six month follow up and 40 accounts on a monthly order keep 156, 48 and 24 more orders a year, worth USD 936, USD 768 and USD 2,160 of margin. A reminder goes only to a buyer who agreed to receive it, and it stops when asked.

The customer card holds five fields: the name as the buyer uses it, the usual order, the rhythm, the channel, the yes, and one occasion the buyer has mentioned. It never holds health, belief, money trouble, family affairs, identity numbers, opinions or guesses. Building one card for each regular costs 10.7, 12.0 and 13.3 hours, or USD 64, USD 180 and USD 267 at the model's hour values. The Personal Data Protection Act is cited as requiring collectors of personal data to register, to be confirmed with the Commission.

A loyalty offer is paid to everyone who qualifies, so it is judged by the margin it costs rather than the sales it touches. The book computes a break even gain for each offer, which is 6.4 points for the shop's USD 1 card, 19.2 points for its USD 3 card, 6.3 points for the clinic's perk and 11.1 points for the supplier's rebate. Because the loyalty offer alone can lose cash, it comes last in the sequence of plays.

The monthly review, the win back sweep and the four plays

A happy customer asked at the right moment is the cheapest source of a new customer. The book values a referred customer at USD 4 at the shop, nothing in cash at the clinic, and USD 120 at the supplier, against USD 13, USD 25 and USD 600 by advertising. At the shop, 100 asks gave 40 names and 16 customers, so the referrals cost USD 64 against USD 208 by advertising. The clinic pays no reward, and the book leaves any reward to the body that regulates the practitioners.

A win back sweep asks lost regulars whether they would return and what would make it easy, and it opens with a question rather than a discount. The shop contacts 40 and wins 8, the clinic contacts 24 and wins 6, and the supplier visits 6 and wins 2, for nets of USD 588, USD 744 and USD 1,740 on the book's assumptions.

The four plays, run for a year, keep up to USD 4,872 of margin at the shop, USD 4,352 at the clinic and USD 10,800 at the supplier, in 90.7, 74.3 and 78.3 hours of work. These are upper bounds, because the gains overlap, and if half of each play works the figures fall to USD 2,436, USD 1,408 and USD 2,160. The one day rule is the largest play in each case, the calls come next, and the loyalty offer comes last.

Networks: weak ties and the twenty people who open doors

Book 4 starts from a finding about weak ties. Work reaches a person more often through someone met occasionally or rarely than through someone met daily. The book cites Granovetter's 1973 study of 54 job changers in a Boston suburb, in which 83.3% heard of their work through a contact seen occasionally or rarely, and later experiments on more than 20 million people, which found moderately weak ties most useful. The lesson is not that close friends have no value, but that a network made only of friends adds little that is new.

The plan runs for ninety days, with five useful messages a week over thirteen weeks, or 65 messages in all, of which 30 are useful gives. The three worked cases are a hire driver, a nurse and a business owner, each on the same plan. In the model they win USD 1,560, USD 1,800 and USD 7,200 a year, and without a win they are short USD 134, USD 234 and USD 2,043 on the cost of the ninety days.

The first step is a map of twenty people who can open doors in one line of work: buyers, bankers and funders, officials and peers. The map fits on one page of four boxes and takes 40 minutes in pencil, with roles written where names are not yet known. Bankers, funders and officials are sources of information and are never offered an advantage. Names that have gone quiet are marked for a single wake up message.

Events are worth attending when they bring people from an empty box on the map. Each room is left with four names and a deadline. In the plan, five events yield 20 names, and each event carries a 48 hour deadline written beforehand. The five events cost the driver USD 52.5, the nurse USD 90 and the owner USD 737.5 in cash and time, and one assumed win repays them 14.9, 20.0 and 4.9 times. More events do not mean more work: the book notes that ten events would give 40 names against 24 follow up slots.

Giving, asking and following up within 48 hours

A give is a small, useful act that takes under ten minutes on average, such as an introduction, a useful piece of information or an offer of help. The plan holds thirty gives across thirteen weeks, costing 297 minutes, or about five hours. A give is never made to a public official as a reward for, or in the hope of, a decision. A giver who repeats the large gives every week would spend 2,340 minutes, or 7.9 times the plan, and the book treats that as a warning about scale.

A request is made so that a yes and a no are both easy to give. The book sets out three sentences: who is asking and what for; one thing, with a size and a time; and a way out. The plan climbs a ladder of eleven asks, six small in weeks 3 to 8, four medium in weeks 9 to 12 and one large in week 13. The model expects 7 of the 11 to be granted for the driver and the nurse, and 8 for the owner. An official is asked for information only, and a no is recorded as information.

Follow up is governed by a 48 hour clock that starts when the event ends or the yes is given. The book cites research finding that a plan written as 'if X, then Y' is more likely to be carried out than a wish. In the model, twenty names followed up inside the clock bring 12 replies, against 4 when none is followed up in time, and the rates are assumptions. The first line of a message carries the person's own words and one useful thing. A price list, a broadcast, a favour or a gift never goes in a follow up.

The contact list is held in three tiers: 20 key names touched six times a year, a warm tier touched twice and a wide tier touched once. Three lists of 100, 110 and 120 names all reach 260 touches over the year, or about five a week. The sheet records the reason for each contact, the last touch, the next touch and the consent given. A phone stores numbers, but a network needs the reason.

An introduction is a small loan of trust, and the rule is to report its result to the giver within seven days, naming what came of it. A budget of yes sets the hours a person gives to others before requests arrive: 3, 4 and 8 hours a month for the three cases, covering 6, 8 and 12 requests of 30 to 40 minutes each. The book sets out five patterns for a kind no, and each is written in advance so that it can be used without strain.

Integrity: faith networks, associations and the corruption law

Church, mosque, alumni and club networks are used with integrity when the member gives first, sells nothing inside the place of worship, and speaks of work only. The book sets four rules that answer the usual complaints, and they apply to church and mosque alike. Alumni circles hold the oldest dormant ties, and waking them costs 33, 39 and 27 minutes in all across the three cases.

The book describes the Tanzania Chamber of Commerce, Industry and Agriculture as a member based, non profit chamber offering fairs, training and advocacy. Its fees are not printed and are to be confirmed directly with the chamber. An association is registered under the Societies Act, which the book lists but did not open, and the Cap number of that Act is to be confirmed.

The Prevention and Combating of Corruption Act is the edge the book treats most carefully. The book's page summary defines an advantage to include a gift, loan, fee, reward or favour, and treats an offer of one to a public official to influence a contract as an offence. It also states a presumption where an advantage passes from a person seeking a public contract. The sections relied on, 3, 15, 16 and 35, are to be confirmed against the text of the Act with the Bureau before any reader relies on them.

Messages to contacts are governed by the Electronic and Postal Communications (Value Added Services) Regulations, 2018, GN 17 of 2 February 2018. The book reads them as setting conditions on unsolicited SMS marketing and telemarketing, and as allowing a consumer to opt out by message, under regulations 9(3) and 7(3). Whether those rules reach a person's own messages to a known contact is left open and is to be put to the Tanzania Communications Regulatory Authority.

Figures, dates and what to confirm before acting

All four books state their positions at 10 October 2026. Most local rows were read between 10 and 12 October 2026, and one telecom statistic in Book 4 was taken from a secondary report dated 13 August 2026.

The figures most likely to move are the reported presumptive tax threshold, the data registration certificate, the mobile money counts and the sections of the Corruption Act. Each sits on a secondary source. Each should be confirmed with the issuer named beside it before anyone acts, and the fee or threshold should be read again at the date of use.

The citations also need a check. Book 1 cites the Fair Competition Act as Cap. 285. Book 3 cites it as Cap 285, Act No. 8 of 2003, and in its text as the Fair Competition Act, 2003. Book 1 read the Act only to section 50, and no penalty is printed in any book. The Fair Competition Commission should confirm which citation applies before a reader quotes it in a contract, a complaint or a notice.

Every worked example is illustrative. The counts of buyers, leave rates, reply rates and win rates are assumptions, to be replaced by the reader's own records before any decision. A reader who meets a figure should ask three things: which seller or case it belongs to, the date it was stated, and whether the rule is Mainland law. None of the four books is legal, tax, accounting or investment advice.

Deals and Partners is the shelf for a reader who has decided to trade with, invest in, or sell to another party. It holds five handbooks, numbered 88 to 92, and each answers one deal question: whether to back someone else's business, how to prepare for a negotiation, how to structure a joint venture or offload an idle asset, how to watch a market before committing money, and how to visit a site, supplier or customer before deciding. The five books share one method. They price a deal from checked numbers, write the terms before any money moves, and test every claim against what can be seen. All five are first editions stated at 10 October 2026. Where a statute is cited, it is Mainland Tanzanian law, and none of the books carries a Zanzibar chapter. Every case is illustrative, and every rate, threshold and date should be confirmed with the issuer named beside it before anyone acts. This summary explains what each book does, a sensible order of reading, and the figures and legal points most likely to change.

What the series covers

The series has five titles, each a first edition and each marked as Book 1 to 5 of Deals and Partners. Investing in Other People's Businesses (Handbook 88) tests a stake in a small company before the money moves. Enter the Deal Room Like a Spy (Handbook 89) prepares the reader for a negotiation over land, supply, partnership or sale. JV Agreements and Asset Offloading (Handbook 90) turns an idle asset into a venture, a sale, a lease or a swap. Watch, Observe, Spot (Handbook 91) sets out the fieldwork done before opening, expanding or buying. Pay a Visit (Handbook 92) covers the site, supplier and customer visits that come before a decision.

The five books rest on the same commitments that appear in the series opening pages: one decision per book, every figure computed from the inputs printed beside it, every rule carrying an issuer and a date, illustrative cases labelled as such, and a kit of tools to fill in at a desk. The tools may be copied into a reader's own notebook and used inside the reader's own business with the source named, but they may not be sold, bundled or re-branded.

Case amounts are in US dollars unless marked otherwise. Where the books convert to shillings, they use TZS 2,600 per dollar, a rate the Kiswahili edition states as its own convention. A reader should not treat that rate as current and should redo the conversion on the day of any decision.

None of the five books has Zanzibar content. Where they cite a statute, it is a Mainland statute. A reader with a Zanzibar deal should consult the current Zanzibar authority and should not assume that Mainland thresholds, fees or registration routes apply.

Choosing the order of reading

The catalogue numbers run from 88 to 92, but the order in which a reader meets the decisions is usually different. A reader who has been offered a stake by a friend should start with Handbook 88, because its first-day screen takes one day and stops many deals before any money moves. A reader who is about to open, expand or buy a site should start with Handbook 91 and then Handbook 92, because the market and the place have to be watched and seen before any figure is trusted.

Handbook 89 belongs at the point where a negotiation is actually scheduled. Its preparation file is one page completed in seven timed blocks, and it assumes that the facts it asks for have been gathered in advance. Handbook 90 belongs where an asset is idle and a partner or buyer is already at the table, because its checklist assumes that the values, the route and the tax steps have been worked out.

A reader who wants the whole cycle can take the books in the order of the decision: watch (91), see (92), check a stake (88), negotiate (89), and structure or offload (90). This order is a suggestion of this summary and is not stated by the books themselves.

Investing in other people's businesses: screen before paying

Handbook 88 is written for savers, salaried professionals and owners who have capital but no time to run a second business. Its central case is a bakery in which ten savers each put USD 300 into 20% of a business priced at USD 12,000 before the money, which is 2.5 times a normalised profit of USD 4,800. The pool bought an oven, a bike and a cold room. Four dividends of USD 3,360 and a buy-back of USD 4,800 returned USD 8,160, a gain of USD 5,160. The same agreement returned USD 960 if sales halved and no buyer came, a loss of USD 2,040. The book keeps the loss beside the gain.

The first-day screen is twelve questions scored from 0 to 2, run in eight hours with a clock, a sheet and the owner's papers, giving a score out of 24. A zero on three named questions stops the screen whatever the total: bank statements that do not match sales, debts missing from the accounts, and no written way out. The bakery scored 21 and went on to full checks. The packaging workshop scored 19 and proceeded only with changes. A haulage company scored 9 with two stoppers at zero, and the stake of USD 10,000 was later worth USD 700, a loss of USD 9,300 or 93.0%.

The book's sharpest lesson lies in the gap between booked profit and operating cash. The haulage company's books showed a profit of USD 18,000, while the bank showed operating cash of USD 5,000, a difference of USD 13,000. Unpaid sales of USD 24,000, less unpaid costs of USD 11,000, explain the gap exactly. A bank loan of USD 20,000 sat outside the accounts altogether. The bridge is built from the statements before any valuation is discussed, and the first-day screen itself leaves this line-by-line reading to Chapter 6.

A stake is priced as a range. It rests on proven profit, net assets after debts, and a discount for a holding that nobody can sell quickly. The packaging workshop asked USD 45,000 before the money for 40%, against an investment of USD 30,000, which is 3.75 times a profit of USD 12,000. The book's grid at 3.0, 3.75 and 4.5 times, with a discount of 20% that the book itself assumes, gives a range of USD 11,520 to USD 34,560 for the 40%. Eight of the nine cells fall below the USD 30,000 paid.

Equity, a loan and a revenue share answer different questions. Equity is a claim on what is left, a loan a claim on a date, and a revenue share a claim on sales up to a cap. On the workshop's USD 30,000, the base case returned USD 35,880 on shares, USD 39,000 on a loan at 15%, and USD 39,000 on a revenue share of 8% up to a cap. The loan beat the base-case shares by USD 3,120, while the shares beat the loan by USD 13,200 in the upside case. Each structure answers four questions: what is owed, when, up to what limit and secured on what.

The shareholder agreement and the exit

A shareholder agreement is the only place where price, use of money and exit are written as obligations. Handbook 88 sets out sixteen clauses in six groups: money, control, information, transfer, exit and disputes. For a minority holder of 20%, the clauses that matter most are the dividend policy, the list of reserved matters that need the investor's consent, a board seat or observer, a key-person clause, information rights, pre-emption, tag-along and drag-along, a buy-back or put, and a route for deadlock and disputes. The Companies Act, 2023 supplies the floor, and the agreement adds the walls.

In the bakery, the dividend policy paid USD 3,360 over four years, a tag-along paid USD 4,800 on a whole sale, and a buy-back at 2.5 times the last profit paid USD 4,800. Had sales halved, the buy-back would have paid USD 1,200, and with no buyer it would have paid nothing. The book treats the notice period, the formula and the payment date as the features that make an exit usable, and each is written before the first payment.

The book sets out five doors out of a stake: an owner buy-back, a sale of the whole company with tag-along, a stake sold alone, dividends only, and no exit at all. In the bakery, the buy-back paid USD 4,800, a stake sold alone USD 3,840, and dividends alone USD 960 a year. Selling alone therefore cost USD 960 against the buy-back. Each door should carry a price fixed before the money moves.

Monitoring is kept light. The investor reads five numbers four times a year, matched to the bank statements: sales, gross margin or ingredient share, cash in months of cost, receivable days, and dividends declared against dividends paid. The book's tripwire is two months of cost, set at USD 13,400 for the bakery and USD 21,000 for the haulage company as an assumption. The haulage company's cash cover fell from 1.3 months to zero across six quarter ends, below the tripwire throughout, and paid none of the USD 2,250 a year pitched.

Deal room: prepare like a spy

Handbook 89 is written for owners and managers who negotiate land, supply, partnership and sale deals. Its method is preparation before the meeting. Its first case concerns a hair salon operator whose rent negotiation shows the difference. A meeting entered without preparation was offered USD 330 and settled at USD 345 a month, which over 24 months came to USD 8,280. After one hour of research, the same operator settled at USD 312 a month, which came to USD 7,488, a gap of USD 792.

The preparation file is one page, completed in seven timed blocks totalling sixty minutes: the other side's pressures, their alternatives, the reader's own alternative and walk-away point, the opening, target and ask, the questions, the concessions, and the closing line with its record. The walk-away point is the best alternative costed to the dollar and spread over the term. In the lease, moving would have cost USD 285 a month for 24 months, plus USD 450 to move and USD 350 of closed trade, which is USD 7,640 over the term, or USD 318 a month. The opening of USD 270 is said aloud with its reason. The walk-away figure is never said aloud.

Signals are reasons to ask a further question, never verdicts on a person. The book scores each side's pressure from 0 to 3 across time, money, alternatives, authority and information, with evidence beside each score. In the lease, the operator scored 5 and the landlord 6. Four empty months at USD 312 would cost USD 1,248, while the rise of USD 48 a month earns USD 576 a year, so 1.8 empty months cancel the rise. Facts are marked as told or seen, and other people's messages, accounts and phones are never touched.

Questions and silence do most of the work in the room. The book's questions come in four kinds, asked in order: what the other side needs, what it can move, and what would stop a signature. A silence of five seconds, planned at three marked points, earns more than a sentence at the same points. Concessions come in shrinking steps and are never given alone, each paid for by something asked in return. In the lease the tenant stepped 15, 12, 9 and 6, and the landlord 15, 15, 10 and 8, and the last gap of 14 closed with a trade of a 24-month term for a repaint at a fixed rent.

The book ends its method with a written echo: a message of about six lines, roughly 110 words, sent within a day, which records the terms agreed. In the lease, the landlord's draft read 'when the landlord's budget allows', where the tenant remembered 'before the term starts', and a repaint worth USD 240 to the tenant hung on that difference. Two cautions attach to the echo. The book says that no rule on recording a conversation was read, and that the weight of electronic records in Tanzania is unsettled. A reader should take advice on both before relying on them in a dispute.

Five scenes in the deal room

The book tests its method in five scenes. In the first, a land seller in a hurry moved a USD 9,000 ask to USD 7,900 once the buyer used three comparables and a costed alternative of USD 8,550. The seller's school fees of USD 2,400 and a rival offer, both told when asked, led to a payment of USD 4,740 on signing and USD 3,160 after a clean official search and transfer. The zone of USD 1,550 was split USD 650 to the buyer and USD 900 to the seller, which a plain split of the two first offers would have cost USD 600 more.

In the second scene, a stated reason for a price rise was tested. A supplier justified a rise of 8% by transport at 10% of the price, rising 20%, which the book puts at USD 80 a month, or 2.0%. A package of USD 4,100 a month for twelve months, with payment in 14 days instead of 30, saved USD 2,640 against the ask, cost USD 394 of carrying, and netted USD 2,246 a year.

The third scene separates money, profit and control, which a partner often asks for at once. A partner with 60% of the capital offered 30% of profit and sole signing. The book's answer gives a 40% share, which pays USD 9,600, with six reserved matters and a signing limit of USD 1,500. The fourth scene allows a large customer to be paid in two parts, USD 6,000 in 12 days and USD 6,000 in 28 days more, with deposits on each order of USD 9,000. The delay cost USD 560 in all, 4.7% of the invoice. A late charge of 1.5% a month is marked as unconfirmed.

The fifth scene is the sale of the owner's own business, with sales of USD 600,000 and a normalised profit of USD 90,000. A first offer of USD 216,000, which is 2.4 times profit, met an ask of USD 324,000 and a walk-away figure of USD 270,000. The price agreed was USD 288,000, or 3.2 times profit, paid 70% at close and 30% over twelve months against a sales base. If both largest customers left, the owner would receive USD 266,400, which is USD 3,600 under the floor. The payment structure is tested before signing.

Joint ventures and idle assets

Handbook 90 is written for owners holding land, vehicles, machines or licences they cannot fully use, and for partners planning a venture. Its first step is a four-column list of each item, its value, its earnings last year and its cost to keep, from which a yield is worked out. In the book's example, a motorbike recorded at TZS 7,800,000 costs USD 600 a year to leave idle, made up of USD 150 of insurance and licence and USD 450 of assumed depreciation. An item that earned nothing is idle.

Four questions choose the route: whether the asset lacks something it needs, whether sharing and waiting are tolerable, whether cash or a good price is at hand, and whether a swap partner exists. Two yeses point to a venture. Where sharing and waiting are not tolerable, the route is a sale if cash or a good price is at hand, and a lease if not. The motorbike ran to a lease at USD 810 net, or USD 570 with a fifth fewer days, against USD 600 lost while idle. A sale left USD 2,850 and a swap cost USD 500. The plot and the truck fleet pointed to ventures, and the plant pointed to a sale.

The form of the venture matters. A contractual venture creates no new person and is quick to draft, but one party acts for both. A company venture creates a separate person that can hold land and sign with a bank, at the cost of registration and duty. Moving the USD 300,000 plot into a company costs TZS 9,749,500, or USD 3,750, on the book's figures, while leasing it costs USD 120. A venture is a set of rules, and the form is the vessel.

Contributions are valued before the shares are set. Cash counts at face value, an asset by comparable sales or three quotes, skill by a fee and a share rather than as capital, and contracts by what they deliver. An owner who believed a truck was worth USD 48,000, against quotes of USD 38,000, USD 40,000 and USD 42,000, overstated the stake by USD 24,000, or 20.0%. A holder of 27.3% loses every simple vote, so protection sits in twelve reserved matters, a board seat, two signatures on the account and milestone releases of the land.

Three asset deals and the title checklist

The three worked deals each carry a loss case. In Deal 1, a plot of USD 300,000 sold nets USD 275,500 after selling costs of USD 9,000 and a 10% tax on a gain of USD 155,000, which is USD 15,500. In the base case, a development venture returned USD 348,218 on equity shares or USD 319,200 on a revenue share. In the loss case, with house prices down 20% and costs up 15%, equity returned USD 251,174 and the revenue share USD 255,360, which is USD 24,326 and USD 20,140 short of the sale.

Deal 2 is a fleet of three trucks valued at USD 120,000. At 40% utilisation it earns USD 1,200, or 1.0%. A lease pays USD 18,000, a sale leaves USD 115,200, and a partnership pays USD 21,000 at 75% utilisation, after an operator's fee of 5% of gross and a 70/30 split of what remains. The partnership beats the lease only above about 71% utilisation. At 55%, the owner receives USD 7,560, a shortfall of USD 10,440 against the lease.

Deal 3 is a plant with a reference value of USD 55,000. A cash offer of USD 48,000 compares with a staged offer of USD 66,000, paid in three instalments that are worth USD 60,764 today at an assumed 10% discount rate, or 26.6% more than the cash offer. The staged offer carries a risk: if the buyer stopped after the first payment with no security, the owner would hold USD 26,400. The book's proposed protection is retention of title with a 60-day trigger.

Title moves only after value, tax, consent, instrument and stamp are taken in that order, then registration or notice. The book flags that the Land Act, Cap. 113, as listed in a secondary source, makes a disposition void where the consents under sections 36 to 40 are not obtained, and it asks the reader to read the text directly. A vehicle sale is notified to the Tanzania Revenue Authority. The merger threshold of TZS 800,000,000, or USD 307,692, under the Fair Competition order GN 17 of 2007, is to be asked about, not assumed. Stamp duty on a conveyance is shown as 0.5% of the first TZS 100,000 plus 1% of the rest, per Handbook 06 with fees stated at 26 September 2026.

Watch before you spend

Handbook 91 is written for owners about to open, expand, buy or invest, and for investors who check a market before committing money. Its premise is that a planned period of watching costs less than a wrong move. Its first case is a cook weighing a USD 300 stall deposit in Dar es Salaam. A seller's claim of brisk trade was tested with counts at fixed hours. The count showed 2,457 passers a day, against a claim of 3,600, and 15 buyers a day against 19 needed. The watcher walked away, spending USD 57 of watching to keep USD 426 out of risk.

The method separates what is seen, counted, heard and guessed. The notebook page has three columns, Seen, Counted and Guessed, and a guess is never added to a count. A single day's count at a lane could have reported anything from 920 to 3,280 people, against a week's counted day of 2,457. The chapters then cover footfall and buyers counted at fixed hours, access, parking, security and neighbours, listening to traders and drivers with weights from 1 to 3, and weekly price tracking of eight items at three rivals over four weeks.

Timing matters as much as place. A shop in Mwanza tracked the customers entering its nearest rival between 17:00 and 18:00 for 28 days, and the weekly totals were 410, 385, 520 and 690. The pay week ran 73.6% above the mean of the first two weeks and held 34.4% of the month. For a shop needing 40 customers a day, the thin weeks fell short, and the pay week needed about USD 236 of extra stock at cost. The watcher builds a 28-day calendar of pay weeks, market days and weak months before any lease is signed.

A gap is a question, not an answer. Four tests ask whether there is demand, whether buyers can pay, whether they can reach the door, and whether the gap has been proved. In Mwanza, 63 people an hour passed the evening with no shop open, but the evening hour would have had to supply 10 of the 40 customers needed, or 15.9% of the passers. Eight criteria, each scored from 0 to 2, give a total out of 16. A total of 12 to 16 moves now, 8 to 11 is amber, and a lower total means wait or walk away.

The fourteen-day watch and the three cities

The fourteen-day plan turns the sheets into one verdict. In the book's model it runs from Monday 12 October 2026 to Sunday 25 October 2026, and holds 38 hours, 20 in the Watch Week and 18 in the second week. Each day has an hour, a sheet and a figure to write down. The ceiling and the verdict day are written on day 1, and a day that slips moves to the next open day. Two lost days would make a plan of 16 days.

The same sheets produced three verdicts in three cities. A stall in Dar es Salaam was walked away from. A shop in Mwanza scored 13 and moved on, with a what-if that a 3.0% price cut would lift its daily need to 47. In Arusha, a worst month at an index of 55 cut Street X to 21 customers per cafe, which was 16 below the need of 37, while Street Y stayed at 39, 2 above it. One street survived the worst month and the other did not.

The book uses the cafe street to show a closure count. On Street X the cafe owner counted 42 fronts, of which 6 had closed and 5 changed hands, a closed share of 14.3% and a churn of 26.2%. On Street Y, 2 of 38 fronts closed. Three of the six closures on Street X were rent raised at renewal, and two were no parking. The street evidence comes before the books of any business on that street, and Handbook 88 applies once the stake is offered.

See it with your own eyes

Handbook 92 argues that reports and calls hide what a visit reveals. A trader's catalogue said 60 drums, but the yard held 12, which is 80.0% fewer, and the advance of USD 900 would have been twice the value of the stock on the ground. The book frames the discipline with the Toyota practice of genchi genbutsu, going to the real place to see the real thing, and it shows that a small buyer can use the practice without a factory.

A visit needs a plan before it needs a gate. The plan names the purpose, the questions and the people to meet, because a visit that begins without a purpose ends with a tour. The log has four columns, seen, counted, heard and photographed, and each sorts a note of different weight. A count carries its unit and method and is made by the visitor, while a photograph is a claim with a caption. Consent is recorded on the page, and each visit is kept in one folder of five items.

The book's cases show the method at work. An office worker with a two-hectare vegetable plot had a manager reporting 9,000 kg picked, against 7,380 kg sold according to the slips. The gap of 1,620 kg, or 18.0%, was worth USD 891. A USD 55 visit walked the plot in the order the work moves, counted what left the gate, and flagged four of twenty checklist items. A packaging maker visiting its largest customer found 9.0 weeks of stock on the shelves against a normal 3, with USD 5,400 of orders at risk. A buyer about to pay a USD 950 deposit on a plot bought from photographs found a 3 m strip in dispute, and the price was made conditional at USD 8,075.

Each visit ends in writing. The report runs to three lines within 24 hours: what was seen, what changed in the numbers, and what is decided. The second line carries a figure and the third a verb, an owner and a date, and the report goes to someone who did not make the visit. The clock starts at the gate, so a visit ending at 15:30 on Thursday 22 October 2026 is reported by 15:30 on Friday 23 October 2026.

Unannounced visits and the quarterly calendar

An unannounced visit shows the ordinary day, which a date can change, and it works best as a second reading after an announced visit. Four questions decide whether to use one: whether the question concerns an ordinary day, whether a baseline visit exists, whether the visitor can arrive at the front gate in office hours, and whether a refusal can be accepted calmly.

A quarterly calendar of six visits, chosen by the money at risk and fixed in dates before the quarter begins, keeps every large paper claim under test. The book's model quarter has three supplier visits, two customer visits and one farm or branch visit, two of them unannounced. It costs 36 hours and USD 438 a quarter, which is USD 288 of owner time plus USD 150 of travel. Each visit in the book cost less than 7% of the money it tested.

The five visits told in full show the same method in five places. The Arusha trader counted 12 drums against 60 and kept USD 900 back. The plot visit found the 3 m strip in dispute. The packaging visit counted 249,600 units a month against 450,000 claimed, leaving 59,600 booked against 120,000 needed. The retailer held 9.0 weeks of cover, and the farm sold 7,380 kg against 9,000 kg reported. A reader who weighs the visit's cost against the money at risk will find the case for each visit in these figures.

Figures and laws to confirm

The five books are first editions stated at 10 October 2026. Some figures come from other sources with their own dates. Handbook 90 uses stamp duty and government fees stated at 26 September 2026, drawn from Handbook 06, and cites a government fee web page dated 12 October 2026. A reader should check the current position with the issuer before any payment, because a fee stated in 2026 can be changed by a later order.

The legal points the books flag most often are these: the Companies Act, 2023, which governs the shareholder agreement in Handbook 88; the Land Act, Cap. 113, under which a disposition is void without consents under sections 36 to 40, as read from a secondary listing in Handbook 90; the Arbitration Act, 2020, which commenced on 18 January 2021 and requires an arbitration agreement in writing, as cited in Handbook 90; and the Fair Competition order GN 17 of 2007, with a merger threshold of TZS 800,000,000, to be confirmed with the Fair Competition Commission. Handbook 90 also cites the Companies Act, Cap. 212 R.E. 2023, section 458, for BRELA fees as set out through Handbook 06.

The tax figures are assumptions in the books rather than settled rates. Handbook 90 applies a 10% tax on the gain on the plot and an assumed 30% tax on the gain on the plant, and it cites a Tanzania Revenue Authority capital gains page that is undated. Handbook 90 also applies an assumed discount rate of 10% to the staged offer. Handbook 88 applies a 20% discount to its price grid and an assumed two-month cash tripwire. Each assumption should be replaced with the rate the Tanzania Revenue Authority confirms in writing before any figure enters a budget.

Currency is the other moving part. The books state their cases in US dollars and convert shilling figures at TZS 2,600 per dollar. That rate is a convention of the Kiswahili edition and is not a current rate. The conversion should be redone on the day of a decision, using the bank's or the issuer's rate, and the reader should keep the source currency beside every converted figure.

Using the books with care

The cases are illustrative. Each is built from stated inputs, and its people and businesses are described by type, not by name. Cases that carry dates after the stated date are projected worked examples and not records of events. A number in a case should not be assumed to apply to the reader's own deal, however close the match looks.

The books recommend no lender, bank, product or provider. Their disclaimers state that they give general business and financial education and are not legal, tax, accounting, financial or investment advice. Each book directs the reader to a qualified adviser, an advocate for any dispute or demand, or Zatra for decisions that carry real money or legal risk.

A reader who is about to sign a stake, a joint venture, a sale or a lease should have the agreement reviewed by an advocate, and should confirm each rate, threshold and date with the issuer named beside it. The series also carries one habit through all five books: a decision is written down before the money moves, whether it is the screen result, the walk-away figure, the venture's exit, the watch verdict or the visit report. That habit is the thread of the series and the first thing a reader should take from it.

The Owner's Mind is the eight-book shelf of the Zatra Handbooks on how an owner acts, decides, recovers and keeps going. It does not teach a trade, a tax rule or an investment. Each book takes one belief or habit that shapes results over months and years, and turns it into a routine that can be counted: a daily tick, a weekly decision rule, a scored list, a dated test or a one page sheet. The books share one method. Every money claim is a modelled case with stated assumptions, and every research finding is named with its source, with a note where the source was read only in summary, second hand or not at all. Most of the series is general and applies to any owner. Mainland Tanzania appears through local appendices in several books, and Zanzibar is almost absent. The series is general information, not legal, tax, financial or medical advice. This summary explains what the eight books cover, the method they share, the points that need confirming, and where a reader should begin.

What the series covers

The series has eight titles, read in order on the shelf. Consistency, Not Hard Work (Handbook 93) argues that steady daily routines outlast bursts of effort. Move Quickly and Reign (Handbook 94) covers decisions, launches and recovery from a mistake. Do You See a Tree or Furniture? (Handbook 95) trains the eye to see value in raw material, waste, land and idle time. The Bamboo Tree (Handbook 96) explains why a business grows roots for years before its sales show. The Leopard's Patience (Handbook 97) argues for one target and a patient strike. Winning Spirit, Quitting Spirit (Handbook 98) sets out the habits that carry an owner through a loss. How Games Train Business Skills (Handbook 99) pairs nine pursuits with business weaknesses. Life-Changing Business Stories From 70 Successful Men (Handbook 100) tells seventy business lives and gives one lesson for each.

The series does not teach a trade or a tax rule. Its subject is the owner's conduct: how often to act, how fast to decide, how to see value, how long to wait, how to recover, how to practise a skill and what the lives of others show. Each book gives a tool that can be counted, and each closes with a plan for a fixed period, which runs from ten to fourteen weeks or ninety days, or in one case a root year of 346 days. The plans are designed to produce a record that the reader can check at the end of the period.

The books are best read by problem rather than in number order. A reader who loses deals while waiting for certainty starts with Handbook 94. A reader who chases every opportunity and closes none starts with Handbook 97. A reader facing a loss, a heavy debt or a failed launch starts with Handbook 98. A reader who wants to see how others have started, failed and recovered starts with Handbook 100. A reader who wants a routine before a larger change starts with Handbook 93, which sets up the counting habit that the other books reuse.

The eight books are sold as one shelf and are read in any order. Handbooks 93 and 94 both use a weekly or daily count, and Handbooks 97 and 98 both use a stop number. The shelf is therefore best read as a set of connected tools rather than as a sequence of arguments.

The shared method

Every book prices its claims with a modelled case. Each book builds its cases around a few modelled owners. Handbook 93 uses a taxi driver in Dar es Salaam on USD 420 a month, a teacher in Nairobi on USD 1,150 a month and a workshop owner in Mwanza on USD 3,000 a month, with figures stated at 10 October 2026. The books call these models and label each assumption. They do not claim that the cases are typical, and the reader should treat each gain as a hypothesis to test in the reader's own accounts.

The second feature is the count. A tick, a conversation, a stalk, a test, a mirror statement or a scored question is recorded so that the reader can see what was done, not only what was felt. Several books use a single page tracker, and their reviews are short and fixed in length, often two, twenty or forty five minutes. The book argues that a streak counter misreads a single miss, while a count records the miss and keeps the record honest.

The third feature is the limit set in advance. A test has a cap, a stop number, a date for reading the result and, for certain decisions, a cooling off period. A decision made without these limits is left to a later review, and the books treat land, loans and partners as the decisions that most need such a limit. The limits are written before the first step, so that the reader cannot move them to suit a good or bad week.

The fourth feature is the research. The books cite journal articles, working papers, news releases and industry reports, and they say which sources were read in full, which were read as abstracts, and which were read second hand or not read. Several findings are used only as context for a habit. The reader should not read a study on habit formation or practice as a forecast for the reader's own business, and the series itself says so in most chapters.

The fifth feature is the honest close. Each book ends with a verdict that is one of a small set of answers: continue, change one thing or stop, or push on, pivot or close. The verdict is read from the count and the limit, not from mood. The reader who uses the same method across the eight books will find that the verdicts are comparable from one book to the next.

Routines, decisions and slow growth

Consistency, Not Hard Work argues that four routines add up to more than bursts of effort. The four are a daily conversation count for sales, a ten minute daily cash check, a twenty minute weekly one to one with a person who can change next week's money, and a learning block that grows from ten minutes a day to thirty. In the three modelled cases the routines take 370 to 410 minutes a week, which the book states as 8.6% to 11.5% of time. The learning block is the slowest to pay, adding about 22 hours by day 90, and the skill gain starts only when twenty hours are ticked. The sales count runs at three conversations a day from day 8 and five from day 43.

The book's chaotic-week case is an illness, examinations and a supplier's trip, which together covered 14 of 270 life days, or 5.2%, and accounted for 45.1% of the missed ticks in the sample. The book's answer is a minimum version of each routine for a bad week, which in the cases took nine minutes a day against fifty. The 90 day tracker holds 259 counted ticks, and the three cases ticked 227, 224 and 224 of them. The monthly gains at month three are USD 87, USD 199 and USD 797 in the cases, and the compounding figures rest on a constant daily rate that the book itself calls an assumption.

Move Quickly and Reign treats slow decisions as a cost that can be written down. It sorts each decision into a two-way door, which can be undone within 14 days for 10% or less of a month's profit, and a one-way door. The distinction is widely reported and attributed to a 2015 shareholder letter, and its thresholds are the book's assumptions. Two-way decisions are made at about seventy percent of the facts, a rule the book draws from an illustrative information curve and not from measured data. One-way decisions, which the book names as land, loans and partners, keep the slow lane, and ten kinds of decision always wait for a cooling off date, among them certain loans at high rates and plots that must be paid for in full.

The second half of Move Quickly and Reign turns the decision into tests. A launch is kept within a quarter of a month's profit, is read in three days and is tied to a stop or go number written first. The reading waits until the next day, so the feeling and the number are separated by a night. The mistake routine has four verbs: admit the mistake in one sentence within ten minutes, fix before explaining, tell everyone touched within two hours, and log it the same day. The rhythm of 180 minutes a week buys 25 tests a year, and the book prices the twenty failures it expects in advance. Tanzanian land and microfinance rules are cited only as secondary accounts or unread texts, and each is to be confirmed with the issuer.

The Bamboo Tree scores four roots, skills, systems, customers and reputation, each from 1 to 5 for a total out of 20. It then tests runway two ways. The first divides cash by monthly cost with no sales. The second divides cash by the present monthly shortfall, and the book warns that the second can mislead when the shortfall is growing. House targets of three and six months are set as rules, and a living cost floor sets the smallest monthly sum on which the household and the work can keep going. Side income is priced as the hours needed to close the gap, and the book warns that work which uses the hours of a root can eat it.

The tailor in the book's case scored 4 of 20 at the start and 14 of 20 after twelve quarters, while sales rose 14%, repeat customers rose from 22% to 51% and payment within 14 days rose from 41% to 74%. The kiosk, by contrast, showed no movement in any of the six measures on the Six Measures Card, and its cash ended the seventh quarter with a shortfall of USD 144. The five signs of a stalled root and the five question Patience or Delay Test separate slow growth from a warning sign. The parable that opens the book is treated as a figure to test, and the botanical sources support only a narrow lesson, to build first and rise after.

Seeing value, patience and recovery

Do You See a Tree or Furniture? trains the eye to see value in what is already at hand. Its main device is the value chain, the list of hands between a raw thing and a buyer. One tree earns USD 90 sold as a log, USD 120 as boards or USD 195 as three tables, and each price belongs to a different hand in the chain. A small maker climbs one rung at a time. Five questions scored 0 to 2 sort an idea into look again, test with conditions or test now, and the score is honest only where each point rests on a weighed number, a described buyer, a quoted price or a rule read.

The fifteen sights in the book are costed by hand on eleven lines: volume, yield, price and sales, then variable cost, fixed cost, margin and start cost, then uplift, payback and the uplift with the price down 20%. The sights cover hides, the cashew apple, coconut husk, maize cobs, fish waste, sawdust briquettes, dung, a digester, idle hours and a delivery round. A digester costing USD 700 that saves USD 18 a month pays back in 39 months in the case, and in 54 months if the saving is 30% lower. A review cited in the book found about 31 to 33% of digesters not working in 2016, so the reader is asked to test the saving before building.

The one week test is the book's exit. Four marks are fixed before day one: three buyers asked, two giving a price, 60% or more sold, and the spending kept within a cap. Four marks met is a go, three a change of one thing, and two or fewer a stop. The three cases show the method at work. One sold 62 of 100 pieces for USD 8 and met four marks, with an uplift of USD 135 that held at USD 90. One sold 16 of 40 kilograms against a cap of USD 30, met one mark and stopped, saving the USD 113.60 that a build would have lost. The third met four marks with an uplift of USD 700 that held at USD 140. The book says that a go is not a forecast.

The Leopard's Patience argues for one target rather than many. The leopard is an analogy, and the book draws field figures from research on the animal: prey in a band of 10 to 40 kilograms, with about 25 kilograms preferred, and a success rate near one in 2.7. The book states where the analogy fails, including scarce prey and hunts that fail more often than they succeed. The method runs in five stages: a Target Card with a not now list, a weighted score of candidates with a cut at 39 out of 50, a terrain map of the people who decide, the calendar they work to and the rivals on the ground, a position built in five rungs, and a Friday count of five signals.

The five signals are Reply, Window, Trigger, Funds and Ask. Each is counted only with a date and the buyer's own words, and mood is not a signal. A strike is made inside three limits: a threshold, a fortnight at 70% of the month's hours, and a cap of half the reserve left after three months of fixed costs. The effort curve and the signal probabilities are the book's model assumptions, and the book says that no study supports them directly. The hunt log counts stalks, abandoned stalks, strikes, kills and misses, so that a reader can tell patience from delay. In one modelled quarter, 24 stalks led to 4 kills, a rate of 16.7%, and the strike rate was 44.4%.

Winning Spirit, Quitting Spirit sets learning, adjusting and continuing against blame, freezing and fleeing. The quitting habits have an outward and an inward form, which the book describes as six signs that can be seen in a working week. Its loss page splits each loss into three columns, caused, not caused and unknown, using three questions: was the information available, was a cheaper action open, and was the decision the owner's own. A thirty statement mirror, fifteen for each set, gives a winning score W and a quitting score Q, and the book reads the result as winning, quitting or mixed without saying what to do next.

The first 48 hours have eight jobs: stop new commitments, secure what is left, eat, sleep and walk, write the loss page, list the next fourteen days, tell one person and fix Monday. The recovery plan runs thirteen Mondays from 12 October 2026 to 9 January 2027, with 70 hours in all, and a ladder of ten small wins, each with a proof another person could see. The stop or go test scores seven questions from 0 to 2 and reads the total as close at 0 to 5, pivot at 6 to 10 and push on at 11 to 14. Sunk money is shown beside the test and never scored. The three owners end the ninety days with a push on, a pivot and a close. The book includes a signpost for anyone with thoughts of self-harm or a low state that does not lift, and it names a doctor, a counsellor or a trusted person as the first contact.

Practice, skill and the lives of others

How Games Train Business Skills pairs nine pursuits with nine work weaknesses and a ten week plan. The book opens with the evidence on practice. It cites a review of 88 studies that finds practice explains 26% of the differences in games, 21% in music, 18% in sports, 4% in education and under 1% in professions. On that basis the book proposes a plan of 720 minutes, or 12 hours, as small against the ledgers of skilled players but real enough to change a work habit. The pursuits are chess, football and golf, piano, guitar, sketching, design, cycling, running and the sprint, each with a habit it is meant to train.

Each pursuit has a different mechanism. Chess trains deciding with a one line plan and a count of blunders, where a sample log falls from six blunders a game in week one to two in week ten. Piano trains patient repetition by cutting a sixteen bar piece into four bar chunks and counting only clean repeats. Guitar trains the reading of mistakes with a three part journal: the slip, its cause and a fix small enough to repeat. Sketching trains seeing, design trains solving for one named user and one task, cycling trains small corrections, running trains an even effort and the sprint trains short bursts with rests. The book notes that far transfer from a hobby to work is rare, so each pairing is a claim tested by two counts, the pursuit count and the habit count.

Costs are kept small. Nine of the ten paths cost under USD 10 a month, and golf on a range costs USD 23 a month, the one path above that level. The book also gives a family section for practising beside children, with a fifteen minute meeting that produces a Family Practice Contract in three parts: who and when, safety, and how progress is shown. No money or goods are staked on any game, and an adult is present at every session. The book's athletics examples are used for illustration only, and the pace and time figures in them are the book's own arithmetic from the sources it names.

Life-Changing Business Stories From 70 Successful Men presents seventy business lives in three pages each. They are arranged in fourteen chapters of five lives and grouped under fourteen problems in a Lesson Index, from starts with no one ahead to cash that ends early, change under the owner, dismissal and the slow road back, old trade houses, factories and giants, patience and loss limits, and long bets such as inventions. Each life follows four steps: the start, the mistake that nearly ended the work, the decision that turned it, and one lesson with a seven day act. The collection leans to the Americas, with 41 of the 70 lives, and six African lives.

Three readers carry the book's fourteen weekly acts, which total 700 minutes. They are a minibus driver on USD 300 a month, a nurse on USD 900 a month and a workshop owner with USD 1,800 of profit. The book's assumed monthly gains of USD 30, USD 54 and USD 340 are stated as assumptions, not findings or promises. It recommends no fund, lender or product and makes no buy or sell call. Each life carries its loss case beside the gain, and the lesson is never said to have caused the success. The collection is one of survivors, and each life tells its own failure too.

The edition note is the most important caveat in the series. The credited sources could not be opened for this edition. The seventy lives were therefore written from the widely documented record, and each fact is flagged to be checked against its credited source before any print. Nineteen of the people are living or not confirmed as deceased, and for them only neutral public facts are given. A reader who wants to rely on a particular life should open the credited source first, and should treat any figure in a life as the source's own, in its own currency and year.

Mainland Tanzania and Zanzibar

The series is general, and most of it applies to any owner. Where a book refers to law, it does so in a local appendix for Mainland Tanzania. The rules named include the Land Act, Cap. 113 (2001), section 37(2), the Microfinance Act, 2018, the Employment and Labour Relations Act, 2004, the Personal Data Protection Act, 2022, the Road Traffic Act, Cap. 168 and section 10 of the Law of Contract Act. Several of these are marked as secondary accounts, not read, or to be confirmed with the named issuer. The Mainland appendices should be read as pointers to the authority, not as the authority.

The series contains no Zanzibar rule. Handbook 100 mentions Zanzibar only as context for a trading house based in Zanzibar and Tanzania, and that history is not a Zanzibar rule. A Zanzibar reader should not take any Mainland figure, rule or appendix as applying in Zanzibar without a check with the Zanzibar authority. Any Mainland figure that the series does print is to be confirmed with the named issuer before use.

Several activities in the series are subject to permits or regulators that the series does not state in full. Charcoal, timber and feed are the clearest cases. The book on sight value says that no licence or transit pass for charcoal is stated and asks the reader to confirm with Tanzania Forest Services and the district forest office before preparing, moving or selling charcoal or timber. Feed is regulated, and the book gives no method for it. Food, soap and animal products are also regulated or hazardous, and the book asks the reader to learn the method from the regulator's guidance.

Common habits and where to start

Across the eight books the same habits recur. Count what is done. Write the limit before the test. Set a dated review. Keep a kill number that ends an action. Where a book adds a new tool, it sits on one of these four habits. The reader who adopts only one habit should adopt the count, because each other tool depends on a record of what happened. A feeling about a month is not a record.

A first step can be taken in an afternoon. The reader can list the decision that is waiting now, sort it as one-way or two-way, write its stop number and set a date for reading it. The reader can start the 21 Day Streak in Handbook 93, or write one target on a card as in Handbook 97. Each first step leads to a fixed plan in the relevant book, and the modelled gains in the plan should be kept as assumptions to test against the reader's own figures.

The series is an honest one on its limits. The books say when a claim is modelled, when a source is second hand, when a figure must be confirmed and when a finding is only context. A reader should keep that discipline. Every figure needs its issuer and date, every gain needs a loss case beside it, and every verdict needs a count. Used this way, the eight books give the reader a method for ordinary decisions that is more reliable than mood.

Money and Terms is a shelf of five handbooks for the moments when a money decision turns on its terms: how a late loan is collected, how a project is financed, how a buyer pays, when a deal should be left, and how a rainy season changes the money a business depends on. Each book is written for an adult facing one decision, and each carries a kit of tools to fill in at a desk. The books are general information, not legal, tax, accounting or investment advice. Their worked cases use US dollars, with Tanzanian shilling equivalents converted at a book rate of TZS 2,600 per dollar, a convention of the books rather than a market quote. The legal references are Mainland Tanzanian rules that the books ask the reader to confirm with the issuer. This summary explains what the shelf covers, where to begin, and which points need checking before anyone acts.

What the shelf covers

The shelf has five books. Debt Collection Strategies for MFIs (Handbook 101) is written for lenders, from microfinance owners to SACCOS managers. Project and Business Financing Models (Handbook 102) compares twelve ways to fund a project. The World's Payment Terms (Handbook 103) sets out how a buyer can pay and who carries the risk at each step. Walk Away Early (Handbook 104) is a method for leaving a deal before it costs more. El Niño: Where the Money Moves (Handbook 105) follows the money through a season of heavy rain. Each handbook has between fourteen and sixteen chapters and between nine and ten tools, and each chapter closes with a brief and six key takeaways.

The books read as a set of decisions rather than a set of topics. The lender needs the arrears ladder before it asks for a pledge. The promoter needs to know what each funder costs before the signature. The seller needs the payment term before the first delivery, and the owner needs a rule for when to stop. The weather book cuts across the other four because it changes the cash on which each of them depends.

The books share a method. Every figure is computed from inputs printed beside it, every rule is named with its issuer and a date, and every illustrative case is labelled as composite. People and businesses are described by type rather than by name. Positions are stated at 10 October 2026, and the books say so on their first pages. Where a rule was read only at second hand, or not read at source, the text marks it to be confirmed with the issuer, and this summary keeps those marks.

The shelf does not recommend a lender, a bank, a product or a provider. It names the institutions that issue the rules that matter, such as the Bank of Tanzania, the Tanzania Meteorological Authority, the Public Private Partnership Centre and the International Chamber of Commerce, and leaves the choice of counterparty to the reader.

Where to begin

Start with the decision, not the number. A lender whose arrears rise month on month should open Handbook 101 and fill the ageing sheet before reading anything else. A promoter who has been offered a loan for a project that needs another structure should read Handbook 102 before the funder meeting. A trader about to agree a price with a new buyer should open Handbook 103 before the quote goes out. A reader already inside a deal that does not add up, or with money paid and a doubt growing, should go to Handbook 104. A reader who sells, farms, rents or lends through a wet season should start with the first chapter of Handbook 105 and the sixty day plan.

Some books share ground on purpose. Handbooks 103 and 104 both test a counterparty before money moves, and Handbook 102 supplies the cash flow page that Handbook 105 uses to test a plan against a wet season. The repeated tables are intended.

Each handbook names on its first pages the tool to print first. The method works best when that tool is filled in on a Monday and reviewed four weeks later, because the shape of a decision shows on paper more clearly than in memory.

Collecting late loans without losing the borrower

Debt Collection Strategies for MFIs starts from the claim that late loans have causes and that the cause decides the response. The book names five causes: timing, shock, over-indebtedness, a loan that did not fit the borrower's cash, and unwillingness. In a composite review of 40 late files, 37 could be fixed without force, which is 92.5%. Three files, or 7.5%, showed unwillingness or fraud. The lender's first task is to sort each late loan into a cause before it chooses a step.

The method is a collection ladder of eight rungs. The first two are cheap contact: a reminder three days before the due date and on it, and a call on the first missed day. Rungs three to five are the field visit, the group meeting and the new schedule, each run by a named officer or the branch manager inside a stated window of days. The sixth rung is the taking of a pledge, which the book treats as a last resort under the Microfinance Act, 2018. The seventh is the demand, mediation and court step, and the eighth is the write-off. The book is careful to separate a write-off from forgiveness. A write-off moves a balance to a register and uses a provision already held. It does not cancel the debt, end the lender's conduct rules or return money to the cash box.

The cost of each rung is worked in the book's model. A reminder costs 20 cents of officer time and 2 cents for the message, and a call costs ten minutes of officer time, or USD 1. A field visit costs USD 18 with a break-even balance of USD 45. The court floor is set at USD 420 in the model, derived from a cost of USD 210 divided by a stated 50% chance of recovery. These are the book's inputs and not market rates. A lender should replace them with its own costs before using the floor for a decision.

The branch model shows how the ladder moves the sheet. PAR 30 stands at 14.0% in the worked branch, where USD 168,000 of a USD 1,200,000 portfolio is more than 30 days late. The provision is USD 105,600, and 45.5% of it sits in the over 90 bucket. After 90 days on the ladder, PAR 30 is 5.0%. It is 6.0% with the write-off added back and 7.0% with restructured loans counted. The book's lesson is that the honest report puts all three figures side by side. Four lenders of very different sizes each removed between 59% and 65% of PAR 30 in 90 days, and the cheap rungs carried most of the cure.

The book also covers the people who run the ladder. A card of eight red lines sets out what no officer may do, a complaint is read by someone other than the officer complained about, and the pay scheme ranks officers by how the ladder is run rather than by the loudest recovery. Digital lenders, whose loans of about USD 100 cannot carry a pledge or a court step, are given a shorter ladder of reminders, one call, a plan and a quarterly letter. For a digital lender the one forbidden step is a message to the contact list, which the book says breaches two red lines and puts the licence at risk.

The Mainland rules that bear on the ladder are set out in a local appendix. The Microfinance Act, 2018 is the main source, with regulation 45 of GN 679 of 2019 for the buckets and provisions of Tier 2 lenders, and regulation 44(3) for group loans. The book flags that the fourteen day written notice, whether early calls count as collection, and whether the Land Act adds its own steps were not confirmed at source. The thirteen week turnaround in the book runs from Monday 19 October 2026 to Saturday 16 January 2027, with the count on Monday 18 January 2027.

Matching funding to the project

Project and Business Financing Models starts from a distinction that many borrowers meet only after signing. Money for a project is either a claim on the project's cash, which is debt, or a claim on the owner's share, which is equity. A project with USD 100,000 of cash pays USD 40,000 of costs, then USD 30,000 to the bank, then USD 10,000 to note holders, and the remaining USD 20,000 to shareholders. When cash falls to USD 60,000, the bank is short by USD 10,000 and the others receive nothing. Each of the twelve models sits somewhere on the line between debt and equity, and each has one card built from the same rows.

The twelve models run from bank and MFI debt, through asset leasing, invoice and LPO finance, equity and mezzanine, supplier credit, grants, concessional funds and crowdfunding, to the state and project finance models, with retained profit at the end. Each card asks the same five questions: risk, cost, control, timing and exit. Cost is written three ways, as total dollars, as a share of the sum and as a yearly rate. The book prices waiting at the project's own profit, so that a delay in a grant or a crowd round appears as a cost. Scores from 1 to 5 are Zatra's judgement and show the trade without choosing for the reader.

The worked figures make the contrast concrete. On a bank card for USD 300,000 over 60 months, the payment is USD 7,137 and the total cost is USD 134,719, or 44.9% of the sum. A supplier credit term of 2/10 net 30 is equal to 37.2% a year. A purchase order of USD 300,000 costs USD 15,600, or 30.1% a year. If a project earns USD 78,000 a year, equity of USD 120,000 for a 30% share returns USD 210,600 to the investor, 1.75 times the sum. A concessional loan of USD 150,000 saves USD 102,877 of interest, and is worth USD 49,377 once a seven month delay, 250 hours of work and the reporting are priced. The rates are the book's assumptions, and each must be confirmed with the funder.

The state and project finance models have a fixed entry cost that decides the size at which they pay. On a cold store of USD 300,000, debt of USD 286,857 could be carried on the cash flow at a cover of 1.3, but the gearing limit holds it to USD 210,000 at a cover of 1.78. Arranging that debt costs USD 54,200, or 18.1% of the project. The book states that the review fee for an unsolicited public private proposal is 0.1% of the estimated cost, capped at USD 50,000, under the Public Private Partnership Act, Cap. 103, as amended by Act No. 4 of 2023, and that the threshold must be confirmed with the PPP Centre.

Retained profit has no lender and no interest. Its cost is the return the project would have earned while the money was being gathered, which the book prices at USD 1,200 for the freezer against USD 226 for the loan. The Model Chooser grid matches a project's cash flow to one of six shapes: steady, seasonal, paid late, build first, long contract and benefit without cash flow. Each shape gives a shortlist, and the grid shortens the choice without making it. The three worked projects, a USD 1,000 equipment purchase, a USD 12,000 seasonal venture and a USD 300,000 build, show that the cheapest rate won none of them. A blend of sources is also tested. In the book's example, a bank loan of USD 120,000 at 15%, a lease of USD 80,000 at 14% and the owner's USD 100,000 cost 14.6% on the USD 200,000 borrowed, with a lowest cover of 1.27, which is worse than either lender alone.

The paperwork and the clauses complete the picture. Twelve documents cover what the twelve models ask for, and building one pack costs far less than a fresh folder for each model. The chapter on term sheet traps sets out twelve clauses that move the money, seven of which can be sized in dollars before anything is signed. The examples include a flat 12% that behaves as 21.5%, a guarantee that can call USD 126,000 against a house of USD 90,000, and fees that lift a 15% loan to 15.6%. The reader is told that three of the traps can be negotiated out before the signature, and that a clause with a number attached is the one a funder is most willing to change.

Two reference points are flagged. The Bank of Tanzania overall lending rate was 11.93% in August 2026, as reported in a secondary source, and the book uses it only as context, since an average is not a quote. The crowdfunding caps, reported from the Capital Markets and Securities Authority guidelines of 2023 (GN 732 of 2023), are USD 500 per investor in one issuer, USD 5,000 per investor in 12 months, USD 150,000 per issuer in 12 months and a USD 300,000 platform cap, and they must be confirmed with the Capital Markets and Securities Authority.

Getting paid: payment terms and Incoterms

The World's Payment Terms begins where many small sales go wrong: a price is agreed and the payment is left to chance. The margin on an order is earned only when the money arrives. At a margin of 15%, an order of USD 5,000 earns USD 750, and one unpaid order of that size takes about seven further sales to repair. The book's answer is a quote with a price and a term with a date, and a buyer scored before the first delivery.

The central instrument is the payment ladder, twelve rungs ranked by the seller's exposure at dispatch. Each rung states who carries the risk at each moment, what paperwork is needed, and the cost in shillings. The rungs run from a deposit and cash on delivery through mobile money, bank transfer, 30 and 60 day credit, documentary collection, letters of credit and escrow, to open account with credit insurance or a standby. Buyers accept the safest terms least, so the value of a quote often peaks at a middle rung. The minimum fees of the bank rungs weigh heaviest on small orders. In the book's example, a confirmed letter of credit takes 333.3% of the margin on the smallest seller's parcel and 13.3% on a larger container.

The buyer is scored, not trusted on sight. The Buyer Risk Scorecard asks ten questions, scores each answer 0, 1 or 2, and counts an unknown as 0. The total places the buyer in band A (17 to 20), B (12 to 16), C (6 to 11) or D (0 to 5). A decision tree then reads the band against the order size and names a rung. A buyer who refuses the rung moves to the next riskier band's term, never to a looser one, and the seller may reduce the order instead. In the book's example, a buyer scoring 13 of 20 is offered rung 2 by one seller, rung 7 by another and rung 5 by a third, because each seller's exposure differs.

Each method has its own fixed cost and its own time. A documentary collection sends a set of papers through four hands, the seller, the seller's bank, the buyer's bank and the buyer, and releases them against payment or acceptance. A collection that is never collected cost USD 1,050 in the book's example, or 140.0% of the margin. Escrow costs 1.5% and depends on a written definition of acceptance. A letter of credit makes the papers, not the goods, the thing the bank examines. The seller reads twelve clauses and three dates: presentation within 21 days of shipment, five banking days for the bank to examine, and expiry. Two discrepancies on a credit of USD 50,000 cost USD 4,325, or 57.7% of the margin.

Mobile money is cheaper than a transfer for a small payment and dearer at scale. For a USD 500 payment, the wallet costs the seller USD 15 and the bank transfer USD 35, but a wallet cap of USD 2,000 a transfer makes the wallet the dearer route at larger sizes, and the bank transfer wins from USD 5,000. The book's rule is that a message saying paid is not money. A reversed payment would cost the whole order, and the seller ships on cleared funds seen in its own account.

Incoterms 2020 divide a journey of eleven steps between seller and buyer and fix the point where risk passes. Under EXW the seller carries none of the steps, and under DDP it carries ten. For one container of USD 50,000, the same goods become eleven prices, from USD 50,000 under EXW to USD 70,259 under DDP, because every dollar of cost handed to the seller adds about USD 1.18 at a margin of 15%. Cost and risk do not always meet. The rules do not decide ownership, price, payment or the governing law, and they apply only where they are written into the contract. The book states that Incoterms 2020 replaced Incoterms 2010 and that the wording of each rule should be confirmed with the ICC.

Money moves when the papers move, and the export rules add a second clock. The Foreign Exchange Act, Cap. 271 R.E. 2023 and the Foreign Exchange Regulations, 2022 (GN No. 294 of 2022) are cited for a 90 day window for export proceeds, which the book dates to Thursday 4 February 2027 for its worked shipment, with forms A and B to be confirmed with the issuer. Exports are zero-rated for VAT under sections 2(1) and 55(1) of the Value Added Tax Act, Cap. 148. The book's five constructed deals lost USD 5,000, USD 4,325, USD 750, USD 160 and USD 5,720. Each loss was preventable with an answer already in the file, and the book's lesson is that the policy must be written down before the first order.

Walking away from a bad deal early

Walk Away Early is written for careful people who stay in bad deals. The book lists seven ordinary reasons for staying: sunk time, a friend's promise, pride, fear of looking weak, pressure, secrecy and urgency. Each reason is a feeling or a history, and each can be answered by one number from the paper. The first rule is to write the walk away number before talks start. For a buyer it is the ceiling and for a seller the floor. The number is taken from the lower of two limits: the best alternative actually available, and 80% of the proven monthly result multiplied by the share and the months.

Money and hours already spent sit in a Gone column and stay out of the decision, because they are the same on both roads. Only the Ahead column counts, and it holds what the next step would cost and return. Hours are priced at monthly income divided by 160, so that they can be added to cash. The Walk Away Charter is one page of three parts: the number and the alternative, the sunk list and the forward question, and the signature with its review date. In the book's three buyer cases, the ceilings come to USD 270, USD 2,419 and USD 24,960. The third is a company case in which the ceiling sat USD 2,040 below the lowest quote, and the answer was a phase one at USD 22,500.

Ten red flags are scored on a card. Pressure, secrecy and urgency are each scored 0, 1 or 2, and a score of 2 is a reason to slow down. Flag 4 scores vague terms, flag 5 scores terms that move after a deposit, and flag 6 scores papers that never arrive. Flag 6 appeared in all twelve cases. Flags 7 and 8 test totals that differ and returns with no source. At 10% a month, USD 300 compounds to USD 942, which the book treats as a claim to test and not a return to accept. Flags 9 and 10 test a story that shifts and a record that cannot be checked. A reference call asks what was agreed, paid and changed, and a refusal to give any reference scores 2.

The Kill Test runs backwards from the promise. K1 compares the price with the median of three independent quotes. K2 back-solves the return against the proven record. K3 is the paper match, the share of the terms on paper against the terms promised in talk, and only 100.0% passes. K4 tests exposure against cash, and K5 applies the forward test. An empty cell counts as a fail, and the twelve cases failed 54 of 60 tests. The Kill Test does not judge the person. It judges whether the deal is what it says it is.

Before a verdict is acted on, the book asks for three independent voices and a night. An independent voice has no stake in the outcome, so introducers, the seller's friends and relatives, and anyone paid by the deal do not count. The same three questions go to each voice on separate calls, because shared answers narrow a group without making it more accurate. The decision then waits 12 hours. The hot note is written after the third call, the cold note the next morning, and only the cold note decides. A deadline that cannot survive a twelve hour wait is, in the book's terms, a sales device.

A polite exit is short on purpose. It has four lines: a thanks, one reason that is a number or a fact, the decision in plain words, and a door that opens only on a condition already in the Charter. The words are the same in person, by telephone and in writing, but the message leaves the best record. Every exit has a loss case, such as the delay of USD 6 for one buyer or the USD 3,900 for a company case. When money is already paid, the audit has three columns, paid, recovered and lost. In the three part-paid cases, USD 3,870 was paid, USD 1,950 was recovered and USD 1,920 was lost, which is 49.6% of the amount paid. The three exits from a part-paid deal are to stop, to restructure or to recover, and a written withdrawal gives the record that makes recovery possible.

The twelve cases cover market stalls, a salon share, a plot, a lease, a silent partnership and promised-return schemes. Four small trades show the rules at the scale of a market stall. Four larger deals kept USD 22,250 of a worst case of USD 24,050 and lost USD 1,800, all of it in the two late exits. Four investment cases failed the one test that needs no insider knowledge, which is to grow the promised return for a year and name its source. All four were left on day 2, 3 or 5, kept USD 19,000 and cost nothing. Across the twelve cases, USD 46,660 was at risk, USD 1,920 was spent in leaving and USD 44,740 was kept. The Charter's review falls on Monday 8 February 2027.

The book cites the Land Act, Cap. 113, under which a disposition that does not comply is void (section 36(1)(b)), and the Law of Contract Act, Cap. 345, on consent. It cites the Fair Competition Act, Cap. 285, Part III, on misleading conduct. A Bank of Tanzania press release of 24 February 2025 warning against an unlicensed overseas scheme is also cited, and the reported arrests are to be confirmed with the issuer.

Money in a season of heavy rain

El Niño: Where the Money Moves follows a seasonal event through the prices, claims and loans that surround a small business. It starts with the signal and a warning about its limits. NOAA's discussion of 8 October 2026 gave the Niño-3.4 region at 2.1°C above normal, with the chance of a strong to very strong event through January to March 2027 stated as above 83%. The Tanzania Meteorological Authority is the author of forecasts in Tanzania under the Tanzania Meteorological Authority Act, 2019, so the book repeats its statements and issues none of its own. The signal is a probability. The book's point is that places, roads and cash decide how much of the damage falls on any one owner.

The book's method is to name the flow of money and the period in which it moves. Food prices usually move first. In the book's model, a bag of maize rises from TZS 70,000 in October to TZS 91,000 in January, which is 30.0% above harvest, before easing to TZS 78,000 by March. Storing eight sealed bags beats selling at harvest by TZS 120,000, while the same bags in a room that floods leave the farmer TZS 244,000 behind. Transport follows the road. A lorry with a monthly profit of TZS 500,000 and fixed costs of TZS 2,500,000 a month runs at a loss once trips stop. Parked for six weeks with three trips, it falls TZS 2,250,000 short, and a detour pays only above five trips.

Building materials and repairs follow the rain as well. A small contractor carries TZS 900,000 a week of standing cost, or TZS 5,400,000 over six idle weeks. A stock of 2,000 cement bags at an assumed TZS 17,000 a bag is worth TZS 34,000,000. The book shows that a flood would lose TZS 2,720,000 on the ground and TZS 680,000 under cover, and that a cover costing TZS 400,000 saves TZS 1,640,000 in a wet season. If the season is dry, the cover is a cost of about 1.2% of the stock. For a landlord with eight rooms, the book states that works of TZS 450,000 before the rain cut the whole bill to TZS 550,000, a saving of TZS 260,000 against a flood bill of TZS 810,000.

Insurance and bank money are the least forgiving. A premium buys a defined promise, and a claim is paid only if the event, the amount, the proof and the time all fit. In the book's Life A case, a premium of TZS 20,000 turns a flooded shortfall of TZS 260,000 into TZS 80,000 when the index is triggered, and adds to the loss when it is not. The market figures in the book give TZS 488.2 billion of claims paid in 2023 against TZS 1.24 trillion of gross premiums, a simple ratio of 39.4%. That ratio mixes all lines and says nothing about one policy, and the figures are to be confirmed with the insurer. Bank money is owed on dates. In a six week cut, an instalment of TZS 1,800,000 leaves one owner TZS 1,200,000 short on the instalment alone, and the book advises a letter to the lender before the due date, since waiting for relief cost another owner TZS 360,000 of rent.

Fuel and imported inputs run through every price. For the book's Life D, TZS 360,000 of a TZS 600,000 trip cost is diesel, so a rise of 10% takes TZS 216,000, or 43.2%, off a monthly profit of TZS 500,000. Imported inputs carry a dollar price and an exchange rate, and the research the book cites gives a direction across countries but not a size for Tanzania. The owner is therefore advised to read three prices a week and not to rely on a forecast of size. The national El Niño plan is estimated at TZS 730 billion, with TZS 10 billion allocated for 2026/27, about 1.4%. Public money tends to reach emergency needs and public works first and small owners last, if at all, so the book treats relief as a late and uncertain line in any plan.

The plan itself is a sixty day programme, from Monday 19 October 2026 to Thursday 17 December 2026. Its core is a cash reserve of six weeks of fixed costs, built in eight equal Monday savings. The Buy, Store, Insure, Avoid list sorts the owner's cash into four pots, and for ten kinds of business the reserve is the largest pot, between 40% and 70% of the cash held. The shop case holds TZS 4,000,000 of stock against fixed costs of TZS 160,000 a week. Six weeks of fixed costs, TZS 960,000, is set apart first, at TZS 120,000 a week. The split holds at a reserve of TZS 400,000 and at TZS 40,000,000, and the pots are ceilings and not targets. No price is forecast and no provider is recommended.

The wet weeks are run on the reserve, with fixed costs paid first and stock second, and with suppliers and lenders told early. Reopening follows a six line checklist that starts with a cash count and ends with the cost of the cut on one page. The book's watch list gives eight lines, each with a trigger and an action.

Figures, sources and what to confirm

The shelf uses one currency convention throughout. Worked cases are in US dollars, with shilling equivalents at a book rate of TZS 2,600 per dollar, which the books describe as a convention and not a market quote. The Kiswahili editions use the same rate, so a shilling figure and a dollar figure in the same book describe the same case.

The legal references that carry weight are collected in local appendices for Mainland Tanzania. In Handbook 101 the main sources are the Microfinance Act, 2018, GN 679 of 2019 and the Personal Data Protection Act, 2022, which bears on contact lists. Handbook 102 cites the Public Private Partnership Act, Cap. 103, as amended in 2023, the Financial Leasing Act, 2008 and the Microfinance Act, 2018. Handbook 103 cites the Value Added Tax Act, Cap. 148 and the Foreign Exchange Act, Cap. 271. Handbook 104 cites the Land Act, Cap. 113, the Law of Contract Act, Cap. 345, the Fair Competition Act, Cap. 285 and the Sale of Goods Act, Cap. 214. Handbook 105 cites the Tanzania Meteorological Authority Act, 2019 and the Warehouse Receipts Act, 2005.

Several sources were read second hand or not read at source, and the books say so. They include the Land Act text cited in Handbook 101, the leasing regulations of 2022 and the crowdfunding caps in Handbook 102, the Sale of Goods Act and the Banking and Financial Institutions Act in Handbook 104, the Warehouse Receipts Act in Handbook 105, and the inflation target of 3 to 5% in the Bank of Tanzania Monetary Policy Statement for 2026/27 in Handbook 105, which is read through a secondary report. The reader should treat each as a point to confirm with the named issuer before it is quoted in a policy, a contract or a loan agreement.

The forecast figures need particular care. The book states the NOAA probability for the strong event in two places in slightly different terms, once as a chance above 83% for January to March 2027 and once in a figure caption as 83% for the October to December quarter. The wording should be checked against NOAA's own bulletin before it is repeated. The reader should take the date of issue of any seasonal statement from the Tanzania Meteorological Authority.

The positions in the shelf are stated at 10 October 2026, and the sources were read in October 2026. A date of this kind has a short life in a subject where rates, thresholds and dates move. Each handbook says that every figure should be confirmed with the issuer before anyone acts, and this summary keeps those marks. The shelf is general financial education and general business information. It is not legal, tax, accounting, financial or investment advice, and a decision that carries real money or legal risk should go to a qualified adviser, an advocate for any dispute, or the publisher, before action is taken.

Mainland and Zanzibar

The shelf applies Mainland Tanzanian rules and says so. Each of the five handbooks carries a local appendix for Mainland Tanzania, and none sets out a Zanzibar rule for collection, financing, payment, exit or seasonal money. The only Zanzibar reference in the shelf is historical. The El Niño handbook records the 1997 and 1998 rains on Pemba, Tanga and Zanzibar as part of the record of a strong event, and it does not draw a Zanzibar rule from that record.

A reader with a business or a loan on Zanzibar should therefore not apply a Mainland appendix without first confirming the Zanzibar rule with the authority that issues it. The Mainland rules on the Microfinance Act, the Public Private Partnership Act, the Value Added Tax Act and the Land Act are not transferred to Zanzibar by this shelf. The two islands are kept apart in every case, and the shelf makes no claim about how a Zanzibar lender, seller or landlord should treat the same decision.

People and Purpose is a series about the people in a money decision. Each of its seven books treats a relationship, a habit or a commitment as a business question, because the way two people talk about money, the way a mentor is chosen, the way a seller listens, or the way a group holds its cash decides what is paid, kept, sold or lost. Every book is a working kit: a method, tools to fill in at a desk, a dated routine and a list of points to confirm. The books give no legal, medical, tax, investment or therapeutic advice. Where a matter needs a counsellor, mediator, lawyer, registry or issuer, the book says so and names the office to ask. This summary explains how the books are built, what each one asks the reader to do, which statements must be checked before anyone acts, and where the examples need their dates moved. It is general information only. Its routines are dated in October and November 2026, and a reader who starts later must move every date.

How every book is built

The seven books share one design, stated at the front of each. Each answers one decision. Each figure is computed from the inputs printed beside it, never from a number typed in. Each rule is named with its issuer and a date, and the reader is asked to confirm it before acting. Illustrative people and businesses are described by type and labelled as examples, and each worked example says so. The books close with a notebook exercise, in which the reader copies a chapter heading and writes one line per point as it applies to the reader's own case.

The result is a series that asks for practice more than agreement. A reader gets a method, a routine with dates and a short list of mistakes to avoid. None of the books names a bank, lender, counsellor, app or school. A reader who needs one must choose and check the provider independently. Books 1 and 2 draw on the Bible, and Book 2 reads Genesis 12 to 19 as its main text. The reading is offered as text, and it does not promise any result in business or in a marriage.

What the reader leaves with

Each book ends its method in a kit. Handbook 106 gives a Weekly Talk Card, a set of money scripts, a six-rung Repair Ladder and a Mediation Folder. Handbook 107 gives a Lot Ledger, a Choose by Values worksheet and two sets of seven marks, one for mentors and one for partners. Handbook 108 gives a five-section daily journal and a weekly review sheet. Handbook 109 gives a countdown kit, a seven-day start sheet and a Handover File. Handbook 110 gives a Gaming Radar of twelve moves, a measurement chain and an experiment design. Handbook 111 gives a research note, a Summary Back script and a same-day summary. Handbook 112 gives a one-page constitution, a five-head study sheet, a shared book and an Opportunity Study sheet.

The tools are meant to be used on a particular day. Each book sets one week or one month of practice, so that the method is tested in the reader's own life before it is judged. A reader who finishes a book without filling in its first tool has read the book and has not used it.

Talking about money without the fight

Constructive Dialogue, Not Arguments (Handbook 106) begins from a plain observation: love and trust do not prevent a quarrel about money, work or family duties. The book holds that the way a talk is conducted, rather than the feeling behind it, decides whether it ends in a fight. Its method is a weekly talk of twenty minutes on a fixed day, in a neutral place, with a timer that belongs to neither person, and with three questions on the Weekly Talk Card. Two habits do most of the damage. Blame passes itself off as a fact when it is a sentence about a person. Excuses answer a charge by explaining why it is unfair, so the fact is never stated. The book names a swap for each that can be said in one sentence.

The money conversation is built as a five-part script, and its first part holds sums only. The book's scripts run from a sum of USD 40 to a sum of USD 30,000, and each sum has its own wording. When a fight occurs, the Repair Ladder has six rungs, to be climbed within 24 hours, and the mending does not require the two people to agree on what the fight was about. A Mediation Folder of six items is gathered over the weeks, so that a counsellor, elder or mediator starts from facts. The book's own arithmetic compares a year of 52 Sundays, made up of 40 plain talks, eight money talks and four reviews, totalling 1,400 minutes or 23.3 hours, with an assumed 36.0 hours of fights in a year. That assumption belongs to the book's model and should not be read as a measured figure.

The last chapters widen the method. Couples who run a business together are asked to keep two tables, one for the marriage and one for the company, with a rule about which decision goes to which. Family partners are asked to keep shares, work and in-laws on separate lines of one signed page. A talk is not enough when the same topic returns three Sundays in a row, when two reviews end without a decision, or when anyone is afraid. In those cases the book points to a counsellor, elder or mediator, and it states that a mediator helps the parties reach their own agreement and does not decide for them.

Choosing the people to walk with

Follow Abraham, Become Lot (Handbook 107) reads Genesis 12 to 19 as a business story. Lot prospered by going with Abram, then lost ground when he chose the well-watered plain by what his eye saw. The book's central idea is that a mentor, partner or patron carries something of value, and that the terms of going with someone are usually the one item nobody writes down. It sets out a Lot Ledger in which each gift is recorded with its date, its helper, its value and the return made. A gift is valued at what the receiver would have paid, and an introduction or a door is left unpriced. The ledger is for seeing, not for collecting, and a gift is never turned into an invoice.

The book then tests people. A mentor is measured by seven marks of what the person has done, and a partner by seven marks, each of which can become a clause in a short note signed before money moves. A decision about direction is made on a worksheet with five weights, two sets of scores and a date for review. Where two partners have outgrown the same land, the book sets out a parting in steps, handling the words, the paper, the money and the people in turn. The two ledgers in the book show returns of 50.4% and 73.2%, and both are illustrations, not findings. The routine in the book runs six Sundays from 18 October 2026, with a review on 29 November 2026, and must be moved to the reader's own dates.

A record that holds up

Journal Everything Important (Handbook 108) argues that memory drops promises, ideas and numbers within days, and that a short daily record catches them. The method is one journal, on paper or a phone, with five sections a day, each taking about two minutes: what happened, what was promised, what was spent, one idea, and one thing to do tomorrow. The first thirty days introduce each section in turn. A weekly review of about twenty minutes then turns seven notes into three lists and three actions. The book's discipline target is 27 ticks in 30 days, with a catch-up note for any missed day rather than a double entry on the next.

The book is careful about evidence. A dated note settles a dispute only as far as it was written on the day. A journal is not a tax record and does not replace receipts. The book cites Tanzania Revenue Authority guidance, as read, that a person liable to tax keeps the documents needed to determine tax for at least five years after the income year, unless the Commissioner notifies otherwise, and it marks that rule for confirmation with the issuer. Privacy rests on three steps: a lock, three copies held in two places, and a one-page handover note that names who may read the journal if the keeper is absent. The final chapters follow five different people through the same system, and they set out the next ninety days from Day 31 to Day 120.

Dreams with a date

Die But Not With Your Dreams (Handbook 109) starts from the deadline every life carries. It asks the reader to count years honestly, list gifts and shelved ideas, and choose three that matter. Each chosen dream receives a first step of under one hour, costing at most USD 20 where the idea allows, and then a ninety-day start plan of thirteen weeks. Its money rule is a start fund of 25% of one year's surplus, sized against the 168 hours of a week and three named people. The book treats late starts as normal. It cites dated records of people who began at 65, 75 and 78, and it describes the first year as a ledger in which the first ninety days cost money and the months that follow repay it slowly.

The fourteen decisions run in a fixed order: the date the countdown begins, the test for each excuse, the sober count of years, the ledger of gifts, the shelf of ideas, the choice of three, the first step, the ninety-day plan, the start fund, the late starter's single move, the first year, the Handover File, teaching one person and writing the letter, and a five-year calendar. A dream is finished, the book says, when a second person could pick up its work the next day. The Handover File, a ten-page document, is the instrument for that, and a yearly count falls on the third Monday of January, beginning on 18 January 2027. The examples start with sums of USD 50, USD 510 and USD 12,500. The Kiswahili edition converts at TZS 2,600 per dollar, a working rate the book states, which the reader should confirm at the time of use. The estate material is flagged for confirmation. The Probate and Administration of Estates Act, Cap. 352 was read for Parts V and XI only, in a TanzLII copy as at 31 December 2023. The business register kept by BRELA, the land laws, the copyright law and the Tanzania Revenue Authority requirements were not read in full, and each must be confirmed with its issuer.

Measuring people without bending the number

People Change When They Are Measured (Handbook 110) starts from the observation that people react the moment they know they are watched or scored, and that the reaction can be either better work or cleverer bending of the figure. The book sets out twelve recognisable ways a number can be bent, each with a sign that can be looked for in the records. Its design rules are that every number sits in a chain of input, output and outcome, and that the safest design pairs an easy measure with a late one. It also covers how often scores are read and by whom, why a target should sit a little above where each person stands rather than at one figure for everybody, and why a small, random, unannounced check of records must be run by someone other than the person checked.

Four case groups show the bending in practice: shop counters, field sales teams, schools and lending or collections. In the lending case, a late loan rewritten as a new loan lowers the arrears figure, so the late share is read honestly only beside the rewritten loans. The chapter on fair performance management and warnings sets an order: the standard is written down and made known, the notice is given, support comes before any warning, and warnings are made in writing. The closing chapter sets a ninety-day experiment. It begins with a baseline fortnight before any change, uses a control drawn by lot, writes a stop rule before day 15, freezes the numbers on day 89 (Friday 15 January 2027) and makes the decision on day 92 (Monday 18 January 2027). The research cited is drawn from several countries and is described where it applies. It is not presented as Tanzanian data unless the text says so.

Listening before the pitch

Make the Client Feel Understood (Handbook 111) holds that a client first buys the feeling that the problem and its cost have been understood, and only then the service. It replaces the twenty-minute pitch with a discovery meeting built backwards from one moment: a summary that the seller says back and the client corrects. Preparation is a thirty-minute timed research note, followed by a short message the day before that sets out the agenda and promises that no pitch will come before the summary. The opening three minutes are scripted in five beats. The questions are open, they put the situation before the problem, and they are tested against past behaviour. The cost of the problem is built in the client's own numbers, line by line, and confirmed aloud before any price is mentioned.

The book treats a hesitation or a polite phrase as a signal for a question, never as a verdict. A pause of four seconds can be waited out, and an objection such as "it is too expensive" is answered with a question. Each meeting closes with a named who, what and when, said aloud and written by both sides. Follow-up is five dated touches, and the thirty-day practice plan adds up to 570 minutes, an average of 19 minutes a day. The book's legal notes are flagged for confirmation. An agenda is not an offer under the Law of Contract Act, Cap. 345 (Mainland Tanzania). The Personal Data Protection Act, 2022 is reported to have come into force on 1 May 2023. The Cybercrimes Act, 2015 governs recorded conversations and was not read for the book. The book recommends paper notes unless the client agrees to a phone or recorder.

From the corner to a circle

Badala ya Vijiwe, Vikundi vya Fursa (Handbook 112) takes the hour a group of friends already spends at a coffee corner and proposes an opportunity circle of eight to fifteen members. It has one book, one vote for every spend, one five-head study sheet for every idea, and a small fund built from weekly contributions. The book separates a circle from a rotating pot, a savings and loan group, a SACCOS, a company, a price ring and any scheme that names a return. Its constitution has twelve short clauses, signed before the first shilling is paid. Four officers, a chair, a secretary, a treasurer and a scout, serve terms of six meetings, and the treasurer never counts alone.

The worked case is a circle of ten paying TZS 5,000 each, beside a market group of twelve paying TZS 20,000 and a trader group of eight paying TZS 50,000, with a fourth group that failed to hold together so that the rules can be seen working. The modelled circle shares out TZS 3,158,400, or TZS 315,840 a member, while the weak case records a loss of TZS 182,400. The book states that its figures are examples, that the fund can lose money, and that a group account needs confirmation with the Bank of Tanzania and the account provider. The year runs as 52 meetings in four quarters, with a count every fourth meeting and a report back from meeting 16. Two circles may study together, each with its own book, vote and loss.

Registration is the last step, and it is not settled. Whether a circle that studies and funds opportunities, and does not lend to its members, counts as a community microfinance group was not found in the sources read. The Regulation 8(2) list in the 2019 Regulations was read from a policy library copy, and a version dated January 2022 was not read. The book therefore sets out the papers a circle will hold in any case, including two signed copies of the constitution and minutes from the first meeting, and it leaves the registration question to the Bank of Tanzania or the local government office.

Shared tools, shared mistakes

Several mistakes recur across the seven books, and each book names them. The first is to talk without a fixed time and place, so that the talk takes over the week. The second is to keep a promise in a phone call, where it is lost, rather than on a line with a name, a task and a date. The third is to measure a team before a baseline is written down, which leaves no means of telling a gain from a normal week. The fourth is to pitch first, which costs meetings the seller never gets back. The fifth is to keep a group's fund in one member's personal wallet, which the book on circles identifies as the pool that failed. The sixth is to start a dream without a handover, which leaves the dream with its owner alone.

The books also share a discipline about money: every amount in them carries a date, a source and a condition. A reader should treat a figure without those three as an open question. Where a figure is a book's own assumption, such as the 36.0 hours of fights or the 25% start fund, the reader should replace it with the reader's own number before relying on the result.

Where the sources stop

Several statements in the series are marked for confirmation, and each should be checked with the named issuer before anyone acts. The Bank of Tanzania and FinScope figures in Handbook 112 are the first. FinScope Tanzania 2023 puts formal financial inclusion at 76% of adults, against 65% in 2017, and community microfinance group membership at 12% of adults, with a 2017 figure of 10% to be confirmed. A news report of July 2026 citing Bank of Tanzania registry data counts 78,453 registered microfinance providers, of which 74,534, about 95%, are Tier IV community groups. These are registration counts from a second-hand source, not counts of members or savings.

The second group concerns Mainland Tanzania law. The Personal Data Protection Act, 2022 is named in Handbooks 109 and 111 and was not read in full. The Cybercrimes Act, 2015 governs recording in Handbook 111 and was not read. The Law of Contract Act, Cap. 345 is applied in Handbook 111 and flagged for confirmation. The five-year record-keeping guidance in Handbook 108 is stated as read. The Probate and Administration of Estates Act, Cap. 352 was read in part. The third group is calendar and rate. Handbook 106 gives Wednesday 9 December 2026 as a Tanzanian public holiday, to be confirmed with the issuer. Handbook 109 converts at TZS 2,600 per dollar for its Kiswahili edition. The fourth point is scope. The series sets out no Zanzibar law. Where a law is named, it is a Mainland Tanzania law, and a reader in Zanzibar must confirm the position with the Zanzibar authority concerned.

Dates in the examples

Each book's routine begins on a dated start in October 2026, most often Sunday 18 October, Monday 19 October, Thursday 22 October or Saturday 24 October, and runs to a first review in the early months of 2027. As at 26 October 2026, several of those starting dates have passed. A reader who begins later should move each dated step, keep the intervals between the steps, and treat the example dates as a pattern rather than a calendar. Where a book gives both a closing date and a day count, the day count should govern. Handbook 108 is an example: its ninety days are given as 18 November 2026 to 15 February 2027, while its routine closes on Sunday 14 February 2027.

Where each book fits

The seven books can be read in any order, because each stands alone. A reader facing a relationship under strain should begin with Handbook 106. A reader choosing a mentor or partner should begin with Handbook 107. Handbook 108 is the record the others rely on, and it is a sound first read for any reader who loses track of promises. Handbook 109 suits the reader who has postponed a dream. Handbook 110 suits the manager, owner or lender who scores other people. Handbook 111 suits a seller meeting clients, and Handbook 112 suits a group of friends that wants to study and fund an idea together.

A reader who runs a household, a family firm or a small team should read Handbooks 106, 108 and 110 together. Each depends on a dated record, a fair talk and a measure that cannot be bent, and the three can share one notebook. The reader who needs one book only should choose the decision in hand and then return to the others when a second decision arrives.

What the series does not do

The series does not give legal, tax, investment, medical or therapeutic advice, and it does not stand in for a lawyer, a registered counsellor, a tax adviser or a mediator. Its examples are constructed, and their figures are not forecasts. It names no provider, lender or app. Where a step depends on a rule, the series names the office that holds the rule and asks the reader to check it there, because a rule that has changed since the books were written would otherwise be repeated as fact.

The series is also not a substitute for a second opinion on a personal matter. A reader in danger, or a reader whose relationship involves threats, should seek help from the services and authorities nearest to them before using any routine in these books.

Kilimo Biashara is the Zatra Handbooks series for people who want to earn from land. It has two books. Hut House for Cultural Tourism (Handbook 113) shows a family or group how to host visitors in a simple village home of two to four rooms, priced and registered before the first guest. Endless Demand: Kilimo Biashara (Handbook 114) compares eighteen crops and enterprises on capital, timing, buyers and risk, then helps a farmer build a one-page plan. Both books treat farming and hosting as businesses and apply Mainland Tanzania rules; Zanzibar is not covered. Every figure is worked from stated inputs and is illustrative, and every rule is to be confirmed with the issuer named beside it.

What the series covers

The series has two books, both first editions with positions stated at 10 October 2026. Book 1 of 2, Hut House for Cultural Tourism, is written for families, community groups and returnees who own land or a home near a park, a market town or a cultural site. Book 2 of 2, Endless Demand: Kilimo Biashara, is written for farmers, landowners and investors who want crops with buyers in every season. The books can be read separately. A reader with a plot near visitors starts with the hut house, and a reader who farms for the market starts with the crop comparison.

The books share one method. Each makes a single money decision, sets out the costs as a written sum, shows a good year and a slow year side by side, and ends with a kit of tools for use at a desk. Hut House has fifteen chapters and eleven tools. Endless Demand has thirteen chapters and ten tools. Both describe themselves as education and general business information, not legal, tax, accounting, financial or investment advice, and both accept that the business can earn less than it costs.

How to read the figures

Every money figure in the series is either an input stated in the text or a result computed from those inputs. The cases are composites described by type, not by name. The hut house follows three costed plans: a family turning two rooms of an existing house into a guest home for TZS 2,000,000, a group of eight building three rooms at a park gate for TZS 7,000,000, and a returnee building four rooms for TZS 15,000,000. A six-room block costing TZS 22,200,000 is included as the case that outgrew its guests. The farm book follows a market trader with 3 acres and TZS 2,400,000, two civil servants with 25 acres and TZS 15,000,000, and a haulage owner with 120 acres and TZS 40,000,000.

Several assumptions run through both books and should be replaced with local figures before any use. The hut house converts visitor prices at TZS 2,600 to the dollar, assumes a booking cost of 6% and sets a village share of TZS 5,000 a guest-night. The farm book assumes every price, yield and cost, and leaves out land rent, family labour, soil and the farmer's skill. Cases dated after the stated date are projected worked examples, not records of events.

Hut House: the order of work

The book follows the order of work in five parts: the offer, the build, the experience, the business and the launch. The offer comes first, because a hut house that cannot describe what a visitor buys will be planned around the wrong rooms. The build follows, with the materials list and the first-night, hygiene and fire rules. The experience covers the guest day, the guide and the house rules. The business part covers licensing, pricing, booking, complaints and the village share. The launch closes with the first ninety days. Each chapter ends with one thing to do that night, drawn from the Hut House Kit, so the reader can stop at any chapter with a concrete step taken.

A reader with a plot and a spare weekend can go straight to Chapter 1 and run the junction count, which is the book's own suggested start.

Hut House: the offer and the site

The opening chapters argue that a visitor buys a home, a meal and a day, and that none of these needs a large building. The worked example has 120 visitors pass a junction in a week, 30 stop and one ask where to sleep. One room-night at TZS 80,000 brings receipts of TZS 108,000 including meals, against variable costs of TZS 46,800. That leaves a contribution of TZS 61,200 towards fixed costs of TZS 50,000 a month. The Tanzania Tourist Board has run a Cultural Tourism Programme since 1996, and its programme counts are cited to show that the model is not new, though they do not measure earnings.

A site is chosen on seven tests, each scored 0, 1 or 2: reach, arrival, story, land, neighbours, water, and safety with quiet. The maximum is 14. A total of 10 to 14 goes forward, and a score of 0 on land or neighbours stops the project whatever the rest of the scorecard says. The story must be three lines the household can check by walking the ground. A low total is treated as a result with three honest outcomes: not yet, a day visit only, or not at all.

Hut House: rooms, materials and build

Rooms are sized to the nights they will actually receive. Plan S costs TZS 2,000,000 and earns a surplus of TZS 3,806,400 in a good year, with a payback of 6.3 months and a slow-year surplus of TZS 134,400. Plan M costs TZS 7,000,000 and earns TZS 5,772,000 in a good year, with a payback of 14.6 months and a slow-year shortfall of TZS 1,288,000. Plan L costs TZS 15,000,000 and earns TZS 13,197,600 in a good year, with a payback of 13.6 months and a slow-year shortfall of TZS 1,950,400. The six-room block keeps 7.1% of the smallest plan's surplus for 11.1 times the build. Each plan holds a reserve of three months of fixed costs, which cushions a slow year but does not cure one.

Materials are chosen from what the land and the village can supply. Plan M's materials list has 19 lines totalling TZS 6,300,000, with a contingency of TZS 700,000 and a build of TZS 7,000,000. Walls, roof and floor are chosen for how they are kept as well as bought. The labour line is a count of days at a day rate that the host asks for and writes down. The contingency is not the reserve. A thatched roof is cool and local but it burns, so the cooking fire stands apart from it.

Hut House: beds, water, toilets and fire

A guest sleeps well when the first night is walked in order: a lit door, a clean bed, a treated net without holes, a covered jug of boiled water and a place to charge a phone. Plan S spends TZS 700,000 on sleep, light and water, or 38.9% of its items. Water is planned at 20 litres a guest a day, so four guests need 80 litres and a 200-litre drum lasts 2.5 days. The boiling guidance the book uses comes from foreign sources and is not a Tanzanian rule, and boiling does not remove chemicals.

The toilet rule is a planning rule, not a legal standard: one toilet and wash place for every two rooms, with a hand-wash point at each toilet and at the eating place. That gives one toilet for Plan S, two for Plan M and two for Plan L, at TZS 140,000, TZS 360,000 and TZS 680,000. The distance of the pit from water is set by the environmental health officer. Fire safety rests on rules rather than equipment: no flame in any room, wiring kept 30 centimetres from the thatch, and a door and window in every room that open from inside. The only sourced study is a United Kingdom leaflet on thatch fires, and it is not a Tanzanian rule. A night drill is timed and recorded each quarter.

Hut House: the guest day, guides and house rules

The guest day is built in six slots from morning tea to the evening fire, each with a price and a cost. A pair's meals sell for TZS 28,000, cost TZS 18,000 and leave TZS 10,000, or 35.7%, before labour. A pair who take a walk, a cooking lesson and a craft hour pay TZS 90,000 and keep TZS 46,000, or 51.1%. Ceremonies, prayers, neighbours' homes and children are not for sale unless their owners agree.

The guide is a neighbour chosen with the village, trained and paid on the day. On a TZS 20,000 walk the guide receives TZS 12,000, or 60.0%. The community decides what it shares, and every photograph of a person is asked for, with a no as a complete answer. Money is paid at the house on a board of prices in shillings and dollars. Two cards carry the rules: a guest card at the welcome and a host card in the kitchen.

Hut House: licensing and registration

This is the chapter where the book is most cautious, and the route it gives is a sequence of questions rather than a set of answers. The route has nine questions, each carrying a label to confirm. The first source is the Tourism Act, 2008, read from the FAOLEX copy as passed. As read, it requires tourist facilities to be registered (section 8), sets a 45-day period for grading decisions (section 11), requires a licence (section 21), and names the Tanzania Tourism Licensing Board as the licensing body (section 22). Amendments since 2008 were not checked, and the book prints no penalty.

The only fee table read is the Tourism (Fees and Charges) Regulations, 2015, GN No. 267 of 2015, in force from 17 July 2015 to 28 December 2017. It is superseded, and a later amendment was not read. A second-hand report says fees were reduced in 2025, so no fee is printed. Whether a hut house of two to four rooms is a registered, licensed or graded facility, and in which class, was not established. A September 2018 warning from the ministry's licensing director to unregistered short-stay providers is printed as history, not as current enforcement.

Village land is handled through the Village Council and the Village Assembly. The Village Land Act was read only as a summary of 1 October 2025. The model sets aside TZS 20,000 a month for licences as a pot of money, not as a fee. The practical sequence is to ask the Ministry of Natural Resources and Tourism which class and duty apply, to ask the Tourist Board how a village home joins its programme, and to write each answer beside its label before the first paying guest.

Hut House: pricing, booking, complaints and the village share

The price of a night starts from its cost. In the case modelled, six nights a month at TZS 80,000 (USD 31 at the stated rate) leave TZS 317,200 after fixed costs. The price floor for a target of TZS 200,000 a month at six nights is TZS 59,220 (USD 23). A resident rate of TZS 48,000 still pays, at TZS 31,120 a night. The book warns that a host who prices from a neighbour's sign can sit below the floor without knowing it, and that no price rescues a house that is too big for its nights.

A stranger books on ten photographs, a plain description and the words of earlier guests. Six channels are named by kind, with no fee printed. The book sets a deposit of 30% and a refund rule that gives full, half or no refund by the days before arrival. In one case file, a no-show's deposit of TZS 72,000 covered 39.2% of the TZS 183,600 the room might have earned. A five-step reply to a complaint listens, separates the fact from the feeling, names the fix, gives a date and invites a return.

The village share is paid per guest-night. A two-night stay at TZS 256,000 pays TZS 109,000, or 42.6%, to people outside the host family, including TZS 20,000 to the village fund. The fund is held by two keyholders from outside the host family and recorded in a book with a quarterly count. A Village Agreement of ten clauses sets out who pays, who holds and who decides. The TZS 5,000 a guest-night is an illustration, not a recommended rate.

Hut House: the first ninety days

The book ends with a rehearsal night with relatives, at an illustrative cost of TZS 48,000, followed by a dated routine. The routine starts on Monday 12 October 2026 and its count falls on Monday 12 July 2027. A room is added only when more than 50% of room-nights fall in the three busiest months. In the book's three plans that share reaches 16.7%, 18.9% and 20.0%, so none passes the test. The host names one decision at day ninety, to keep, change or stop, and writes it down with a date.

Endless Demand: demand is not the buyer

The opening chapters separate demand from a buyer. Demand for food, fibre, fuel and export does not end, but a farmer sells to one buyer at one gate in one week. In the book's example, ten farmers who each plant 3 acres of maize at an assumed TZS 1,000 a kilogram earn TZS 1,140,000 each. When the price halves, each makes a loss of TZS 210,000, and the village's revenue falls from TZS 27,000,000 to TZS 13,500,000. The National Bureau of Statistics counted 8,971,794 agricultural households in its 2023/24 survey, of which 8,821,600 are on the Mainland and 150,194 in Zanzibar.

Case S compares a neighbour's 3 acres of maize, with a ten-year net of TZS 6,000,000 and TZS 2,922,000 in three bad years, against a four-crop plan on 2.3 acres that nets TZS 16,810,000 and TZS 11,922,250 in bad years. The book ranks crops by stated rules and prints every gain beside its loss.

Endless Demand: filters, capital and the crop profile

A plot passes three filters in order: water, soil and the road to a buyer. A fourth filter, money, removes crops whose capital per acre is more than the savings. Twelve of the eighteen crops can be funded on one acre with TZS 2,400,000, from sesame at TZS 380,000 to sugar cane and timber at TZS 2,000,000. The book shows the cost of skipping the filters: TZS 10,500,000 lost by apples on a warm site, TZS 6,300,000 for oil palm with no mill within reach, TZS 3,200,000 for cane with no contract, and TZS 7,500,000 for moringa with no steady buyer.

Each crop profile has six lines: regions, capital per acre, months to income, buyers, the assumed price and risk. Capital per acre is the larger of the first-year outlay and the deepest cumulative shortfall, so timber needs TZS 2,000,000 an acre, avocado TZS 4,000,000, banana TZS 2,600,000 and maize TZS 520,000. The loss case is the same for every crop: three bad years, in years 3, 6 and 9, at 70% of the yield and 75% of the price, which keeps 52.5% of revenue. Buyer depth is scored from 1, meaning one or two buyers, to 5, meaning buyers in every village. The book asks the reader to call three buyers before planting.

Endless Demand: trees, fruit, grains and thin markets

Trees cost the farm every year in which they earn nothing. Timber needs TZS 2,000,000 an acre and its first revenue arrives in month 66. Mango, oil palm and cashew need TZS 2,600,000, TZS 2,800,000 and TZS 1,750,000 and first earn in months 40, 40 and 36. Timber nets TZS 8,350,000 over ten years, the lowest of the four, and is ahead only in year 10. A fire in year 8 costs timber TZS 950,000, and an oil palm plot with no mill nearby loses TZS 6,300,000.

The fruit chapter treats banana as the soonest payer: first bunches near month 14 on TZS 2,600,000 an acre, with a ten-year net of TZS 23,000,000. Avocado needs TZS 4,000,000, waits 42 months and nets TZS 29,200,000. Apples need TZS 4,300,000 and net TZS 21,300,000, but a wrong site costs TZS 10,500,000. Grapes need TZS 4,500,000, net TZS 25,700,000 and carry rain risk at harvest. Among grains, maize nets TZS 200,000 a season, rice TZS 780,000 and silage TZS 550,000. In a bad season each loses money: TZS 142,000, TZS 160,500 and TZS 305,000 respectively. Storing 2,700 kg of maize gains TZS 513,000 at an assumed price of TZS 750 a kilogram and loses TZS 162,000 at a flat price, with a break-even of TZS 560 a kilogram.

The small-crop chapter finds beans the quickest and steadiest: first income in month 4 on TZS 480,000 an acre, a ten-year net of TZS 5,100,000, and a small gain even in a bad season. Groundnuts carry aflatoxin risk and can lose the whole outlay of TZS 750,000 an acre. Moringa, honey and sugar cane are thin markets in which one or two buyers can set the price. A mill contract is needed for cane, and without one the book shows a loss of TZS 3,200,000.

Endless Demand: ranking for small, medium and large farms

The ranking first gates crops on capital, then scores five measures from 1 to 5: return, speed, buyer depth, loss and fit. For a small farm of 3 acres with TZS 2,400,000, the weights are 2, 4, 2, 3 and 3, with a maximum of 70. Beans score 70, sesame 65 and groundnuts 63. The book attaches a loss to each leading crop, and it notes that the score ignores land rent, family labour, soil and the farmer's skill.

For a medium farm of 25 acres with TZS 15,000,000, the weights change to 4, 2, 2, 3 and 3, which doubles the weight on return and halves the weight on speed. Beans score 67, sesame 65 and banana 63. The top three together need TZS 17,300,000, which is TZS 2,300,000 more than the savings, so the table gives an order of preference rather than a plan. For a large farm of 120 acres with TZS 40,000,000, capital rather than land is the limit. Rice at TZS 1,200,000 an acre funds 33.3 acres. The large weights are 4, 1, 3, 3 and 4, with a maximum of 75. Rice scores 69, sesame 68 and beans 67.

Endless Demand: quick crops, trees and the one-page plan

Quick crops pay for trees while they wait. In Case M, a 25-acre farm with TZS 15,000,000 of savings, rice, beans and sunflower earn a surplus of TZS 6,080,000 a year against tree costs of TZS 6,300,000. The plan as first written, with every tree planted in year 1, closes year 1 with a loss of TZS 220,000 and a reserve of TZS 260,000. It trails the all-annual plan for five years and leads from year 6. Its ten-year net is TZS 110,650,000, against TZS 65,600,000 for the all-annual plan.

The one-page plan records the land, crops, outlay, buyers, calendar, bad years, reserve and review dates. Its Case M page delays the banana and timber to year 2, giving a year 1 outlay of TZS 12,040,000, a year 1 gain of TZS 2,480,000 and a reserve of TZS 2,960,000. Its ten-year net is TZS 99,700,000, or TZS 69,509,000 in three bad years, and the timber harvest then falls outside the ten years. Case L shows the cost of concentration: an all-rice plan nets TZS 31,450,000 more than the mixed plan in the base case, but TZS 39,830,000 less when rice fails in years 3 and 4.

Endless Demand: reading each plan in three columns

The book asks that every plan be read in three columns: the base case, three bad years and the failure of the main crop. A plan that looks strongest in the base case can be the weakest when the main crop fails, and the three columns make that visible on one page. Case L makes the point. Its all-rice plan is the stronger of the two in the base case, but the mixed plan is the stronger when rice fails in years 3 and 4, so the choice depends on how much of the farm's living rests on one crop.

The book also reads the first year closely. A year 1 outlay must fit within the savings, and what is left is often thin. The reserve is written as a number on the plan, not as a hope, and the first review is set for a fixed date so that the leading figures are tested before the long-run ones are relied on. Planting months are counted back from the months of income and confirmed with the district officer before any seed is bought.

Endless Demand: the Mainland rules a farm must ask about

The local appendix on Mainland Tanzania lists the instruments behind ten questions a farmer would ask. Several rows were read for another handbook and not re-checked for this book. Land for a non-citizen is allocated only for investment, under the Land Act, Cap. 113, sections 19(2) and 20, as read on 8 October 2026. Farming one's own produce falls outside the Business Licensing Act, s.3, as read in Handbook 06's appendix at fees stated 26 September 2026. Traders, processors and warehouse operators in cereals and other produce are registered under the Cereals and Other Produce Act, 2009. Pesticides and plant health permits fall under the Plant Health Act, 2020, and fertilizer registration under the Fertilizers Act, 2009. The Warehouse Receipts Act, Cap. 339, was read from summaries only.

The tax rows come from the TRA guide 2025/2026, which gives a corporate rate of 30%, VAT of 18% on the Mainland and zero-rated exports. The Finance Act 2026 was not read. The Cashewnut Industry Act, Cap. 203, was not read, and no rule for timber, sugar, honey or oil palm was established. Growers of those crops must therefore ask the cashew board, the forest agency and the crop boards before a nursery order is placed, which the book dates to Monday 2 November 2026. The appendix prints no fee, cess, levy or processing time. The order of asking is the district agricultural officer first, the land office next, the council on market and road charges, and the Tanzania Revenue Authority before the farm sells beyond its own produce.

Mainland and Zanzibar

Both books cover Mainland Tanzania only. Hut House states that Zanzibar has its own authorities, and that a Zanzibar headline on non-hotel tourist accommodation was read as a headline only, not as an article. Endless Demand states that Zanzibar has its own regime and is not covered, although its household count is reported beside the Mainland figure. A reader with a plan in Zanzibar should not rely on either book's rules and should take the question directly to the Zanzibar authorities.

Figures and rules to confirm before acting

Several points in the series are provisional and should be confirmed with the named issuer before any decision. The Tourism Act, 2008 is read from a copy that predates later amendments. The 2015 fee regulations are superseded, and no current fee is printed. The Village Land Act is read from a summary dated 1 October 2025, not from its text. The TRA rates omit the Finance Act 2026. Dollar prices use TZS 2,600 to the dollar, which the reader should replace with the rate of the day. The 6% booking cost, the TZS 5,000 village share and every price are illustrations. Crop prices, yields and costs are assumed.

The dated steps are the Ministry and Tourist Board questions before the first guest, the cashew board and forest agency questions before the nursery order on Monday 2 November 2026, the first farm review on Monday 7 December 2026, which tests the leading figures rather than the ten-year ones, and the hut house's day-ninety count on Monday 12 July 2027.

Where to start

A reader with land near visitors and a modest budget starts with Hut House, runs the site scorecard, and asks the licensing questions before any building begins. A reader with a farm plan or savings to invest starts with Endless Demand, applies the capital gate, and asks the district officer about soil, water and buyers for the three crops that remain. A reader with both a plot and a farm can take both books, but should keep the two budgets separate so that a shortfall in one is not hidden by the other.

Each book closes with one step to take in the coming week, written in a notebook under a heading copied from the chapter. The series asks for answers to be written down, because the pattern shows on paper and not in the head. The series is general information and not advice. Where a decision carries real money or legal risk, a qualified adviser, and Zatra, should be consulted before acting.

Cash That Works is a two-book set in the Zatra Handbooks on the weekly routine that lets a Tanzanian firm pay its people on Friday. Book 1, Handbook 115, builds a six-week cash page, prices work so that it pays, sets out a collection routine and places the dates that leave regardless. Book 2, Handbook 116, tells the same habit through five owners in five towns, so that the reader can learn the tools through a story before reading the rules. Both books are written for owners and managers of shops, workshops, contractors, cooperatives and service firms that earn a profit on paper and still run short of cash. The series rests on one principle: every rule carries an issuer and a date, and every case is labelled illustrative. Mainland Tanzania and Zanzibar are kept apart throughout, and several rates are marked Verify because they move each 1 July. This summary explains what each book covers, the method both share, the points that need confirming before use and the order in which the reader should approach them. It is general information, not legal, tax, accounting or investment advice.

What the series covers

The series has two titles, and it is designed to be read as one routine. Cash That Works (Handbook 115) is the working manual. It runs to nine tabs in four parts, from seeing the next six weeks, through money in and money out, to the page in other people's hands. It closes with a Cash Binder of ten tools, a season calendar, a Kiswahili section on records and credit, a Figures to Confirm table and a glossary. The Friday Page (Handbook 116) is the companion. It follows one Dar es Salaam printer and four other owners through the same six weeks and sets out the habit as a story. Its nine tools are gathered in The Friday Folder.

A reader who has never kept a cash plan should begin with Tab I of Book 1 or with the opening story of Book 2. A reader who already runs a plan but lacks a price or a collection routine should go straight to Tabs III and IV. The books do not replace a bookkeeper. They ask for a record that the owner keeps every week and that a lender, a tax officer or a partner could check without help.

Both books rest on one claim, stated at the front of Book 1. Profit tells whether the work was worth doing. Cash tells whether the next week's work can be done. A firm can show a profit for a month and still be unable to pay wages, because a sale is counted when it is delivered and the money arrives when the customer pays. The series teaches a weekly page that counts money on the day it moves, so that a gap is seen while it is still small.

Every figure in the series is either computed from inputs printed beside it or taken from a named issuer with a date. The cases are composites. People and firms are described by type, and the towns are real. Figures marked Verify have been read from an issuer or from commentary and must be confirmed again before use. A small number are marked Unknown, where the sources read did not settle the point, and in those places the book prints no figure and names the office to ask.

Profit is not cash, and the page shows the difference

Tab I opens with a workshop in Mabibo whose owner delivered a large March order, recorded the sale at once and was paid in May. The paper, wages and electricity for the order left the bank in March. The profit is real, but the cash is two months away, and for those two months the workshop has in effect lent money to its customer without interest or agreement. The book's answer is the Friday page: six weeks ruled on one sheet, in which each receipt goes into the week the customer is likely to pay and each payment into the week it leaves.

The page takes six steps in a fixed order. The first is the opening balance, counted on a Friday across the bank account, the mobile money wallet and cash in hand, less every cheque issued and not yet cleared. The second lists receipts, placing each unpaid invoice in the week the customer will actually pay, taken from a habit table of past payments. Cash sales use the lowest of the last four weekly totals rather than the average, because the average includes the best week. The third lists the payments that cannot move: wages, rent, statutory dues and loan instalments first, then suppliers by due date, then the owner's own fixed wage. The fourth works the balances. The fifth marks the lowest week and the first week below zero, which are often different weeks. The sixth writes one action beside the first negative week while it still costs a telephone call.

The habit table is the core of the method. In the book's sample of eight paid invoices the median customer was promised payment in 30 days and paid in 31.5 days, and the slowest customer took 52 days. The book's point is that a due date on an invoice is a promise, and the date in the bank is a habit. The page should be built from the second. The sample is small, and the book presents it as an illustration of method, not as a benchmark for any trade.

Six weeks is the horizon the book recommends for most small firms. It is longer than the median payment delay, so most invoices already issued fall inside the page, and it is short enough to be kept honest. Invoices that land beyond the sixth week are not dropped. They are written below the page with the customer's name and brought in when they come into range. The page extends to thirteen weeks where a firm carries seasonal stock, faces a large tax date or pays a quarterly loan instalment. Each Friday the past week is dropped and a new sixth week is added, so the page rolls forward rather than being rebuilt each month.

The worked page closes week 4 at a shortfall of TZS 1.0 million, computed as a TZS 14.0 million opening balance, TZS 4.0 million coming in and TZS 19.0 million going out, all in the book's worked units. Two telephone calls close the week in the book's account: one asks a customer for a payment day, and one moves a supplier date by agreement. A second copy of the page, with the largest receipt moved two weeks later, shows dependence on one customer.

The book then lists five mistakes that the page catches, each shown through an owner's case. One is a customer who always pays late, such as a customer who takes 52 days against a promised 30, which the page catches by using the days actually taken. Another is a wage paid from cash meant for another payment, and a third is school fees taken from the till with nothing written down. The statutory dates are the part most often missed. PAYE and the Skills and Development Levy fall by the 7th of the following month, NSSF, PSSSF, NHIF and WCF at month end, and VAT by the 20th of the following month for registered firms. Each date is marked Verify in the book's figures table, since rates and dates move with the Finance Act each 1 July.

Closing a gap and what the close costs

Tab II takes the first negative week from the page and prices each way of closing it. The book ranks five levers by cost, and the cheapest is almost never a loan. The first lever is to bring a receipt forward, which costs goodwill and perhaps a late fee if the contract allows one. The second is to move a payment by agreement, asking the supplier for a date rather than for money in general, and asking before the due date and not after it. The third is to take a deposit on a new order, so that a 50% deposit on an order of TZS 8.0 million gives TZS 4.0 million at the start, at the cost of the sale if the customer refuses. The fourth is an early-payment discount, and the fifth is borrowing.

The early-payment discount is the book's clearest piece of arithmetic. A discount of 2 for 7, net 30, gives 2% off for payment within seven days, with the balance otherwise due at thirty days. Paying 23 days early for 2% off is worth about 32.4% a year on a simple basis, computed as 0.02 divided by 0.98, multiplied by 365 and divided by 23. The rule is that the yearly cost of any discount is worked out first and judged afterwards. A seller who offers the discount as a habit therefore pays the customer far more than a lender would charge for the same cash.

The loan section compares three quotes for the same TZS 5,000,000 borrowed over twelve months, as illustrated in the book. An 18% rate on a reducing balance has total interest of TZS 500,800 and a true cost of about 19.6% a year. The same 18% charged flat on the amount first borrowed gives total interest of TZS 900,000 and a true cost of about 36.8%. A rate of 3.5% a month flat gives total interest of TZS 2,100,000 and a true cost of about 98.0% a year. The book's conclusion is that every monthly figure must be converted to a yearly cost before comparison, because a flat rate roughly doubles the figure the borrower sees on the screen.

Who regulates the lender is the next question. The book separates four kinds of provider. A commercial bank is a Tier 1 deposit-taker supervised by the Bank of Tanzania, and its deposits are insured up to TZS 7,500,000 per depositor (Deposit Insurance Board, Verify). A Tier 2 non-deposit lender is also supervised by the Bank of Tanzania and takes no deposits. A SACCOS is supervised by the Tanzania Cooperative Development Commission, and its deposits are not covered by the Deposit Insurance Board. Savings groups and VICOBA are supervised by local government authorities. The register of licensed Tier 2 lenders stood at 2,938, with 19 approved digital lending platforms, at December 2025 (Bank of Tanzania, Verify). The reader should check the licence on the Bank of Tanzania register before signing anything.

The book sets out eight checks before signing: the licence, the rate basis, the fees, the repayment schedule, the security, the guarantors, the credit record and the complaints route. A guarantor is told before signing what the guarantee means, because a guarantor who has not been asked is treated as a debtor who has not been told. Credit reports are available once a year without charge, and negative data is held for six years from the date a loan is cleared (Creditinfo Tanzania and Dun and Bradstreet Credit Bureau Tanzania, Verify). A complaint goes first to the lender with a reference number, then to the Bank of Tanzania through SEMA NA BOT.

The book states that 3.5% a month is expected practice in the microfinance sector, set out in an official statement of 13 March 2025. It is not a legal ceiling, and the 2025 Regulations were not found to contain a price limit (Bank of Tanzania and the Microfinance Act, 2018, Verify). Limitation of contract claims is put at six years under the Law of Limitation Act, Cap. 89, a Mainland statute marked Verify. Any demand, guarantee or court step belongs to an advocate. The chapter's final rule is that the total repayable in shillings, with every fee included, is asked for in writing before a decision is made.

Price a job so that it pays

The price is built from the whole cost of the job: materials, labour, machine and power, delivery and a fixed overhead share. The overhead share is set once a year as a share of direct cost, applied to every job and reviewed each quarter. The price is then the total cost divided by one minus the margin wanted, and any turnover tax rate joins that divisor. VAT, where it applies, is added on top and is not part of the price the firm keeps. The book sets the formula out as a price sheet and asks the reader to work it for one real order before any other step.

The book separates markup from margin. A markup of 25% on cost earns a margin of 20% on the selling price, because a quarter of cost is a fifth of the price. The two words are not interchangeable, and a firm that prices at a 25% markup is earning a margin of 20%, not 25%. The worked order of 2,000 printed menus sets materials, labour and overhead first and the margin last, so that the number on the quotation is the result of the cost and not the starting point.

A discount is paid from profit. The book's example is a 5% discount on a sale carrying a 20% margin, which takes a quarter of the profit and needs about a third more orders to recover the same sum. Break-even is the fixed cost divided by the contribution share, and a turnover tax lowers that share. The worked lunch case, at TZS 9,000 a plate for 120 guests, shows a loss of TZS 40,540 after tax on each lunch (illustrative, computed from the inputs printed).

A quotation carries nine lines of terms, with a deposit first among them, and the margin is reviewed every quarter. The reason for writing the terms is that an unwritten price protects nothing. The book's rule is that a deposit buys the materials before the work is bought, and that finished work is invoiced on the day it is delivered. The price must also cover the wait. In the book's words, a price that does not cover the wait brings the gap back every month, and no lever closes it for good.

Collecting what is owed

Late money is a loan the firm never chose to make. Tab IV builds the routine that stops that loan being made by default. Five documents make a debt collectable: a quotation with terms, the order, the signed delivery note, the tax invoice or receipt, and a statement of account. A firm that holds all five has a debt that an advocate or a court can act on. A firm holding only a promise has, in the book's phrase, found out afterwards what the promise is worth.

Credit limits are set before supply rather than after the debt has grown. The book's rule is one week of a small buyer's purchases, or the unpaid work the page can carry for 30 days. The cash that a customer holds is worked as sales multiplied by days to pay and divided by 365. In the book's example, a council holding TZS 106.8 million of TZS 126.2 million tied up with three customers shows the scale of a slow public buyer. Two cases illustrate the waits. A builder's council certificate was paid on day 130 after it was billed, and a second invoice, Dorothy's, was paid on day 71. Neither was cheated, but both were made poorer by a wait that had never been priced.

The late-payer ladder runs in the same order for every customer. The first step falls three days before the due date. Reminders follow on day 1, day 7, day 14 and day 30 after it, and each step is logged. The steps move from a reminder to a call, then a formal letter, then a visit, and finally a stop in supply. The aim is that the same steps are taken in the same order every time, so that a call becomes a routine and not a fresh decision each time. When the ladder fails, supply stops in writing, and the file goes to an advocate within the six-year limitation period (Verify).

Public buyers need a line of their own in the page. The book notes that VAT withholding by designated agents runs at 3% of the price on goods and 6% on services, with remittance within 10 days after the period (TRA and the Finance Act 2026 text, Verify). On a TZS 10 million service invoice the 6% withholding holds back TZS 600,000 (illustrative). Payment terms and interest on late payment for public bodies are Unknown in the sources read. The Public Procurement Act was read to section 69(10), and the procuring entity and the Public Procurement Regulatory Authority are to be asked in writing before a contract is priced.

The dates that leave regardless

Tab V turns to cash that leaves on dates other people set. The sentence most often heard when the first notice arrives is that the staff were paid and the other dues were not expected. A wage slip shows the employee's side, and the employer's side sits elsewhere. The book's example is a gross wage of TZS 800,000. For a Mainland employer with ten or more employees, the employer adds TZS 80,000 of NSSF, TZS 4,000 of WCF and TZS 28,000 of skills levy, so that the wage costs TZS 912,000 a month. Of that sum, TZS 252,000 goes to the State and the funds in the book's worked split, with the NSSF and PAYE rates marked Verify. Below ten employees the same wage costs TZS 884,000 (TRA, Verify).

The dates are the core of the tab. PAYE and the Skills and Development Levy fall by the 7th of the following month. NSSF, PSSSF, NHIF and WCF fall at month end, which is a different date and must be placed on the page separately. VAT for registered firms falls by the 20th of the following month. Provisional income tax falls in instalments by accounting period, on 31 March, 30 June, 30 September and 31 December (TRA, Verify). The book sets a weekly reserve of TZS 581,538 for ten such wages, computed as an annual cost spread over 52 weeks, as an illustration of how a reserve is sized.

The Mainland skills levy threshold has been ten employees since July 2023, having been four before. The book notes that this threshold is often printed wrongly. Mainland VAT is 18%, with a registration threshold of TZS 200 million a year and a return due by the 20th of the following month (TRA, Verify). Presumptive tax at 4% of turnover applies to resident individuals with a turnover of TZS 11 million to 200 million a year, as reported in commentary on the Finance Act 2026, whose text was not read. The payment dates for presumptive tax are Unknown, and the first-year relief reported for new individual businesses is Unknown in its eligibility.

Zanzibar is kept separate throughout the tab. The book prints Zanzibar figures only where a source states them, and it marks them for confirmation with the Zanzibar Revenue Authority. They are VAT at 15%, with 18% for banking, postal, telecom, insurance and digital services, a registration threshold of TZS 100 million (PwC, 2026, Verify), and a skills levy of 4% from four employees (Verify). Whether the presumptive regime applies in Zanzibar is Unknown. Zanzibar employer contributions to social funds are Unknown, and no Mainland figure stands for them.

A VAT line is also a timing point, and the book treats it as one. VAT collected is the State's money and is not earned. The worked example shows TZS 3,600,000 charged less TZS 1,440,000 paid, leaving TZS 2,160,000 owed, which must not be spent. Tab V ends by placing each tax date on the Friday page beside the rent date, so that a tax falls on the page as expected, sized and already set aside.

Money that stays in the business

Tab VI turns to the money that sits inside the business: stock on the shelf, customers' balances and suppliers' terms. The book counts it in days. Stock days, plus the days customers take to pay, less the days suppliers allow, give the cash cycle. The worked hardware shop holds 54 days of stock and waits 24 days for customers, less 21 days of supplier credit, which gives a cash cycle of 57 days. In shillings the cycle ties up TZS 63.0 million, made up of 54.0 in stock and 30.0 owed by customers, less 21.0 owed to suppliers, in the book's worked units.

Growth makes the cycle bite harder. A 20% increase in sales in the same shop needs TZS 12.6 million more cash, moving the tied-up figure from 63.0 to 75.6 million in the worked units. Repaid out of extra profit, that cash takes about 36 weeks to return, or about 90 weeks once a clerk is paid. The tab opens with the point that stock and credit grow before the extra profit arrives. An owner who has been short of cash for years is inclined to turn any balance into something visible, and the book does not oppose the purchase. Its test is the page.

Supplier terms have a price. Two per cent for seven days against thirty is about 32.4% a year, and sixty days at a price 3% higher is about 36.5% a year, both on the simple basis used in the book. A bulk purchase is run on the weekly page in at least two versions, with a stress test. The book's seed example pays 42.0 in one payment, or 21.0 and 21.0 in two, and the second version moves the lowest week from a shortfall of 5.5 to a balance of 15.5 in the worked table's millions. Staging the payment, not only the purchase, is the lever.

The tab's measure is therefore the lowest week and not the price. A purchase is written on the weekly page, stressed and judged by the shillings left in the account in its worst week. The book's rule is that a larger order is a plan with a date, and a plan with a date can be checked against the bank on the following Friday.

One page, five trades

Tab VII fits the page to the trade. The book's point is that the page stays the same while the rows, the length and the week to watch change by trade. A shop watches the week after its season order, when stock has been bought and receipts are thin. A workshop watches the deposit against the materials it must buy before it builds. A professional firm watches the gap between delivery and invoice. A cooperative watches the thin week before the buyer pays, and a contractor watches its unpaid certificates. Each trade has one trap, one change that removes the trap and one week that a lender reads first.

The members of the circle in the tab are a shop, a joinery workshop, a practice, a cooperative and a contractor. Each has built the page of Tab I and used it through Tab V, and each has felt that the examples belonged to someone else. In the joinery case, a 30% deposit on a school order covers 63% of the materials. In the practice, recurring income covers 84% of the payroll. The shop's lowest week is TZS 3.9 million when it holds two lots of stock, against a shortfall of TZS 4.1 million with one order. The figures are composite and illustrative, and the book shows the sums beside them.

Records, and the people who read the page

Tab VIII makes the page true. The opening figure is only as good as the records behind it, so the book sets out a day book that records every shilling on the day it moves, with the pocket it moved through. Money is kept in three pockets: the firm's bank account, the business mobile wallet and the cash tin. The owner's money meets the firm's only on a written line. Each pocket is reconciled line by line each week, because a net difference can hide far larger errors. In the book's case a net difference of TZS 45,500 hid TZS 1,565,500 of errors across offsetting lines.

The reconciliation is walked like a surveyor's traverse. The owner takes the book back to the bank, the wallet and the tin, and if the figures meet, the opening balance on the page is true. If they do not, the error is found while it is still a week old. The book sets six tests in a fixed order to find a difference: tick each line, look for the exact amount and half of it, halve the amount, divide by nine, check the edges of the period, and ask the people concerned. A difference that nobody can explain is written as unexplained and carried, and never filed as a sundry item. Some money never enters the firm's pockets, so unpaid invoices are confirmed with clients every month.

Tab IX carries the page to people who lend, invest or govern. The book says that all three ask three questions, phrased differently: whether the firm can see ahead, whether it acts when it sees a problem, and whether what it says is true. The page answers the first and second, and a run of pairs answers the third. Each Friday the firm predicted a closing balance, and each following Friday the bank showed the real figure. Six such pairs, side by side, are offered as the strongest evidence a small firm can give that its numbers mean something. A negative week with no action is treated as a warning heard and not answered. An action that names a person, an amount and a date is treated as evidence of management.

Readers lean differently. A lender leans on the repayment: which week it is paid from and what cushion remains after it. The book's loan example is TZS 20,000,000 at 18% on a reducing balance over twelve months, with a total repayable of TZS 22,003,198 and twelve instalments of about TZS 1,833,600 (illustrative, computed). The lender is judged by the total repayable in shillings, and each instalment goes into the week it falls. Guarantees, security and any disputed demand go to an advocate before anything is signed or answered.

The Friday Page: five owners and one habit

The Friday Page is a story companion to Book 1. Its owners are composites with invented surnames, and their firms are described and never named. The central case is Neema Mushi, who runs a printing shop in Kariakoo, Dar es Salaam. On the last Friday of March her shop has delivered its largest order, an invoice of TZS 38,000,000 for a hotel group that pays in sixty days, while her account holds TZS 1,200,000. The order cost TZS 32,568,000 to make, and the paper was bought in March. Between those dates the cash gap is visible six weeks early on a single sheet.

Book 2 follows the same six weeks through seven numbered Friday chapters, each adding one tool from The Friday Folder. Friday 1 rules the page and rings the lowest week with one action beside it. Friday 2 builds a price from cost with a floor beneath it. Friday 3 separates the stall's cash from the household's and writes down what is owed. Friday 4 lays out twelve months to find the reserve a seasonal business must hold before it buys anything large. Friday 5 prices the wait on a large, slow customer, using sales multiplied by days to pay, divided by 365, and multiplied by the rate paid on borrowed money. Friday 6 shows a cost shock reaching five towns and orders four calls, the cut first and the wage last. Friday 7 takes the page to a treasurer and a bank.

The five owners are placed in five towns: a printer in Kariakoo, a fish trader in Mwanza called Mama Zawadi, a lodge owner in Arusha facing a rival's pool across the valley, a builder in Dodoma with forty workers and a council that pays late, and the treasurer of a maize and bean cooperative in Mbeya. Their scale runs from a stall selling TZS 640,000 of fish in a week to a builder selling TZS 480,000,000 in a year. Each learns the page first and then the lesson that belongs to their trade, such as the reserve for a seasonal lodge, the wait on a council contract or a repayment timed to the harvest.

Two calculations recur through the book. The first is that a customer's wait is a loan. A council holding TZS 300,000,000 of sales over 130 days ties up TZS 106.8 million, computed as 300,000,000 multiplied by 130 and divided by 365, as worked in the book. The second is that a price carries a floor. The price is built from cost, divided by one minus the margin, and held above a floor fixed before the telephone rings, so that a rival's lower sign is answered with a figure and not with a discount.

The rules are kept apart in a section called The Rules Behind the Stories. Each rate is given with its issuer, the date it was read and its status of Settled, Verify or Unknown, and a figure that could not be confirmed is left unprinted. The page is the habit that all five stories share. It is ruled once in about forty minutes and then carried forward each Friday. The opening proverb, haraka haraka haina baraka, meaning that haste brings no blessing, makes the same point about speed in a shorter form.

Mainland Tanzania and Zanzibar

The two books apply the same method to Mainland Tanzania and treat Zanzibar as a separate set of rules. The Mainland figures are drawn from the Tanzania Revenue Authority, the Ministry of Finance, the Bank of Tanzania and the Ministry of Labour, and most are marked Verify. A Mainland figure does not stand for the islands. Where no Zanzibar source was found, the book says so and sends the reader to the Zanzibar Revenue Authority or to the fund concerned.

The Zanzibar items printed in Book 1 are VAT at 15%, a registration threshold of TZS 100 million and a skills levy of 4% from four employees. Each is marked to confirm with the Zanzibar Revenue Authority. The presumptive tax regime in Zanzibar is Unknown. A reader running a business in Zanzibar should take every figure from Zanzibar's own authorities and use the Mainland figures in these books only for comparison.

Figures to confirm, and the order of reading

Book 1 closes with a Figures to Confirm table that gathers the rates, dates, thresholds and open questions of every tab on one page, with the office to ask and a status for each. Settled marks a rule of structure that rarely moves. Verify marks a rate or date that moves, usually each 1 July, which has been read from the issuer or from commentary and must be confirmed again before use. Unknown marks a point the sources did not settle, for which the book prints no figure. The table gives positions as at 12 October 2026, while the scope note on the cover states 10 October 2026. The reader should note the difference.

The status of several items needs attention before use. The Finance Act 2026, assented on 30 June 2026, is read through commentary from Clyde & Co and EY, because the Act text was not read. Items marked Unknown include the payment dates for presumptive tax, the rate of income tax withholding on services, the VAT tax point on credit sales, the statutory period for keeping records and the late-payment penalties on payroll dues. A reader should ask the named office in writing before relying on any of them.

The reading order is practical. A reader with a fresh plan starts with Tab I and builds the six-week page before reading further, since every later tab reuses it. A reader with a negative week goes to Tab II, a reader setting a price to Tab III, a reader with unpaid debts to Tab IV and a reader with statutory dues to Tab V. Tabs VI to IX serve growth, trades, records and lenders. A reader who wants a story first can begin with Friday 1 of The Friday Page.

The series is general information based on the official sources at the dates stated. It is not legal, tax, accounting or investment advice. Where a rate differs from the issuer's current notice, the notice prevails, and the difference should be written beside the relevant tab so that the sum can be rerun. Disputes, demands, guarantees and court steps belong to an advocate. Registrations, tax filings and licences belong to the issuing authority, and the reader should confirm each one with that authority before acting.

Kiswahili Originals is the fifth series in the Zatra Handbooks. It is published in Kiswahili and English, and this summary works from the English editions, which carry the same figures. The five titles are Gone by September: Where the Retirement Payout Went (Handbook 117), The Good Wife Idea, and How It Was Twisted in Africa (Handbook 118), And Let It Be You, Their Commander (Handbook 119), Start Raising Capital: Not Only for Parties and Troubles (Handbook 120) and Sinza Was Farmland and Tegeta Was Bush (Handbook 121). Each starts from a decision that a household, a group or a buyer must make, and works it in Tanzanian shillings with the loss printed beside the gain. The books state their positions at 10 October 2026. Their legal and regulatory references are to Mainland Tanzania. Handbooks 117 and 118 say directly that Zanzibar has its own regime and is not covered, and no figure from the series should be applied to Zanzibar without a separate check. Nothing in the series is legal, tax or investment advice. Its figures are illustrative, and every rate, threshold and date should be confirmed with the issuer named in the book before anyone acts.

What the five books cover

The five books share one structure. Each opens with a short statement of who it is for, the three pains that reader brings and the three gains the book promises. Each then follows illustrative households, groups or buyers through a sequence of steps, nine or ten in each book, and closes with a one-page tool the reader can complete. Each also names an open starter: a first action for the coming week, such as writing a one-page plan or listing the services within five kilometres of a plot. The fuller kits, worksheets and templates that follow are described as available to members of the Zatra membership at zatra.co/guides.

The books are grouped by the decision they address. Two concern money in a household: the lump sum from a pension or end-of-service benefit (Handbook 117), and the sharing of decisions and accounts between a husband and a wife (Handbook 118). Two concern groups: the building of a producing group with written ownership and pay (Handbook 119), and the pooling of savings into capital that earns (Handbook 120). The fifth concerns the purchase of land on the edge of Dar es Salaam (Handbook 121). The books stand alone, and each says so. Handbook 117 is the natural starting point for a reader facing a payout, because the first task it sets is to list what the money must cover and the date each need falls due. A reader considering a pooled group should read Handbook 120 before any money is collected, because its tests of the purchase and of the offer come before any money moves.

Every book sets its figures out in the same way. A figure is the book's own arithmetic from stated inputs, and where it rests on a published source, the book names the issuer and the date read. Where a figure could not be found, the book says so and names the office to ask, rather than offering an estimate. Each illustrative case carries a loss case beside its gain: a payout that stops early, a fund that falls, a group that runs short, a buyer who pays before the title is checked. The books also repeat the same caution, that an investor can lose money. The series therefore works as a set of tests to apply to a decision, and it does not recommend a product, a provider or a purchase.

The Kiswahili and English editions are meant to carry the same figures, and the English text is the source for this summary. Both state the same position date of 10 October 2026. Readers in either language should use the same checks and confirm the same issuers.

Lump sums and the household: Handbooks 117 and 118

Handbook 117 treats a final payment, such as a pension lump sum or end-of-service benefit (mafao), as a problem of timing. Its teacher received TZS 36,000,000, and the ledger follows the money from July. By the end of September in the illustrative ledger, the balance stood at TZS 1,620,000, or 4.5% of the payout. Thirteen lines account for the rest: six requests from others worth TZS 15,100,000 (41.9%), four personal purchases of TZS 9,700,000, three debts of TZS 4,800,000, a scheme deposit of TZS 4,000,000 and a household gap of TZS 780,000. The book's point is that no single line was foolish, but none was set against the others. The six requests alone equalled 126 months of the planned monthly gap.

The second half of the book gives the same money jobs. The cover list totals TZS 17,640,000, or 49.0% of the payout: debts of TZS 4,800,000 paid in a written order, a roof of TZS 3,600,000 on a written quote, a reserve of six months at TZS 640,000 a month (TZS 3,840,000) and a year-one account of TZS 5,400,000. The remaining TZS 18,360,000 is divided into an income pot of TZS 17,000,000 and a learning pot of TZS 1,360,000. The income pot pays a gap of TZS 120,000 a month. Tested in five cases, it lasts 141 months with no return, 244 months at the assumed 6%, 165 months if the gap rises 4% a year, 183 months after a 15% fall in the first year, and 137 months in the loss case with a rising gap. The book calls every rate an arithmetic label rather than a forecast.

The book then runs the method at two other sizes. A driver's payout of TZS 9,000,000 yields a bridge pot of TZS 3,400,000 that pays 34 months with no return, 37 months at 6% and 31 months in the loss case. A clinic matron's payout of TZS 40,000,000 holds an income pot of TZS 14,100,000 and a growth pot of TZS 11,900,000. In a 25% fall the growth pot loses TZS 2,975,000, or 7.4% of the payout, while the income pot still pays 242 months at the assumed 6%. The request envelope in the teacher's plan is 5% of the payout, TZS 1,800,000, with a cap of TZS 300,000 for any one request, and the book holds that the envelope answers six requests without breaking the plan. A family meeting held before the payout moves the argument to a time when no money is on the table.

The quick-scheme chapter is concrete. A deposit of TZS 4,000,000 promised TZS 8,000,000 in thirty days, and a doubling each month for twelve months, would on the book's arithmetic owe TZS 16,384,000,000 at the end of the chain. A chain of three needs 59,049 people at the tenth level and 43,046,721 at the sixteenth. The book cites the Bank of Tanzania and the CMSA as having said that an unlicensed operator is outside the law and that promoting or joining pyramid schemes is an offence. Its scheme check asks five questions: whether the return is fixed and large, whether there is a deadline, whether a trusted name is attached, whether the reader is asked to pay now, and whether the reader is asked to bring others in.

Handbook 118 reads a related problem through the household partnership. Its starting claim is that the good wife idea began as a description of productive work. Chapter 1 cites the Proverbs 31 passage on a woman who works wool and flax, trades, buys a field and sells linen, and sorts its verbs into seven kinds of work, only one of them house work in the narrow sense. The book says that nothing in the sources it reads states that a wife has no say over resources. Chapter 2 names four doors through which the partnership narrowed: land, inputs and advice, paid work and the breadwinner, and paper such as registers and titles that carry one name. No figure for Tanzania was found for these doors, no door is dated, and the book blames no sex, tribe, faith or custom.

Chapter 4 costs the narrowing in four households: a highland farm, a town shop, a city home with rooms to rent, and a widow's five months. On the farm the wife works 70 hours a week against 56 for her husband. Her 40 unpaid home and care hours are valued at TZS 2,080,000 a year at an assumed TZS 1,000 an hour, or 52.8% of a household income of TZS 3,940,000. Inputs on the wrong plot cost TZS 160,000 a season, or 7.5% of net. The city home was owed TZS 5,760,000 in rent, collected TZS 5,200,000 and lost track of TZS 560,000, or 9.7%. The widow's case shows what a missing name costs: a funeral of TZS 1,800,000 less contributions of TZS 1,100,000 leaves TZS 700,000, and with household costs and a harvest of TZS 600,000 the gap is TZS 1,200,000, or 66.7% of the need.

Chapters 5 to 7 build the partnership at home, in a shop and on a farm, and chapters 8 to 10 move from principle to practice. Five skills, namely the cash book, pricing and margin, phone, wallet and bank, reading a paper and supplier talk, come to 24 hours, or three hours a week over eight weeks, and the book holds that the partner who knows a skill teaches it. Twenty household decisions are sorted into three accounts. In the model, on a household income of TZS 2,880,000 a month, the husband alone holds 12 decisions, the wife 3, and five are joint; the day-90 target is 6, 6 and 8. A second signature on the shop wallet and a shared supplier list would have cut a loss of TZS 1,200,000 to TZS 175,000. Resources then move in four rungs, use, manage, sign and own a share, each written down at a Sunday meeting before the next.

The law is read in part and the book says so. It reads the Law of Marriage Act, 1971, on the property of a married woman (sections 56 and 58), on presumptions where property is held in one name or two (section 60), on division of property by the contributions of each spouse in money, property or work (section 114), and on the consent of a spouse before the matrimonial home is sold, leased or mortgaged (section 59). It reads the Village Land Act, 1999, section 3(2), only through a journal paper, and says so. The consent rule for a mortgage under the land laws was not read. Each family's position is for an advocate, and the land office and the lender can say what the land laws require.

Groups that produce and raise capital: Handbooks 119 and 120

Handbook 119 is written for a person who can already do the work and now wants to bring others into it, whether a tailor with a machine and a waiting list, a teacher with savings and colleagues, or a farmer with a press and neighbours. Its central figure is the jemedari, a leader who goes in front by consent and carries six loads: purpose, money, records, people, risk and the handover. The book opens with a loss. Ten friends pooled TZS 5,000,000 with nothing written and, after 14 months and a forced sale, received back TZS 2,400,000. Each lost TZS 260,000, or 52.0% of the pool. The book's answer is a sequence of nine stages: choose one productive activity, hold a first meeting, write who owns what, pay people fairly, keep the books open, deal with a member who stops pulling, grow from three members to thirty, and step back.

The choice of activity is made before any meeting. Eight activities are scored, and one is chosen by what three people can run together. The first meeting is timed at 90 minutes in six items of 10, 15, 20, 15, 20 and 10 minutes, and it settles seven decisions before any money moves. The chair, the minute-taker and the timekeeper are three different jobs. In the ward tailor's case, the meeting fixes capital at TZS 3,000,000 (TZS 1,500,000, TZS 900,000 and TZS 600,000), shares of 50%, 30% and 20%, and a work commitment of 660 days a year at a day rate of TZS 12,000. Ownership follows what each member puts in, valued by agreement and written in a register before money moves. A further case shows the cost of skipping this step: a mill with one name on the receipt and no page showing whose it was lost TZS 2,600,000, or 52.0% of its pool.

Pay follows a fixed order: the leader's allowance, work at the day rate, the reserve, and then the dividend by shares. In the tailors' good year the three members take TZS 12,252,000 and TZS 348,000 stays in the reserve. In a short year the day is paid at TZS 6,484 and no dividend is paid. The clause bank in Part F gives twenty-four clauses in plain words, and the one-page charter carries twenty numbered lines. The book leaves the choice of legal form to the registrar or a lawyer. Growth is handled in stages. The oil press group grows from 3 members at TZS 1,000,000 a share to 30 members at TZS 2,000,000 a share. The founders' proportion falls from 100 to 10.0%, while each founder's share rises in value from TZS 1,000,000 to TZS 2,000,000.

A founder hands over the lead without losing the business when the handover has a date, a pack, a six-week absence test and an allowance that follows the office. In the oil press model the allowance is TZS 4,800,000 a year, the founder's week falls from 14 hours to 6, and the founder's share, worth TZS 2,000,000, stays with the founder. The reader is thereby shown that stepping back is a planned transfer with a price, not a loss of control.

Handbook 120 starts from a strength that most families already have. A wedding sum of TZS 13,200,000 was raised in 42 days through a michango, with TZS 14,400,000 pledged and 91.6% collected. A funeral sum of TZS 2,400,000 was raised in three days. The book argues that the same method can serve capital, and it sets twelve michango beside twelve capital versions. Ten of the twelve have capital twins that earn. Two, a funeral fund and a hospital fund, are insurance and a revolving loan and not income. Four reasons keep capital waiting, among them shame, the lack of a date and the fact that nothing is visible to see.

The purchase is tested before any shilling moves. The Purchase Case Sheet sets the price, the setup and a reserve against the base net and a bad-year case. The three-wheeler example uses TZS 11,800,000 for the vehicle, TZS 1,200,000 for setup and TZS 1,400,000 as a reserve, a target of TZS 14,400,000 and a base net of TZS 3,600,000. An equal monthly sum of TZS 100,000 asks a rider for 22.2% of income, a nurse for 7.6% and an owner for 1.6%, so the circle of forty uses units of TZS 50,000, and members take one, two or four. The circle holds 800 shares of TZS 50,000 in a register. The first month of the Dodoma circle closed at TZS 3,758,000 after TZS 4,000,000 fell due, TZS 200,000 was missed, and TZS 18,000 of charges and TZS 24,000 of stationery were paid. A one-page monthly record with eight lines and three signatures keeps the book, and three different people keep the page.

Terms come next, and the Dodoma roof shows the cost of having none. The roof needed TZS 3,000,000 against a reserve of TZS 1,500,000, so a call of TZS 1,500,000 fell on forty members, and an evening was lost on whether to charge TZS 37,500 each or to charge by shares. The book's fourteen-clause agreement covers purpose, members, dates, lateness, money, signatures, votes, books, change of mind, the operator, title, reserve and sharing, heirs, exit and dissolution. A fair share-out is one division applied to every share. In the village bank, a base fund of TZS 19,920,000 gives TZS 5,804 a share and a gain of TZS 167,232 in the book's case, while a bad fund of TZS 16,520,000 gives TZS 4,813 a share and a loss of TZS 38,896.

A member who needs the money back is handled by a set of written doors. In the Dodoma case, a nurse with 20 shares receives TZS 600,000 less a 5% fee of TZS 30,000, which is TZS 570,000 before the purchase. After it, at the value of the circle at cost of TZS 41,500,000, she receives TZS 1,037,500, which the reserve of TZS 1,500,000 covers. At a fallen value of TZS 36,000,000 she receives TZS 937,500, and after a 10% discount TZS 843,750, paid in four instalments. A sale to another member costs the circle nothing, and hardship is met by a pause or a loan from the reserve.

The book closes on danger. A circle in a town on the lake promised 20% a month, or 240% a year, on contributions of TZS 500,000. Its payouts passed the money coming in during month 4, at TZS 5,100,000 against TZS 4,500,000, and its cash ran short in month 8 by TZS 3,900,000 on TZS 33,500,000 contributed. Of 67 members, 25 gained, 14 broke even and 28 lost, and the members who joined in month 6 each lost TZS 400,000. The ten signs and three questions the book sets out test an offer before any money moves. The final chapter explains when a circle should change form, whether to stay a circle, register as a community microfinance group, become a cooperative or form a company. The texts it read name groups of 10 to 50 members with a common bond for community microfinance, and 20 persons for a savings and credit society and 10 for other primary societies under the Cooperative Societies Act. The company minimum is left to BRELA.

Buying land at the edge of Dar es Salaam: Handbook 121

Handbook 121 treats the growth of Dar es Salaam as a set of dated steps. The built edge was about 5 kilometres from the centre until the 1940s, about 10 kilometres in the 1950s and 1960s, nearly 20 in the 1970s and 1980s and nearly 30 from the 1990s. A 30-kilometre edge therefore holds 36 times the area of a 5-kilometre one. The population was 851,522 in 1978 and 5,383,728 in 2022. The book reads roads as the factor that decides where growth lands, and it notes that much of the land at the edge is farmland held under customary tenure and sold through brokers, so the question of who may sign comes before the question of price. The saying behind the title, that Sinza was farmland and Tegeta was bush, is a family recollection, and the book treats it as a memory rather than a record. What the sources record is a road and a date. A Nairobi counter-case, in which a city rebuilt its core rather than spreading outward, is used to test the model.

Area is judged through ten signs, each scored 0, 1 or 2 on one sheet, for a total out of 20. The total sorts an area into the early band (0 to 7), the forming band (8 to 13) or the formed band (14 to 20). A sign counts only when it is built and working. The cases price each band at assumed values of TZS 6,000, TZS 14,000 and TZS 40,000 a square metre, and the book labels these as assumed prices. On those assumptions TZS 5,000,000 buys 500 square metres in an early area, 210 in a forming area and 70 in a formed one, once a tenth is kept back. The size falls faster than the price rises, because survey, title and lawyer costs of TZS 1,300,000, as the book assumes them, fall on any plot whatever its size. Three buyers decide at three budgets: a minibus driver and his wife with TZS 5,000,000, two teachers with TZS 12,000,000, and a company owner with TZS 40,000,000. Twelve Edge Files score twelve kinds of area.

The title and the seller are the heart of the book. A sale should proceed only when six checks agree: the kind of right on the paper, the plan matched to the beacons, the official search, the person who signs, the village step and the lawyer's reading. A seller's paper shows what the seller says, and a stamp shows that a payment was seen. Neither shows ownership. Case D lost TZS 2,400,000 when a second buyer paid the same seller. For Case T, paying in full before any check would lose TZS 8,400,000, failing at the third rung would lose TZS 3,360,000, and failing at the second would lose TZS 840,000.

Payment follows four rungs of 10%, 30%, 50% and 10%, each opened by a check, so that the amount at risk grows only as proof grows. For Case T the rungs are TZS 840,000, TZS 2,520,000, TZS 4,200,000 and TZS 840,000. A seller who fails at the second check costs TZS 840,000 instead of TZS 8,400,000. The book says the percentages are its own design, that the long instalment can be tied to the rungs, and that a retention of TZS 840,000 waits for the registered title. Building follows the same logic. A board of TZS 150,000, a fence at TZS 20,000 a metre, a caretaker's room of 12 square metres at TZS 4,800,000 and a house of 60 square metres at TZS 27,000,000 are taken in that order as proof arrives. A room built before title and then lost costs TZS 4,800,000, and an encroached strip of 72 square metres costs TZS 1,008,000.

The losses are priced for the three buyers. The book prints five: TZS 8,400,000 for paying in full, TZS 4,800,000 for counting a plan as a road, TZS 2,400,000 for a second buyer, TZS 4,800,000 for a room built before title, and TZS 1,008,000 for a boundary nobody marked. Each loss has a matching check, from the ladder to the beacons. Even the careful buyer carries a flat-price loss, which the book states for each case.

The plot plan is one page in three parts: the plot, the money and the next three dates, with the base-year gain and the flat-year loss on the same line. Case D's outlay is TZS 3,849,500, leaving TZS 1,150,500. Case T's outlay is TZS 10,224,500, leaving TZS 1,775,500. The book states that the plan does not know the price, the signs that will be kept up or the people in the file, and it keeps the flat-year loss on the page so that the reader sees it.

How to use the figures, and what to confirm

Every figure in the series is the book's own arithmetic from stated inputs, and it should be read as a worked example, not as a quotation from an issuer. Before any step is taken, the reader should confirm the items below with the issuer named in the relevant book. The list is not complete, and the books name further offices where they apply.

Handbook 117 quotes a fund's page for retirement at 60, voluntary retirement from 55, 180 months of contributions and factors of 0.33 and 0.67. It records that newspapers reported shares of 25 and 75 and later 33 and 67. The reader should confirm the current factors with the fund, and with the NSSF where it applies. The tax position on a lump sum should be confirmed with the Tanzania Revenue Authority. The book also cites insurance and securities references from 2025 and 2026, which the reader should check directly with the issuers named.

Handbook 118 reads the Law of Marriage Act, 1971, from a consolidated copy whose date it does not show, and says that the current text should be checked against the official copy. The Village Land Act, 1999, is known only through a journal paper. The consent rule for a mortgage under the land laws was not read. These points go to an advocate, to the district land office, to the ward or village office, or to the lender, as the book directs. Each family's position under the law is for an advocate.

Handbooks 119 and 120 leave several points open. The legal form of a group is for the registrar or a lawyer. The community microfinance thresholds of 10 to 50 members were read in a text summary of the gazette on 13 October 2026, and the book notes that Bank of Tanzania amendments must be checked. The Cooperative Societies Act thresholds, Act No. 6 of 2013, sections 20 and 29 to 31, were read on TanzLII on the same date and must be confirmed with the Registrar of Cooperative Societies. The company minimum for members and directors is not printed, because it was read second hand, and BRELA should be asked before any meeting is planned. The Companies Act, Cap. 212, is likewise to be confirmed with BRELA.

Handbook 121's prices, title costs and build costs are assumptions, not market data. Before a plot is bought, the reader should confirm current prices with local sellers and brokers, check the title with the land registry, have a survey checked by a licensed surveyor, and have the legal reading done by an advocate. The prices in the series are the book's assumptions, and no figure from it should be used as a valuation.

The series gives no recommendation on any provider, product or purchase. Its value lies in the tests it sets: a payout listed against its dated needs, a household's decisions counted in writing, a group's rules agreed before money moves, a scheme checked for its arithmetic and its licence, and a plot checked for title before a shilling is paid. The reader should begin with the open starter in the book that matches the decision, and let each figure be tested against the reader's own household or group before any step is taken.

Set Your House in Order is a series of six handbooks for a household or a small business owner who needs the money, the commitments, the papers and the people of a home or a firm put into an order that can be checked. Each book takes one decision: review a year, carry relatives without ruin, leave a former life, repair a wrong, check a deal before paying, or prepare a family for the day after a death. The books share one method. The record comes before memory, every figure carries its issuer and its date, and every worked case is labelled as illustrative. Two books draw on religious texts and one on Tanzanian tax. Mainland Tanzania and Zanzibar are kept apart throughout, and every legal or fee line that was not read at its issuer is marked for confirmation. This summary is general information. It is not legal, tax, accounting, financial or religious advice, and rates, thresholds and dates must be confirmed with the issuer named beside them before anyone acts.

What the series covers

The series has six titles, and they are not a ladder that must be climbed in order. Each answers a different question that a household or a small business meets. The Year-End Postmortem (Handbook 122) asks how to review a year honestly and plan the next one in a single weekend. From Dependants to Pillars (Handbook 123) asks how the family earner can give to relatives by a plan rather than by the ring of a phone. Feet in Zoar, Heart in Sodom (Handbook 124) asks how a person leaves a former life and actually arrives somewhere else. Metanoia and Teshuvah (Handbook 125) asks what real repentance requires, and it sets out the parts of a change that can be tested. Trust, Yes. But Verify (Handbook 126) asks how to check people, papers and deals before money moves. If I Die Today (Handbook 127) asks what a family will need on the day after a death, and how to have it ready within four weeks.

The books sort naturally into three pairs. Handbooks 122 and 126 are practical tools for a desk: one reviews the record of a year, the other tests a deal. Handbooks 123 and 127 concern the family and its money after a death or in the long run. Handbooks 124 and 125 concern personal change and the repair of a wrong. A reader with a single urgent decision can open the book that matches it. A reader planning the whole household will find that the same habits recur across the six: a dated record, a named owner, and a check that can show whether an action was taken.

The series does not promise a verdict on any faith, family or person. Its cases are composites, described by type rather than by name, and its figures are worked from stated inputs. Where a case records a date after the stated date of the book, the case is a projected worked example and not a record of events. Each book ends with a list of what it settles and a note of the sources that were read, which the reader should treat as the limit of what the book can claim.

The method all six books share

Each book opens with a short statement of what it settles, then sets out a sequence of sittings, chapters or stations, and closes with working papers, a glossary and a table of sources. The sources table is the most important back page. It records each instrument, its issuer, the date it was read and a status, which is one of read at the issuer, read second hand, partly read or not read. Where a line was not read at the issuer, the book says so and prints no figure from it. This keeps a reader from taking an estimate for a rule.

The books also share a rule on figures. Money is computed from the inputs printed beside it, and each computed total can be checked by the reader against the same inputs. Where a fee, rate or threshold is not confirmed, the book leaves the place blank and names the office to ask. The reader should expect this in every book of the series, since no handbook can settle a question that an issuer has not answered in a public text.

The tools in each book, whether sheets, forms, cards or ledgers, may be copied into the reader's own notebook and used inside the reader's own business with the source named. They may not be sold, bundled or re-branded. Each book states that it gives general information only, that no adviser relationship is created by reading it, and that a decision carrying real money or legal risk needs a qualified adviser, an advocate for a dispute, or the publisher's own advice before action.

Book 1: The Year-End Postmortem

The Year-End Postmortem is set over one weekend. Its thirteen timed sittings fill sixteen hours, from Friday 18:00 to Sunday 20:30, and the book begins by agreeing three rules before the first paper is taken out of the box: nobody is on trial, the record speaks before memory, and the clock decides when a sitting ends. The method borrows from engineering postmortems and from the army's after-action review. Both look for causes in systems, such as a missing check, a signal or a rule, and both refuse to end a review in a finding about a person's carelessness.

The illustrative household is a honey business in Tabora run by two adults. Its record, rebuilt from statements, messages and photographs, shows money in of USD 8,640 and money out of USD 8,820, a change of minus USD 180. That change reconciles to the cash balance, which was USD 420 on 1 December 2025 and USD 240 on 30 November 2026. The business also won a school contract and kept a savings habit through every month, and the book uses this to make its central point: a year usually holds gains and losses together, and memory rarely keeps both. In the household's own scoring, the adults' gut feeling put the year at 43 of 70, while the record put it at 36.

Three failures are traced with the five whys, and each is written up on one blameless page with the same headings: summary, impact, timeline, causes, what went well, where the household was lucky, and actions. An action is accepted only when it has an owner, a date and a check, which is a paper that will show whether it was done. The worked example is a bulk purchase of raw honey made on credit in August, which left USD 270 of stock unsold at the end of November and USD 330 still owed. The cost of the three failures is given as USD 551 in money and 120 hours. The book values the hours at the household's own rate, which brings the total to USD 762.

The weekend closes with a one-page plan for the next year, which has seven parts and fits on one side, four dated checkpoints, and a premortem. The premortem assumes that the coming year has failed and asks why, in silence and apart, before the reasons are written down. The reader finishes with the record filed, the dates copied to phones, one outside reader named, and the people affected told. Every figure in the household is labelled as illustrative and is computed from stated inputs, and the book makes no claim about any real business.

Book 2: From Dependants to Pillars

From Dependants to Pillars is written for the person whom a family calls first when money is needed. Its starting point is a map of giving. In the illustrative file, one earner's pay of USD 1,300 a month sent USD 7,200 in twelve months to nine names. That is 46.2 percent of everything the earner took home, and the oldest of those commitments had been running for 91 months. The pay was already short of what it was asked to carry, so the gap was borrowed, in the file USD 1,200 from a staff society after 24 months. The book's first claim is that support given in small transfers is invisible until it is written down as names, amounts, dates and months run.

The second step sorts each transfer into one of three kinds: an emergency, a duty or a habit. The sort matters because each kind needs a different reply. An emergency has an end, a duty has a standing claim that the earner may choose to keep, and a habit is a payment that continues only because nobody has stopped it. Once the sort is done, the earner sets a giving line from their own floor. The floor is what the earner's own costs, debts and reserve require, and the line is what remains. In the illustrative file the line is USD 380 a month. The book is careful to present the line as a limit the earner sets for their own safety, not as a verdict on the people who ask.

The third step is the reply. The book supplies replies for the door that end in a date, so that a request is answered with a sum, a timetable or a refusal that can be kept. Relatives who can carry a share are brought in one at a time through three instruments: a start, a page of written terms and a test. A family meeting agenda, a one-page set of terms, a monthly skill ladder and a four-line graduation test give the stages. Twelve case files, lettered A to L, each show a worked sum, and six other earners at pay levels from USD 300 to USD 30,000 a month have their line computed in the same way.

The book also reads the local rules that bear on family support and records at which issuer each was read. Where a rule could not be confirmed, the book says so. The projected outcome in the illustrative file, after a year of the new line, is a giving figure of USD 4,560, the debt repaid, and two relatives making a contribution. The book presents these as worked results from stated inputs, not as a promise of what any family will do. Its lasting claim is that a limit set from the earner's own floor protects the earner and the family together, and that a duty with no end date becomes dependable only once it has a date.

Book 3: Feet in Zoar, Heart in Sodom

Feet in Zoar, Heart in Sodom is about leaving. A person leaves a former life, which may be a habit such as betting, a post, a yard or a commitment, and the book asks how to arrive somewhere rather than hover at the edge. It takes its frame from the Genesis account of the leaving of Sodom and from the warning in Luke 17:32 about looking back. Seven stations follow: the warning, the delay, the hand, the road without a way back, a little town, the look back, and a station on Lot's wife. A reader who does not share that frame can still use the stations as a sequence of decisions, each with a date and a test.

The central figure is the gap between hearing a warning and acting on it. The book follows four walkers, who waited 75, 332, 34 and 85 days before a first act, an average of 131.5 days. The four waits together came to 526 days. One of them, a welder who stopped betting, is the book's clearest case: two slips cost him USD 35 across 293 days, against about USD 1,172 at his old rate of spending. The first slip was met with a twenty-four hour protocol and not with a restart, which the book presents as the difference between a slip and a relapse. The waiting itself has a cost, and the book counts it.

The tools are practical. A warning log records what was heard and what was done, with the gap between them in days. A wage stub names what the old life paid beyond money, such as company, status or relief, and then names what will replace each of those. A ninety-day runway combines a cue map, which lists the places, people and times that set off the old pattern, with if-then plans for each cue and a rule against camping on the plain, meaning a long stay in the uncertain middle with no destination. A little town is chosen by three tests: it must be near, small and it must end on a date. A card records each warning as a date, an act and a cost.

After the move, three records keep the reader honest. A forward ledger shows the plan for each coming month. A look-back log records each time the reader returns in thought or deed to the old life. An arrival review, held at fixed points, tells a lapse from a return. The book makes no claim that the process is simple, and its figures are illustrative and worked from stated rates. It is a book about a decision and the steps that make it stick, and the reader who finds that a cue or a debt needs help from a family member, a faith community or a clinician should take that help as part of the plan.

Book 4: Metanoia and Teshuvah

Metanoia and Teshuvah takes its two words from Greek and Hebrew. Metanoia is the change of mind and teshuvah is the turn. The book holds that real repentance is both, and that it is complete only when the same situation returns and the choice is different. Five movements run through its fourteen chapters: see it, own it, stop it, repair it and prove it. Four composite people carry a taking of USD 1,120, a short-paid sum of USD 6,000, a rumour that cost USD 1,200 and a betting habit through each movement. While their promises were being renewed, the harm grew at USD 80, USD 150, USD 150 and USD 162 a month, which the book uses to argue that a promise without a plan wears out the person who makes it first.

The method begins with the plain sentence. A wrong must be written as one sentence with a verb, an object and a number or a date before any apology is drafted, because a vague wrong cannot be repaired. An excuse ledger lists each excuse beside the fact it hides. A confession has two directions, toward the person harmed and toward those who need to know, and the book does not treat the second as permission to tell the whole town. An apology has six parts. Where a wrong is public, the statement made in public is separated from the private repair, and the repair is measured by what was taken and not by the wording of regret.

Repair has a price, and the book prices it. In one case the repayment is USD 1,344, paid at USD 56 a month. The stop plan carries a date, a cue map and if-then rules, in the same form as Book 3. The measure of a complete change is the proof year, a test that returns the person to the same situation, with no fear of discovery, and checks whether the choice holds. The last chapters deal with falling and getting up again, and with the view of the person who was harmed. Eight working papers, a proof-year calendar and a table of sources, each with its reading status, close the manual.

Public examples are used with care. The manual describes a South African commission set up under the Promotion of National Unity and Reconciliation Act, No. 34 of 1995. Its amnesty figures, as summarised by South African History Online, were 849 granted out of 7,112 petitioners, or 11.9 percent. The manual draws only one point from them: a public turn was conditional on a full telling. The manual also reads the Tanzanian remission order that applies to Mainland Tanzania, the Tax Administration (Remission of Interest and Penalty) Order, 2018, as a model of structure only. The two summaries the book read give different windows for applications, and the order's own text was not read. The order is history and not a current rule, Zanzibar is outside it, and a reader with a tax liability should confirm the position with the Tanzania Revenue Authority. The stages model used in the chapters is summarised from a 1994 book that the manual did not read directly, and that is stated in its sources table.

Book 5: Trust, Yes. But Verify

Trust, Yes. But Verify is for a buyer about to pay, sign or hand over money for land, goods, services or a partnership. Its opening case is a nurse on USD 900 a month, who was ready to pay USD 14,000 for a plot at a handshake. That price was 15.6 months of her pay. The seller was not the registered owner, who had died. The book's first principle is that the depth of checking should follow the amount that could be lost, counted in months of pay, and that the depth must be decided before the first payment, because money paid before a check is money the check can no longer protect.

The seven checks follow the order in which a deal moves. The first is to name the claim as one sentence that a registry or a visit can test. The second checks the person by identity and address, and by two references that the buyer finds rather than those the seller supplies. The third reads each paper at the office that issued it, with a note of what that office can and cannot say. The fourth follows the money, so that a payment is made only to a name that matches the registered owner. The fifth is a visit to the site. The sixth puts the agreed terms in writing. The seventh pays in stages, so that the money at risk is always less than the value of what has been delivered.

The staged payment is the book's chief protection. In the illustrative case, the plan pays in stages of 10, 30 and 60 percent, which leaves USD 5,600 at risk against the USD 14,000 price, for USD 114 of travel and leave. A broker's holding fee of USD 700, requested on 9 September, was declined in the case. The book treats a holding fee and a feeling of certainty as reasons for slower steps, not faster ones. It holds that a good feeling is not a check, and it cites research on lie detection to make the point, without offering any verdict on the person in the case.

The proportion rule sets how many checks a deal needs. Where the amount at risk is under a quarter of one month's pay, two checks are enough. Up to three months' pay needs four, and anything above that needs all seven. The book's eight kit sheets and twelve exhibits give a form for each check, and its table of sources marks which items were read. Several search and copy fees are not printed, because none was read at the issuer. The reader must ask the registry for them. The book recommends no lender, provider or product, and its checks are a method for any deal, not a guarantee of the outcome.

Book 6: If I Die Today

If I Die Today asks what a family will need on the day after a death, and it asks for the answer to be prepared within four weeks. The method is twenty-eight diary days in sixteen entries, about 1,165 minutes in all, and a Family File of twelve sheets to tear out, with a letter to be opened. It is written for a parent, a partner or the owner of a small trade who has never listed what the family would need. Its opening question is what happens when a person dies without a will, and the book answers it by showing how much a will actually governs.

In the illustrative household, a family earning USD 2,350 a month held eleven assets worth USD 31,720, none of them in a joint name, and four debts totalling USD 2,445. A guarantee of USD 2,400 was found on day five. A will governs only USD 10,600 of the USD 112,520 in play, which is 9.4 percent. Five forms with names on them move USD 88,420. Most of the money therefore passes by the name on each form, not by the will. The book checks seven forms in all, and it treats the name on an account, a policy or a register as the thing to correct first.

The four weeks run in order. The first days build a register of what is owned and owed, open the accounts that sit behind screens, write a one-page sheet for the first night, and find the debts nobody listed. Guardians for the children and an executor are named, each asked before being written down, and a reserve is named too. Money for the first month is set aside as USD 3,050, within reach in week one, against USD 2,870 of first-week needs. A chapter covers a kiosk and its phones, and a handover sheet names the person who does the work. Three copies of the file are kept in two kinds of place, one trusted person is briefed, and a drill is run with a stopwatch. In the illustrative case, the drill was passed 8 of 10 in 17 minutes.

The legal lines need the most care, and the book marks each by how it was read. For Mainland Tanzania, the Probate and Administration of Estates Act, Cap. 352, sections 58, 88 and 89, was read from a private site's copy of the text rather than the official text published by Parliament. Section 58 concerns caveats against a grant, with a period of four months unless withdrawn. Section 88 concerns the law applied to an estate by tribe, religion and intention. Section 89 concerns wills that the court declares valid and whose original it keeps. A death benefit page of the Public Service Social Security Fund, read by summary and dated 20 October 2026, gives survivor shares of 60 percent to children and 40 percent to a spouse where both survive, a spouse taking all where there are no children, and a qualifying period of 180 months of contributions. The book marks that page for confirmation with the Fund.

Several other lines were not confirmed. The Local Customary Law (Declaration) (No. 4) Order, GN 436 of 1963, is recorded by title only. The National Social Security Fund site would not open, so its survivor benefits were not read. A search for a published Judiciary figure on probate timing found nothing that could be printed, and the Registrar's fee for depositing a will was not read. For Zanzibar, the probate and administration lines, including those applied by the Kadhi's courts, and the Zanzibar survivor benefits were not read, and the book prints no Zanzibar rule at all. The reader should confirm every one of these lines with the issuer before relying on them.

Mainland Tanzania and Zanzibar

The series keeps the Mainland and Zanzibar apart, and the reader should not take a Mainland line to Zanzibar. Most of the method in the six books, including the record, the reconciliation, the staged payment and the giving line, is not tied to any jurisdiction, and it applies in either place. The legal and fee lines are another matter. Handbook 127 reads Mainland probate and survivor rules and marks the Zanzibar lines as not read. Handbook 125 reads a 2018 Mainland tax remission order and states that Zanzibar is outside it. Handbook 126 describes registries by what they do, such as the land registry or a company registry, and leaves their fees to be confirmed.

The distinction matters most in Book 6. Mainland Tanzania's probate law and the social security schemes it names cannot be assumed to apply to the same family in Zanzibar, which has its own courts and rules. A family with property or members in both places should treat the two as separate questions and take advice on each.

What to confirm before acting

Several lines in the series are not settled by the books, and the reader should confirm them before acting. The first is the Mainland remission order of 2018. The two summaries read disagree on the application window, and the order's text was not read. The Tanzania Revenue Authority should be asked whether any part of the order is still open, and whether the reader's own position qualifies.

The second is the Mainland probate material in Book 6. Sections 58, 88 and 89 of Cap. 352 were read from a private copy. The official text published by Parliament, and any current practice of the courts that apply it, should be checked. The third is the Public Service Social Security Fund death benefit page, read by summary and dated 20 October 2026. Its survivor shares and its qualifying period should be confirmed with the Fund. The National Social Security Fund survivor benefits, the Judiciary's probate timing, the Registrar's deposit fee and every Zanzibar line remain open.

The third group is the research and the second-hand material. The stages model in Book 4 is summarised from a 1994 book the manual did not read directly. The South African amnesty figures in the same book come from a published summary and are presented only as a point about full telling. The scriptural readings in Books 3 and 4 are the books' own, and the reader who holds a different text or tradition should check them against that tradition's sources rather than against this summary.

Reading order and first steps

A reader who knows the decision should open the matching book. For a household or small business that reaches the end of the year without a clear account, Handbook 122 supplies the record method and the weekend structure. For a family earner under constant calls, Handbook 123 supplies the map and the giving line. For a person leaving a habit, a job or a commitment, Handbook 124 supplies the warnings, the cue map and the arrival review. For a wrong that has been named only in outline, Handbook 125 supplies the plain sentence and the repair. Before paying for a plot, goods or a service, Handbook 126 supplies the seven checks. For the family's protection after a death, Handbook 127 supplies the register, the will question and the drill.

The first step in every book is the same: fill in one line of the first sheet, then stop. The Year-End Postmortem begins with three rules on a lid. From Dependants to Pillars begins with one month's transfers. Feet in Zoar begins with a dated warning. Metanoia and Teshuvah begins with one plain sentence. Trust, Yes. But Verify begins with the amount at risk in months of pay. If I Die Today begins with the names of the person to ring first and of the person who minds the children. Each first step needs a sheet of paper, a date and a single named person, and each can be done in an evening.

Every book ends with a list of what it settles, a sources table and a set of questions to confirm. The series is general information. It does not replace the advice of a qualified adviser, an advocate for a dispute, or the issuer named beside a figure, and its figures should be checked against the issuer before anyone acts.

Money That Stays is the shelf of the Zatra Handbooks that deals with what happens after money arrives. Its seven books share one problem: a lump sum, a sale or a harvest is spent faster than the plan meant to keep it. Book 128 handles a lump sum, Book 129 the requests that follow it, Books 130 and 131 the first year of a shop and the price that keeps it solvent, Book 132 a smallholder's harvest, Book 133 a shop's profit over ten years, and Book 134 a restart after a fire, theft, closure or debt. Each book works one running case in full, gives dated pages to copy, and sets its decisions in writing before the pressure of a request arrives. The cases are illustrative composites and the sums are worked from stated inputs. This summary explains what the series covers, the order in which to read it, and the figures that depend on rules which must be confirmed with the issuer before anyone acts. It is general information, not legal, tax, pension or investment advice.

What the series covers

The seven books are numbered 128 to 134 in the Zatra Handbooks and carry the label Money That Stays, Book 1 to 7 of 7. They are written in English for a reader who already earns, sells or farms, and who has just received money or has lost some. The subject is not how to make more money. It is how money already held stays in the business or household that needs it.

The series rests on three working habits that recur in every book. The first is a dated page written before the decision, such as the ninety-day hold in Book 128 or the family rule in Book 129, because a rule written in calm is kept more often than one made under pressure. The second is a ledger that closes: cash and stock are counted separately, and any difference is traced before the next purchase. The third is a test with a number in it, such as a cash floor, a break-even sale, a payback month or a gate before each stage of a farm, so that a verdict comes from arithmetic and not from mood.

Each book carries tools to copy, from notebook pages to one-page cards, and a set of dated case files. The handbooks state that the cases are composites and that the figures are computed from stated inputs. A reader should treat them as a model to test against the reader's own figures, not as a forecast of return.

Which book to open first

The order matters less than the situation. A reader who has just received a pension lump sum, an end-of-service package, an inheritance or the proceeds of a plot should open Book 128 at once, because the first sixteen days decide how much of the sum remains. A reader whose relatives, staff or committee are asking for money should read Book 129 before the next request is answered, since its ceiling is set on paper before any reply is sent.

A reader opening a new shop, kiosk or counter should read Book 130 for the first year's cash and Book 131 for the price that turns sales into profit rather than stock. A smallholder paid once or twice a year should read Book 132 before the next sale. An owner who has traded for years without the business growing should read Book 133. An owner whose business has closed, burned, been robbed or fallen into debt should read Book 134, which runs sixty days day by day.

Book 128: Rich in June, Broke by Christmas

The book opens with the reason lump sums disappear. A payout arrives in one morning, in front of people who knew the date, and it is requested at once. The illustrative case is Yusuph Kijazi, a retired mathematics teacher in Songea, who was paid a pension lump sum of USD 7,920 on 15 June. The fund's letter set out the calculation: a lump sum of 33 percent of the annual pension, commuted at a factor of 12.5, alongside a monthly pension of USD 107.20. The sum equalled 73.9 months of that pension, and the book notes that windfalls of this kind are usually asked for and spent readily.

The first defence is a page of six lines copied from the letter before any request is answered, with a count of ninety days from the morning the money arrived. Within sixteen days seven claims were made on the sum, totalling USD 5,293, which is 66.8 percent of the payout. Only two of the seven carried a written amount and a date, and together they came to USD 808. The urgent list sorts each claim into one of three lists. Pay is a written amount and date. Answer is a real need without a date, which is met in writing within two weeks. Park is an opportunity that can wait. In this case the lists were USD 808 to pay, USD 1,835 to answer and USD 2,650 to park, and they add back to USD 5,293.

The ninety-day hold is the central rule. Written on Day 0, it names what may leave the sum and sets caps in dollars. Four kinds of payment are allowed: a dated debt, a noticed fee, living costs drawn from the pension, and an exception. Exceptions are capped at 10 percent of the sum across the ninety days, which was USD 792 here, and the hospital took USD 230 of it. On Day 90, Sunday 13 September 2026, the balance stood at USD 6,882, which is 86.9 percent of the sum. The book is clear that the hold is Zatra's method and not a rule of any regulator or fund. It works because the reader writes it down before the first request arrives and keeps to it. The delay is priced too: at an assumed 10 percent, the waiting cost USD 175.36, and the hold breaks even against a single USD 1,500 loss at about one chance in 8.6.

On Day 17 the remaining USD 7,112 was divided into three accounts. The reserve held six months of core cost at USD 154 a month, which is USD 924. The allowance, equal to the gap between income and core spending plus a little spending money, was fixed for twelve months and not topped up, at USD 912. The working fund, the remainder, was USD 5,276, which is 74.2 percent of the balance left after the claims. The three lines add back to the balance and are checked against it each week. The book's habit chapters set a routine of 990 minutes over ninety days, about eleven minutes a day, with a daily entry, a weekly hour and a monthly close. In the book's case a USD 7 difference found in the second week was traced and corrected, and that discipline was kept for the rest of the ninety days.

The tax and paperwork chapter rests on one payout file and four written questions sent to the fund, the bank and the tax office by Day 30. A sale price can leave three different tax bills depending on what paperwork is held. On a plot sold for USD 24,000, the book's example gives tax of USD 400, USD 720 or USD 1,700 depending on the proved cost, and a flat route of USD 720. The final chapter gives the working fund one purpose. Juma Lyatuu of Temeke had USD 7,800 and three offers. The book costs the grain mill he was offered, including three months of fixed costs, at USD 6,240 in all, and tests it against a 30 percent fall in contribution, which moves payback from 24.0 months to 60.0 months. The one-page decision has five lines, purpose, sum, owner, maximum loss and review date, and a month-6 line that stops the project if it falls short. Paying off a loan at 3.5 percent a month is treated as a purpose too.

Book 129: Stop Being the Family ATM

Book 129 deals with the requests that come after a payout and with the regular ones that come from relatives, staff, a committee or a sender abroad. Its core test is written before the reply is sent. Four questions decide whether a request is a hook, which earns a return; a not yet, which waits; or a fish, which is eaten. The questions ask what the money buys, who earns from it and from when, what the asker puts in, and what the giver can lose without missing a bill. The household ceiling is set first at one month's leftover pay, and the asker's stake must be at least one fifth of the amount asked.

The book sets a ladder of five rungs between a gift and a partnership. Rung 1 gives, Rung 2 lends, Rung 3 shares, Rung 4 grants a tool paid to the supplier, and Rung 5 partners. A partnership carries the giver's time, so one that would need less than four hours a month is closed. A loan is tested first: the instalment plus living costs is set against net income. A loan is sized at no more than one third of the borrower's own net estimate, tested by two working sellers, and paid to the supplier in stages. Once a request clears the test it goes on a written page of ten lines that both sides sign and a keeper holds. The book's example is a USD 750 family loan with fifteen instalments of USD 50 and a review after the sixth. A keeper who scores under three is replaced, and a handshake loan is sized as a gift.

The case files apply the method to a salaried earner abroad, a shop owner with staff, a community committee and a school. In one, Lukas Shija in Leeds sent USD 250 a month for three years, thirty-six transfers in all, or USD 9,000, and the family held nothing more at the end. The book's model applies an average cost of 7.9 percent to the transfers. Its answer was one funded start of USD 1,500 for a cold box and stock for his mother's fish stall, kept on one page with a review date. In a staff case, an advance is capped at wages already earned, which is wage times days divided by 30. Hadija's USD 300 was therefore not an advance, since only USD 80 of a USD 200 wage had been earned by day 12.

A fish is right only where the need cannot wait for any return. Treatment is paid against a quote and a burial through the treasurer. Food is paid weekly, because Habiba's sister was fed by three weekly payments of USD 20 and not one payment of USD 60, since food is eaten each week. A reserve of half the pot keeps emergencies from competing with hooks. A school message is three requests, arrears, a fee ahead and a way out of school. Arrears are paid to the school against its statement. A yearly pot, in one case 8 percent of take-home pay, answers every request from one figure written in January, with a quarter of it held for a child in the last year of school.

The book closes with ten hooks that recur in family requests, each tested by four questions: skill, customer, downside and title. Pricing them side by side shows payback from 3.3 to 15.0 months and a loss at month six from USD 80 to USD 310. A payback over twelve months needs a rung that carries the giver's time, and a hook the giver cannot check from a distance waits. Community requests are answered from a line set before they arrive, a share of net profit, and a pledge is a loan to the committee unless it is paid within the month. The book is a guide to making each reply leave something behind, not a rule that refuses every request.

Book 130: The Money Ran Out in March

Book 130 is the first-year capital plan for a new shop, kiosk or counter. Its premise is that a starting sum is not one balance but five places: stock, fit-out, rent paid ahead, registration and a cash float. The illustrative shop, run by Furaha Tarimo in Buguruni, opened on USD 2,500. On its first morning the bank account showed USD 912, while the worth of the shop, counting stock at cost, was USD 2,440. Three limits are fixed before the first sale and twelve month-end checks are written in advance, so that a shortfall appears as a number in the month it happens.

The ledger has eight headings and records only money that moves. A count of the drawer, the phone and the bank at each month-end is set against the arithmetic, and a difference is traced before the next order. A profit can rise while cash falls, and the book shows it in a school kitchen contract. In that month profit was USD 108 and cash fell by USD 269, because USD 301 of new stock was bought before the first delivery. Stock is counted at cost, and any withdrawal in cash or goods is a draw. An asset is entered once at cost, given a life in months and charged by division, so that profit carries a fair share of it every month. A motorcycle at USD 720, charged at USD 15 a month over 48 months, is one such example.

A contract adds sales and a claim on cash on the same day. Before the school kitchen contract is signed, the shop tests its unpaid invoices against cash less a floor of one month's running costs. An unpaid balance of USD 700 against USD 645 of spare cash fails, while USD 400 fits. The motorcycle passes the cash-floor test in month 11 by USD 19 and fails in month 10 by USD 31, and it pays back in cash in 25.7 months. An asset is bought only from cash the month will still hold above the floor, and the book reads the year afterwards in five numbers: worth, return, payback, lowest margin and assets.

The payback table adds monthly profit to a running total and divides the sum still unearned by the latest monthly profit. At month 12 the illustrative shop had USD 1,196 of its sum unearned, and at the latest monthly profit of USD 179 that gives payback in month 19. The year-end review sets plan, actual and gap side by side. The shop's USD 349 gap splits into USD 141 of volume, USD 133 of margin and USD 75 of costs. It passed one of four scale tests, so the verdict is hold. A phone-repair counter in the same book, which pays back in month 18 on USD 1,800, passed all four and is told to scale. The verdict rule is written in month 0: scale needs all four tests, stop needs any one of three, and hold is everything else. A draw rises only when profit to date has outrun draws to date by a stated sum.

The book's last chapter is a page for the next person. It holds the reasons beside each part of the plan, the first-year figures and the review dates, so that someone else can keep the shop running. The case of a hardware counter in Bagamoyo, which opened in January with USD 2,000, shows why the dated checks matter. Its owner read USD 18 in the bank on 14 March and could not say where the rest had gone. The book's approach is to answer that question in advance, with a number for each month.

Book 131: Sold Out, Zero Profit

Book 131 starts from a shelf that sells out and still leaves nothing in the pocket. Its central lesson is that margin is not markup. A markup divides profit by cost and a margin divides it by price, so a 50 percent markup is a 33.3 percent margin, and a 50 percent margin needs a 100 percent markup. In the illustrative Mbeya counter a phone-charger seller sold all 300 chargers at USD 6.00 and ended the month with nothing, because the share taken by the supplier and the rent left nothing. A trial month at USD 8.00 sold 190 and earned USD 160. Profit rose at every price tried, even though buyers fell from 300 to 190.

Each price is built from the landed cost of one unit, which is everything paid to bring it to the shelf in a state to be sold, divided by the units that can be sold. Normal spoilage stays in the cost and selling costs stay out. Every payment of the month then belongs in one of three bins, decided by one test: whether the payment would change if 100 more units were sold. The contribution of each unit is its price less its variable cost. A screen repairer named Nuhu needs 45 repairs a month to cover his fixed costs, and his 62 repairs leave a margin of safety of 17 repairs, or 27.42 percent. A rent rise of USD 80 costs ten repairs a month, because each repair contributes USD 8.

Discounts, rises and rival moves are priced by the volume each must win. A discount of d on a margin of m needs d divided by (m less d) more units to hold the same profit, and a rise of i needs i divided by (m plus i) fewer. A 10 percent discount at a 30 percent margin needs 50 percent more units, and at a 20 percent margin it needs 100 percent. A volume the shop has never sold, or cannot serve, ends the discussion before it starts. The book's tiers, baskets and first-visit offers are tested the same way, and a floor price and a card of grades keep the answer to a rival or a dearer load from being made in a queue.

Hours are priced as well as goods. A worker named Amina has 100 billed hours that must carry her whole month. Her rate is the take-home target plus fixed costs divided by billed hours, which gives USD 6.00 in the example, and not by hours worked. Unbilled hours, such as time at the bench that is not invoiced, are a cost of the business in the same way as rent, so the billed hours have to carry them. The book's price notebook asks for a log of unbilled hours, a clock to correct the estimate, and a road or quote-visit cost in every price. The customer sees one number, and every line behind it is an hour or a cost.

The book's pricing kit is a notebook of six items that keeps a price honest in thirty minutes a month. In one case a can of a product moved from USD 11.40 to USD 12.00, because the break-even price was 10.86 and the target price 12.05. The method is simple and repeated: one cost, one contribution, one break-even line and one target, each checked by a figure written before the next sale.

Book 132: The Harvest That Bought the Land

Book 132 follows a smallholder's season from the sale to the next reinvestment. Its rule is that the first sale of the year pays next season's inputs before anything else. In the illustrative case at Katesh, a maize farmer named Zuberi Tarmo sold 1,200 kilograms at the market scale at USD 0.20 a kilogram and held USD 240. Seed and fertiliser for three acres took USD 90, USD 100 was set apart for the first animals, and USD 50 was kept for the lean months. The seasons are set out in three stages, and each stage has a gate.

Stage one is seed and the first cycle, stage two is livestock bought from the harvest, and stage three is leased land and hired labour. Before each step the test is run at the expected price and again at 10 percent less, and a stage is climbed only after a second clean season. Credit counts as a claim on the harvest, and the book shows its cost. A neighbour who took the same inputs on dealer credit repaid 731 kilograms of maize for what 450 kilograms would have bought in cash.

Leased land is priced before the crop is sold. At placeholder prices, which the book states are to be replaced with local quotes, one leased acre costs USD 74 and returns USD 46 at 600 kilograms, breaking even at 370 kilograms. Two acres need USD 148 in cash before a kilogram is sold, and a lease must be paid in full before planting. A share rent protects only a small part of a poor season, which is USD 2.20 an acre in the poor case. A written split settled USD 12 between two brothers.

Cooperatives and offtake contracts are the book's route to better terms. A group buys together, sells together and stands between its members and a buyer, at the price of time, a share of the price and some control. Registration under the Cooperative Societies Act needs at least twenty members and a volume that covers costs, and the registrar should confirm this before a group is formed. An offtake contract is a buyer's promise to take a stated quantity at a stated price on a stated date. The paper must state quantity, quality, price, delivery, payment and default. In one example an offer of USD 200 paid USD 190 after a moisture clause and an advance charge, so the highest printed price among three offers gave the lowest cheque.

Holding grain for a better price is priced too. For 800 kilograms of maize held five months, the price rise must cover storage loss, bags, treatment and interest on the money borrowed to hold the grain. In the book's example that bar is 18.4 percent, and the national rise of 11 percent falls short. Farm records are three one-page sheets, for each crop, each kind of animal and labour, written on the day money moves. In one case memory understated a season by USD 35, which is 31.8 percent of its true cost. The reserve is sized to the largest single gap in the season, USD 333 in the case of Devotha Nnko, and it is filled in order: seed, gap, stage.

Land and the written page close the book. Land on the Mainland is let and held through the village, and the cases show what an unwritten arrangement costs. A lease with no term cost one farmer USD 36, and a sale without a minute left a buyer with USD 600 paid and no paper. The Mainland rules on village land and leases must be confirmed with the village and the district land office before any lease is signed. Zanzibar land is governed separately and is not covered in this book.

Book 133: Busy for Ten Years, Still Small

Book 133 deals with a shop that is busy every day for years and yet is the size it was five years earlier. Its argument turns on a fixed share of profit. Two shops in Temeke, run by Subira Mwakasege and Hamidu Kimaro, sell the same household goods from the same wholesaler, and each opened with USD 2,000. Subira put back 40 percent of each year's profit into the shop, and Hamidu took everything out. At an assumed return of 25 percent on the capital kept in, Subira's shop held USD 5,187 of stock, shelves and cash after ten years. She had taken USD 4,781 out for the household, of which USD 339 sat in a reserve she owned, so USD 4,442 had reached her household. Hamidu had taken USD 5,000, which leaves his household USD 558 ahead in cash, while Subira's shop is USD 3,187 ahead on the shelves.

The book's first tool is a one-page monthly ledger of fourteen lines. It turns sales, stock and bills into one profit figure before anything is taken. Stock is counted at cost, the margin is checked against its band, and the till is predicted and counted. The drawer answers a different question. In one month Hamidu's drawer showed USD 342, while the month had earned USD 42. The reserve is sized by a stretch of trade that can be named, built from the household's share and kept in a place that passes four tests. A USD 125 month met from stock cost Hamidu USD 31.25 of profit in every later year, and met from a reserve it cost Subira no profit at all. Holding the reserve cost USD 39.58 a year in the book's case, and it is released only by five written calls.

The owner's draw is a fixed monthly sum equal to half of last year's profit divided by twelve. It is paid on the same day each month and reset once a year in January. Months above it repay earlier calls on the reserve. Months below it may call on the reserve for up to three months, and in the fourth the draw stops and the ledger is read for the cause. The book's example shows a remainder that swung the household by USD 300 between two months, which is why the draw is a fixed number and not a mood.

Growth follows a ladder of four rungs: stock, a leased asset, paid labour and a signed contract. Each is a larger promise than the one below it. Stock stays the first call until cover is 10 percent above the shop's own starting level. Every other rung is priced in the sales it needs, tested by a count of real demand, and refused or halved when the cash needed to carry it exceeds the year's kept share. The next asset is chosen by payback time and not by appearance. An urgent bill is met first from the reserve and then from the draw, because USD 300 taken from stock and cash would have cost Subira USD 1,012 of her ten-year result.

The habit outlives the owner. The closing chapter sets a handover page that a successor can run, with the monthly ledger, the reserve rule and the draw written in the same terms. The shop's growth then does not depend on one person's memory, and a new owner can answer the same three questions the first owner answered: how much profit to keep, how much to draw and which asset to buy next.

Book 134: Sixty Days After the Fire

Book 134 is the restart book. It takes an owner whose business has closed, burned, been robbed or fallen into debt, and it sets out sixty days day by day. Its opening point is that capital is a process rather than a balance: a loop of buying, selling and buying again. A fire breaks some parts of that loop and leaves others standing, and the book counts the parts before it plans anything. The illustrative case is Salha Mbilinyi, who sells household goods from a wooden stall in Tandale, Dar es Salaam. A fire took TZS 1,768,000 (USD 680) of stock and frame, and left her owing USD 435 to four creditors. She restarted on USD 135 of her own and borrowed nothing at the outset.

The first days count what is left. Day 1 counts the goods and visits the market office, day 2 collects the creditors' own statements, and day 3 rebuilds the customers' list and fixes the order in which five people are told. Every figure is marked as a document, a statement or a memory, so that the restart rests on evidence and not on recollection. The sixty days then run as a calendar on which close day falls the day before payment day, and the sum for each creditor is set aside on close day.

Debt comes first. Each creditor is ranked by what he or she can do next, and each is offered a small fixed sum that the weakest week can still pay. Every loan is priced at its true cost, and the microfinance loan in the example charged 3.5 percent a month. Only one of the four creditors could stop the stallholder selling on the row where her customers know her, and the book's order reflects that. The offer is written in seven lines, signed with a witness, and a copy is kept for each side.

Stock is bought in the units customers take and paid for from cash already in hand. The first order is sized by weekly demand and not by the old shelf. The stall had never sold more than TZS 260,000 (USD 100) of goods at cost in a week, so a shelf of TZS 371,436 (USD 142.86) was all its demand could use. Prices are set at cost plus 30 percent, and no line may leave under its floor, which is the weekly claims divided by goods sold at cost. Credit from a wholesaler is asked for in steps of 10, 20 and 30 percent of the weekly order, and cash lots on a fixed day build the record that earns it.

The records chapter uses five short records and two proofs. Opening stock plus purchases less cost of goods sold must equal the shelf counted, and sales less costs and pay must equal the cash counted. A difference is traced before the next restock, and a copy of each page is kept where a fire, flood, thief or lost phone cannot reach it. The restart is then tested against a slow week and a thin one. Each instalment must fit within the cash the weakest week can produce, and a late payment is announced beforehand rather than moved to wait for good sales.

The later choices follow the ladder of stock, asset and contract that Book 133 sets out. A loan is taken only where it earns more than it costs, and four written tests decide: payment record, unmet demand, cash first and repayment source. A lump sum goes to the dearest debt first. Day 60 ends with a choice between borrowed money and a payout, and the book's last chapter explains what to tell the next generation about a failure. The account has five parts: what was lost, what was owed, what was left, the sequence of the first weeks, and the one decision to reverse.

Where the figures need confirming

The series states figures in US dollars, with shillings where the handbook uses them. The Tanzanian appendix to Book 128 is the only part written in shillings, at TZS 2,600 to the US dollar. Its rules were read on 22 October 2026 from the NSSF and PSSSF benefit pages, the Bank of Tanzania tender documents, the Tanzania Revenue Authority pages and the Finance Act, 2026. The appendix states that every row was carried from the earlier edition and was not re-read in this edition. The lump-sum factor of 12.5, the commutation of 33 percent and the accrual of 1/580 a month should therefore be confirmed in writing with the fund before any figure is relied on.

The same appendix records that the public service retirement age is compulsory at 60 and voluntary from 55, and that the NSSF and PSSSF benefit rules, including the earnings-base tests, were not re-read. Tax rates and thresholds move, often each July. The sale-tax examples in Book 128 depend on proved cost and approved value, and each such figure must be confirmed with the Tanzania Revenue Authority before tax is paid or a sale is signed. Fees for registration and licences named in Book 130 must be confirmed with the registering authority before a budget is set.

The books apply Mainland rules and say so. Zanzibar has its own revenue authority, the Zanzibar Revenue Authority, and its own rules on registration, licensing and pensions. Book 128 states no Zanzibar rule and works its Unguja case without one. Book 133 states that no Zanzibar figure is printed in it. A reader in Zanzibar should take the rules from the Zanzibar authorities and not from these books.

The illustrative people, shops and farms are composites, and the sums are worked from stated inputs. The placeholder prices in Book 132 and the assumed 25 percent return in Book 133 are assumptions to be replaced with a reader's own figures. Nothing in the series is a forecast of return, and the cases show method rather than results.

Hold Your Ground is the set of six Zatra Handbooks about the people a business or household depends on, and the decisions that follow when one of them leaves, asks, fails or misleads. The books cover losing a key person, refusing without losing a customer, hiring and parting with staff, controlling a heated reaction, listening before speaking, and auditing advice before money moves. Handbooks 135 and 137 rest on Mainland employment rules, Handbook 136 sets out hours rules for both sides, and Handbooks 138, 139 and 142 name other instruments mainly to show where an advocate is needed. Mainland Tanzania and Zanzibar are kept apart throughout. Money figures are illustrative. Most books work in US dollars at TZS 2,600 per USD as a convention for arithmetic, and Handbook 142 works in shillings. Every legal row carries a status that shows how the text was read. This summary is general information. It is not legal, tax, employment or investment advice, and each rule should be confirmed with the issuer or an advocate before a step rests on it.

What the series covers

The six titles are Be Ready to Lose Anyone (Handbook 135), No Is Not Disrespect (136), Hire for Character, Fire Fast (137), Calmness Is Mastery (138), Say Less, See More (139) and Honest Critics, Paid Flatterers (142). Each takes one decision that a small business or household makes when a person matters to it, and each is built the same way: a count the reader can make in an afternoon, a worked case from stated inputs, and a set of tools to fill in at a desk.

The cases are composites described by type, not by name, and their figures are illustrative. Money is worked in US dollars at TZS 2,600 per USD, which the books describe as a convention for arithmetic and not a quotation. Handbook 142 works in shillings. Cases dated after 20 October 2026 are projections, and the books say so.

The books are workplace and household education. None is legal, tax, accounting or investment advice, and none creates an adviser relationship. Handbooks 135 and 137 mark each statute with one of four statuses: R for a text read as enacted or consolidated, J for a rule read as quoted in a judgment, S for a summary only, and C for a rule not read. Handbooks 138, 139 and 142 use plain statements instead, such as read at source, read second hand or not reachable. In every book, a reader who acts on a row that is not fully read confirms it with the issuer or an advocate first. Most sources were read on 25 or 26 October 2026.

Holding people: Be Ready to Lose Anyone

Handbook 135 starts from a count rather than a fear. Its first reading sorts what a person holds into five kinds: hands that do the work, keys that open doors, knowledge kept in heads and phones, ties that bring work and help, and money that passes through names. A hold map lists each one, so that single-holder items show before anyone leaves. The case that carries the book is a two-partner solar business at Namanyere, where one technician had led 41 of the last 57 jobs, a share of 72 percent, and 12 of 14 weekly tasks had a single holder.

Supporting readings follow. The loyalty reading holds that a person stays in a place for three reasons: links to people there, a fit between the place and the person's life, and what the person would give up by leaving. The pause test reads a pause in giving or chasing as a measurement only where its cause is something other than the wish to measure, and sets out six readings of a quiet tie. The stay conversation asks once, with six parts: what has been noticed, what is wanted, what can be offered, by when, what follows a no, and what stays between the people involved. A seventy-two hour rule governs silence in the first days after a departure.

The dry run is the book's most concrete test. At Namanyere the owner stepped out of the work for three days, 13 to 15 October 2026, with a card that kept the telephone for three fires only. Of eleven items, seven ran, three stalled and one went wrong. The wrong item was a quotation that gave away USD 60 on one job. The four gaps all sat in the holds of keys, knowledge and money, and by the book's arithmetic cost USD 15 each. Hands and ties had none. The book states that the purpose is not to find fault with the staff who stayed, but to find, on a quiet Tuesday, the tasks that only the owner's presence had been holding.

Five departures are then worked as dated files, each from the day before to the seventh day after: a partner who held the keys, a cutter who held the patterns, a customer who took 62 percent of an egg farm's trays, a friend who held lunchtime cover and a savings group's notebook, and a son who left for the mine when a press earned USD 24 a day against a household need of USD 370 a month. The key-person plan asks for five parts, one for each hold: a second pair of hands, a second key, a page, an introduction and a reserve. The Week-One Sheet is a grid of the five holds against five days. Its filled example supposes a departure on Friday 15 January 2027, which the book labels a projection.

The Mainland employment rules sit in File B and in the local table. Read on TanzLII on 25 October 2026, the Employment and Labour Relations Act, 2004 gives written particulars (section 15 as enacted), a minimum notice of leaving that is 28 days after the first month for an employee paid monthly (section 41), and final payments with a certificate of service (section 44). The consolidation numbers these sections one higher from section 13 onward, so the reader who opens that text finds them at 16, 42 and 45. The Zanzibar rules are not read for this edition, and the book directs a Zanzibar reader to the issuer or an advocate practising in Zanzibar.

Refusing without losing the customer: No Is Not Disrespect

Handbook 136 prices each yes before it is given. A nurse gave 46 hours in one month against a yes budget of 36 hours, a figure set at one third of free hours, or 9 hours a week. A guesthouse owner found that a 15 percent discount needs 27.3 percent more nights to earn the same contribution at a 70 percent margin,. A rice miller refused 14 requests in 30 days and was asked again on 9 of them. Thirty days later the same sentences drew a second ask on only 5 of 14.

The reply has three parts: acknowledge the request, answer once, and leave a door. The book supplies 72 lines by situation and 36 counters for a second ask, and sorts a repeated request into one of seven recognisable forms, among them the second ask, the smaller ask and the deadline. Discount requests come in four kinds, the haggle, the loyalty ask, the volume ask and the threat, and each has a reply. Favours come in five kinds: free work, a free stay, a name, a signature and an introduction. The extra task usually arrives as an addition to a day that is already full.

Three rooms need three registers: the boss, the relative and the elder. A short passage covers the minutes after a hurt reply, with three moves: hear what the refusal costs the other side, hold the answer without repeating the argument, and offer help with the next step. A ninety-day ledger in the book counts 34 requests, of which 29 were answered yes and 6 of those came back. The limits sheet closes the book with five written limits, each a number, a line, a door and a name, kept for thirty days.

The hours rules are in the local table, with statuses. For the Mainland, the Employment and Labour Relations Act (Cap. 366), read on TanzLII on 25 October 2026, sets six days a week, not more than 45 hours a week and not more than nine hours a day. No work may be required beyond 12 hours in a day. Overtime is by agreement only, capped at 50 hours in a four-week cycle, and paid at not less than one and a half times the basic wage. Each row is marked for confirmation of the current text. For Zanzibar, the Employment Act, No. 11 of 2005, read in the Gazette of 7 October 2005, sets not more than eight hours a day or 42 a week, and overtime of ten hours a week, which a collective agreement may raise to fifteen. Later amendments were not checked. The two sets of rules are not applied to each other, and the book states no rate of pay.

Hiring, probation and parting: Hire for Character, Fire Fast

Handbook 137 is the book that most often leads to a decision with legal consequences, so it is the one to read with most care. Its case is a garage at Mikese, where the owner hired a second diesel mechanic who scored 9 of 10 on skill and 5 on character in the interview. The first 24 days of his employment cost customers USD 418: a billing skim of USD 228 on six jobs charged USD 38 above cost, and a set of swapped injectors costing USD 190. The book is explicit that the figure stopped only because the file stopped it, and that the pace of the skim over a thirteen-week probation would have reached USD 988.

Hiring uses five interview questions about character, each scored from 0 to 2. The line is 7 of 10, with a refusal whatever the total if either of two named answers scores 0. Three cheap checks sit between the interview and the offer. The first is a call to the last direct supervisor, which the book times at four minutes, and the second is a check of papers with the body that issued them. The written line has five behaviours, covering theft, lying, disrespect, bullying and what is not the worker's business, read aloud and signed before the first day.

During probation each day gets at most three lines in two columns, one for what was observed and one for what was inferred. A first breach is then sorted by four questions: has it happened before, was it hidden, who bore the cost, and how did the person react when shown. The book's seven-day route runs from a found breach to a parting. Within a day of a finding there are four routes: warn, restrict and hold a hearing, pause for the police, or separate the people first. In the Mikese file, the notice given on the Friday allowed 96 hours before the hearing, more than the 48 hours that the Mainland Code treats as the normal minimum.

The Mainland rules carry the book's weight, and their statuses differ. Read as enacted, the Act's unfair termination part does not apply to an employee with less than six months' employment with the same employer (section 35, R). Notice is seven days in the first month, then four days for daily or weekly pay or twenty-eight days for monthly pay, given in writing with reasons (section 41, R). Severance is seven days' pay for each completed year from twelve months, counted to a maximum of ten years, and is not payable on a fair dismissal for misconduct (section 42, R). The Code of Good Practice Rules, 2007 were not read in the Gazette. Its probation rule, that a probation is of reasonable length and not more than twelve months, is read only as quoted in High Court judgments (status J), and the 48-hour minimum for a hearing is also read in judgments (J), while the outcome and appeal periods rest on a law-firm summary (S).

The 2025 amendments are known only through summaries, and their figures disagree. The compensation bands by ground are, for procedure alone, 6 to 12 months' remuneration; for reason alone, 12 to 18; and for both, 12 to 20 in one summary and 12 to 24 in another. The book does not choose a band and marks the question for an advocate. Expiry of a probation does not by itself confirm an employee, according to a Court of Appeal decision read on TanzLII (Nzaligo v NMB, 2019), but decisions point in different directions on the remedy a probationer has.

Zanzibar is kept on its own card. Under the Employment Act, 2005, read in the gazette as far as section 88, a permanent written contract carries a probation of six months, extendable by up to six months more, and a probation that ends without a letter of confirmation leaves the employee treated as confirmed (section 60, R). An employer who wants to end a probation on fourteen days' notice or fourteen days' wages must act before the sixth month ends (section 60(4), R). The sections from 89 onward, including those on discipline, summary dismissal and unfair termination, were not reached, so the book states no Zanzibar hearing or termination rule. The book also warns of the reverse error. A Mainland owner who believes that Zanzibar probation ends on fourteen days' wages, because someone in Zanzibar said so, is working from the wrong card. The fix is a file for each side and an advocate who practises where the worker is employed.

Heated moments: Calmness Is Mastery

Handbook 138 is built on a log rather than a lecture. The owner records every provoked moment for seven days, then for thirty, with five fields: when it happened, what happened, how hot it was, how long the first move waited and what it cost. A workshop owner on the Arusha to Nairobi road logged 14 moments in the first week. She answered 13 with the first impulse and counted USD 118 of cost, of which insults and staff moments took USD 63. The first Sunday count sorts the week by trigger, hour and cost, so that the owner sees the pattern before changing anything.

The method has seven steps, drilled one a day: pause, name the feeling, wait, observe, ask, decide and review. The pause is planned in advance as an if-then rule. Naming the feeling in two words and a number turned three heated moments into held ones in the case. Observation keeps three facts and one guess on separate lines. The wait is set by the wait ladder, from a ten-second pause for a small move that can be undone, up to two nights and a second opinion for a large one that cannot. The Sunday review takes thirty minutes in six parts: count, sort, read, choose, fix and sit.

Four drills test the method against common situations. A late payment is handled with a statement of account in three lines, a single date and a ladder of three contacts. A rude customer is met with four sentences and a cap. A rumour was answered in four ways in the trial: attack, a reply in the group, ignore and invite. An insult is tested against fourteen sentences said in a hostile call, of which four carried claims about money and ten carried abuse. The book also covers the case where the other party does not play along, since a game is a move that invites a matching move.

The second count on the same sheet shows 14 moments, 2 first impulses acted on and USD 8 of cost, with the median wait rising from 2 minutes to 15. The book notes the limits of one log and writes them beside the result. Its legal passages name Mainland instruments: the Employment and Labour Relations Act, 2004, read at source, where fair reason and fair procedure are cited at section 38; the Law of Contract Act and the Law of Limitation Act, read at source with no period printed; the Cybercrimes Act, 2015, read at source, where sections 16 and 23 are the nearest to a reply to a rumour; and the Fair Competition Act, 2003 for refunds and consumer complaints. The Tanzania Revenue Authority's receipt rules were not reachable. Zanzibar's rules on discipline and dismissal are not stated.

Listening and watching: Say Less, See More

Handbook 139 asks an owner to measure first how much of a meeting he or she takes up. The opening case is a maize mill and feed shop, where the owner spoke for 41 of the 60 minutes of a Monday meeting on 7 September 2026, a share of 68.3 percent. A speak-last rule followed, under which the owner gives a view only after the others have spoken. Over five Mondays the owner's share fell to 31.7 percent. In the same weeks the team found that a scale had given away 0.8 kilograms of maize on every bag, which the book prices at USD 307.20 by the Monday it was mentioned, and at USD 166.40 a month had it gone on.

The thirteen watches move from the quiet mouth through what is said, what is not said and what people do, to the screen, the meeting room, negotiation, group chats and family gatherings, and then to the choice of staying out. Fifty-three signs are carded, each with its look-alikes, a cheap test and a quiet move. A quick key of one page for each family of signs lets the reader find a sign at a desk, and eleven field sheets go with it. A ten-minute log tracks how often the first reading of a sign was right, and ten situations are rehearsed.

The negotiation case turns on silence. A price concession was made in 40 seconds, USD 1.30 a set on 640 sets, which cost USD 832 against a margin of USD 2,688. The book's point is that a negotiation is two people deciding who will speak into the next silence. The guide times the silence after a price in seconds. The book is careful to say that a sign is a reading to test, not a verdict, and each card gives a cheap test that could show the reading is wrong.

The legal table is the shortest in the series. The Personal Data Protection Act, 2022 is listed for an owner's private log about named staff, and is marked not reachable at the issuer on 25 October 2026. The Cybercrimes Act, 2015, and the communications regulator's Electronic and Postal Communications (Online Content) Regulations, 2020 are also marked not reachable, and whether a Zanzibar instrument applies is to be confirmed with an advocate in Zanzibar. The guide advises against recording calls and against copying or photographing group messages, and it keeps the log and any record about a member of staff to what the owner needs.

Auditing advice before money moves: Honest Critics, Paid Flatterers

Handbook 142 is written for the owner who has many voices around a purchase. Its case is a mill owner in Masasi weighing a used oil press at TZS 5,400,000. Twenty-three people spoke about it in the week of 5 to 11 October 2026, nineteen of them unasked, and the owner had asked four. The first sheet is a register: one line for each person who spoke, marked as asked or volunteered, with the stake that person carried. The share of voices actually asked is then a fact the owner can see, not a feeling.

Each adviser is scored on an eight-column scorecard. The columns run from aim, which asks whether a voice is aimed at the work or at the person, through how often a voice has disagreed in the last ten decisions, to whether the owner owes the voice anything, what its praise cost in a quarter, whether its claim has been checked with local people, and whether the advice has survived a trial. The scores add to a total from 0 to 16, and the total falls into one of four bands for how far to rely on the view. The flip test asks one question twice, with opposite tilts.

The quarterly ledger prices praise. The cost of praise in one quarter in the case, counted as money, gifts and hours, was TZS 3,360,000. Of that, TZS 2,340,000 bought agreement and TZS 1,020,000 bought reach. Column E scores each voice by its cost as a share of that quarter's profit, with under 1 percent scoring 2. The book then asks who is being persuaded. Every purchase has a payer, a user, a gatekeeper and onlookers, and in one case eleven of seventeen hours of persuading went to onlookers and two to the payer.

Before money moves, the owner asks twelve local people five plain questions about last time, and sets the answers against the seller's claim. The book's column G scores 2 when a claim has been checked with at least five local people the adviser did not choose. Ten tests with stop signs follow, and the scorecard scores a small trial 2, a paper check 1 and no test 0. In one case a second lorry stopped at three of ten tests, passed five and left two to run. The book closes with a circle of three advisers, a reply clock for public criticism, a one-page rule for who may speak, and a quarterly review that re-scores every voice on the same eight columns.

The legal passages are limited to two areas. The Prevention and Combating of Corruption Act, 2007 (Act No. 11 of 2007, Chapter 329) was read in part on 26 October 2026, and no section or penalty is printed. The communications regulator's Online Content Regulations, 2020 (Government Notice 538 of 2020) were read in part, together with a 2026 amendment cited as Government Notice 158i of 2026 and said to be in force from 1 July 2026, although the 2025 amendment page could not be opened. Defamation law was read second hand only. The Zanzibar Anti-Corruption and Economic Crimes Act, 2023 (Act No. 5 of 2023) was read at source, and Mainland instruments are not treated as standing in for it.

Mainland and Zanzibar, kept apart

Every book keeps Mainland Tanzania and Zanzibar separate, because each has its own law, offices and registries. The series never applies a Mainland figure to a Zanzibar workplace or the reverse, and it says where a row for the other side was not read. Handbook 137 shows the cost of the mistake. A Zanzibar car hire office followed the Mainland rule of twenty-eight days' notice, a rule that the book says was lawful in the doing but cost more than the Zanzibar card required. The lesson is a file for each side, marked with the card that governs it.

Handbook 135 reads the Mainland employment rules and states that the Zanzibar rows were not read for this edition. Handbook 136 reads the Mainland hours rules from the Act on TanzLII and the Zanzibar hours rules from the 2005 gazette, with later amendments not checked. Handbook 137 holds the two sides on separate cards: the Mainland Act as enacted and the Code as quoted in judgments, and the Zanzibar Act as read to section 88, with later amendments not checked. Handbook 138 names Mainland instruments and states that the Zanzibar employment rules on discipline and dismissal are not stated. Handbook 139 marks Mainland data, cybercrime and online content instruments as not reachable at the issuer, and records no Zanzibar instrument as read. Handbook 142 reads the Mainland and Zanzibar anti-corruption statutes separately and refers defamation to an advocate admitted in the place concerned.

Figures and rules to confirm

The series prints no government fee, rate or penalty, and where a book touches a receipt, a licence or a registration, it says so. The figures that do appear are illustrative cases worked from stated inputs. The reader should still confirm five groups of material before any step rests on it.

The first group is numbering. Handbooks 135 and 137 cite the Employment and Labour Relations Act as enacted, so that notice is section 41, final payments section 44 and unfair termination section 37. The TanzLII consolidation numbers the same sections one higher from section 13 onward. Handbook 138 cites fair reason and fair procedure at section 38, which is the consolidation number. A reader who opens either text should match the number to the edition in hand, and should not mix the two.

The second group is the 2025 amendments. The Labour Laws (Amendments) Act, 2025, which the book says took effect on 14 March 2025, was read only through summaries. The compensation bands and the bar on discipline once a dispute is filed rest on those summaries, and two summaries give different upper bands for the both-grounds case. An advocate should confirm the band before any dismissal decision is priced.

The third group is instruments whose status differs between books. The Personal Data Protection Act, 2022 is read in Handbook 137 but listed as not reachable in Handbook 139. The Cybercrimes Act, 2015 is read at source in Handbook 138 and not reachable in Handbook 139. Each book states its own reading, and neither binds the other, so the reader confirms both with the issuer.

The fourth group is Zanzibar. The Employment Act, 2005 is read in the gazette only as far as section 88 in Handbook 137, and its later amendments were not checked in Handbooks 136 and 137. The Zanzibar anti-corruption statute of 2023 was read, but its authority's own site was not. The fifth group is the regulator's 2026 amendment, cited in Handbook 142 as Government Notice 158i of 2026. It should be confirmed with the communications regulator before any paid promotion is built on it.

Reading order

A reader with staff or partners who might leave should begin with Be Ready to Lose Anyone (Handbook 135), then build the key-person plan and the Week-One Sheet. A reader who is about to hire, or who has a first breach on file, should go to Hire for Character, Fire Fast (Handbook 137) and confirm the Mainland or Zanzibar card before the first warning is written.

A reader who is asked for money, time or favours every week should begin with No Is Not Disrespect (Handbook 136) for the three-part reply and the yes budget. A reader who answers the first impulse in a quarrel, a late payment or a rumour should read Calmness Is Mastery (Handbook 138) and start the thirty-day log, since the second count is the main test of the method. A reader who talks first in every meeting should read Say Less, See More (Handbook 139). A reader about to spend a large sum on advice should read Honest Critics, Paid Flatterers (Handbook 142) before the money moves.

Each book stands alone, though all six share one discipline: count before deciding, and confirm the rule before acting on it. The reader should note the status of every legal row in the book in hand, and take the rows marked S, C or not reachable to the issuer or an advocate before a step rests on them.

Think Like the Owner is a series of six Zatra Handbooks, numbered 143 to 148, each written for one decision that an owner or an employee in Tanzania has to make about money. The books do not teach a method in the abstract. Most follow one person or one business across a set period, usually ninety days or a quarter, and record what happened week by week. The working is shown in full, so that a conclusion can be checked against the same records that produced it. The six books cover different ground. One asks whether a salaried worker should fund a first business from the pay date and when to leave the job. One prices the money beliefs that keep income small. One costs supply businesses for the mining, farming and building trades. One puts a price on the gifts and samples that a counter hands out without a bill. One rehearses twelve real company crises as decisions. The last separates what an advocate, an accountant and the owner should each decide. Positions are stated at 25 October 2026. Every money figure is illustrative and worked in US dollars, and where a shilling figure appears the book states the rate it uses. Where a rule differs between Mainland Tanzania and Zanzibar, the books keep the two apart. This summary explains what each book covers, the order in which a reader might take them, and the figures that must be confirmed with the issuer before anyone acts on them. It is general business education. It is not legal, tax, accounting or investment advice.

One method across six books

The series rests on one premise: a belief about money can be priced against records the reader already holds. A payslip, a till count, a bench sheet, a call card and a decision log are all ordinary documents. Each book asks the reader to use them before deciding. The result is a set of figures that can be argued with, rather than an opinion about how money ought to behave.

The books share a common structure. Most are built on a dated diary, a set of costed models, a run of real or composite cases, or a desk of working sheets. Every case is worked from stated inputs and carries its own arithmetic. Each book closes with a table of what was read and what to confirm, and each states that its tools may be copied for the reader's own use but may not be sold, bundled or re-branded.

A second habit runs through all six. The reader writes the test, the number or the rule down before the evidence arrives. Book 1 sets its eight tests on Day 1. Book 5 requires a call card with a figure, a date and a way back before the outcome is read. Book 6 keeps a decision log written before the result is known. The discipline matters more than any single figure, because it stops the reader from changing the test when the answer proves inconvenient.

Book 1: The Paycheck That Steals Destiny

Handbook 143 follows a layout and customer coordinator at a commercial printing press in Vingunguti, Dar es Salaam. She nets USD 320 a month and wants to start a design business without losing her job. The diary runs from 19 July to 17 October 2026. The book is set in Mainland Tanzania, and it treats the salary as both a funder for the first business and a meter for the hours the job claims.

The first rule is applied on the pay date. Of a gross of USD 410, tax takes USD 49, pension USD 41 and the net pay is USD 320. A savings rule of 15 percent moves USD 48 of that net on each pay date, before any spending. A six-month runway of USD 1,740 is worked from the household's costs. By Day 90 the runway stood at USD 528, which is why the book treats the runway as the first test.

The job itself is measured. Hours are sorted into paid hours, unpaid extra, commuting and sleep. The diarist's real hourly pay is USD 1.82 on paid hours and USD 1.32 on every hour the job claims, and the job takes 55 hours a week with the commute. The contract is read clause by clause. Clause 11 requires the employer's prior written consent to any other paid work, whether or not it competes. The owner replied in writing within five days and consented to design work on six conditions, with a review in January.

The side business is tested for ninety days. It brings in USD 348 from 14 jobs, with 153 hours logged and USD 237 kept after costs. Five offers for a first business are scored on four tests before one is chosen. Two of the five pass all four, and both are design only, with no stock and no machine.

The diary also tests the pension. It records USD 41 a month paid into the pension account, but it does not assume the employer's share, because the diary does not state it. A voluntary contribution is treated as a figure to be priced with the issuer before it is relied on. A colleague who left a job paying USD 560 a month with USD 1,400 in cash shows the cash trap that follows a resignation made before a runway exists.

Eight tests written on Day 1 are scored on Sunday 11 October. Four pass: direction, repeat and referred work, written terms and hours kept. Four fail: the runway, the profit of the last thirty days, the price of cover after leaving and the waiting work. Under the rule written in advance, a score of four means the pay date stays. A re-score is set for Saturday 16 January 2027. The book's projections for later dates are worked examples, not records, and the reader should treat them as such.

Book 2: Broke Mindset: The Real Culprit

Handbook 144 tests twelve money beliefs that keep income small when money arrives. Each belief is stated in the holder's own words, such as 'I will save when I earn more' or 'one good sale will set it right'. Each is tested in four parts on records the holder already has, and each ends with a verdict that keeps the part that is true and names the part that is not.

The illustrative cases put a measured cost on each belief. The book states the cost in US dollars and as a share of the owner's own income. In its worked cases the twelve beliefs cost between 1.5 and 25.7 percent of income, from a better jacket and phone at the low end to a start that is put off at the high end.

Seven habits are named, and each is matched to one practice. The habits are a plan too small to matter, spending set by others' expectations, avoiding the numbers, waiting for the big break, borrowing for what should be saved for, keeping money problems from the people who share them, and counting what shows instead of what stays. The practice that replaces them is a weekly bench of forty-five minutes, with a twenty-minute version for a week that cannot hold the longer one.

The ninety-day workbook follows one household through thirteen benches, from Sunday 12 July to Sunday 4 October 2026. Its baseline was USD 3,690 for a year, 24.4 percent of its income. Its Day 90 run-rate was USD 1,972, or 13.0 percent. The book describes that as a run-rate over a short period, and it says the figure says nothing about the year to come. The reader should take it as the shape of a result that a household might look for, not as a forecast.

One chapter is about two partners. Each kept a sum from the other out of care, and 159 days of silence cost USD 140 in a missed supplier discount and six days without sugar and oil. The chapter shows that a money belief can be shared as easily as it is held, and that the cost of not speaking is measured in the same dollars as any other habit.

The tools are ten Bench Sheets, each with a worked column and a blank column for the reader's own figures, together with a scoreboard that keeps measured figures apart from planned ones.

Book 3: Sell to the Miners, Diggers and Farmers

Handbook 145 asks whether to supply the people who dig, plant and build, or to join them. It costs eight supply businesses across seventeen plates. Each plate reads a pit, a field or a trench by customer type, and each rule carries a status mark. Eight small plates repeat the sums in other places in Kenya, Uganda and Tanzania, and a gazetteer lists 49 products with their rule marks.

The book opens with a choice of two doors. The saying that suppliers get rich in a rush is traced in print well after the rush it describes, and it names no author, no sum and no customer who failed to pay. What it leaves out is the clock, the margin, the loss line and the rule, which the plates then supply for each product and place.

The central model is a pit shop at Rwamgasa. It opens with USD 3,600 and earns USD 410 a month after paying its owner USD 150. That holds only while no more than 22.1 percent of what crews owe goes unpaid. Because the shop sells to crews that buy before they earn, its profit is partly the return on a loan book. Other models follow the same logic. An engine bench at Katoro opens with USD 5,750 and earns USD 702 a month. A pump pool at Ushirombo opens with USD 3,100 and earns USD 267 a month at the 57.7 percent use the model assumes.

Credit terms change the answer. A tractor-parts van at Sikonge that sells half its work on account for 120 days earns USD 755 a month and ties up USD 25,550. The same van on tighter terms, with 25 percent on account for 60 days, earns USD 1,023 a month and ties up USD 15,500. A USD 300 default at a 24 percent margin needs USD 950 of new sales to win back.

The stress test is a rush that ends. When sales fall 40 percent for three months, four of the eight monthly models fall below zero. The pit shop's profit falls from USD 410 in the model month to USD 2 in the first month of such a rush. The book closes with a fourteen-day plan for the first ten customers, each offered a small test order. A pit shop of the modelled kind breaks even at about five customers.

The book keeps Mainland Tanzania and Zanzibar apart. It claims no Zanzibar mining rule, because it found no dedicated Zanzibar mining law when it was read. That point should be confirmed with the Zanzibar authority before it is relied on.

Book 4: Give Coffee for Free

Handbook 146 is set at a coffee and tea counter at the Mombo bus stand in north-eastern Tanzania. In its first counted week the counter gave away USD 39.42 at cost. That was 21.2 percent of the profit it could have kept, and most of it had no recorded reason. The book keeps the counter's giving ledger for thirteen weeks. Each hand-over is written at the moment it leaves the hand, at cost, with one of six reasons beside it.

Cost, not price, decides how generous a gift can afford to be, and cost differs by item. One extra mandazi needs one and a half paid ones behind it to repay its cost. One extra tea needs only half of one paid cup. A sample is treated as a small loan against a future visit. A ginger sample paid back in its second week with USD 0.36 to spare. A passenger sip never paid back.

A daily gift turns into the standard a giver is measured by, so removing it is felt as a loss rather than as the end of a favour. The book's remedy is scarcity. A gift that is rare, counted and explained is noticed in a way a daily gift cannot be, and it keeps its value to the receiver.

Recurring gifts receive a budget, a repay table and an end date before they start. Staff gifts are handled separately, because a bonus that is given every year becomes a wage in the staff's reading. Favours of time, credit or access are costed too, even though no cup changes hands. A stop is made by written notice, with a script and a monthly review. Gifts that belong to a town, such as a contribution to a funeral, are judged by a cap rather than by a return.

The first quarter closed with USD 445.66 given away at cost, 7.5 percent of the quarter's sales. The book does not treat that figure as a verdict on generosity. It treats it as a count, and it sets the next quarter's projection at USD 14.22 a week of known gifts. At the first week's rate, the counter would give away USD 2,049.84 a year at cost, before a single gift is questioned.

Book 5: The CEO's Seat

Handbook 147 rehearses the chief executive's worst days. The reader sits in the chief executive's seat for a week of twelve dated decisions. Each is a real company crisis, presented as a memo with its numbers and options, followed by a call card and then a reveal. The reader writes the call first, with a figure, a date and a way back, and only then reads what the company did.

The cases are set in a sequence of sectors, not in the reader's own business. They cover a payroll cut, a deposit-taking bank where money can leave in an afternoon, supplier credit that funds growth, a regulatory deadline, a product recall, a replacement product that fails, a profit figure that was too high, a bonus that rewarded units, an out-of-stock outlet, a price rise, a rival's public comparison and a public apology. Each case is worked against a dairy with a monthly turnover of USD 30,000, so the arithmetic is shown at the scale of a small business.

The payroll case shows why counting comes first. A cut of jobs pays for itself in months, and those months can be counted before the cut is made. The sheet tests the choice against a cut in pay at a dairy, where the answer turns on how long the shock is believed to last.

Each call is scored on four tests: number, worst case, first to hear and door. The book judges a decision by the information that was available on the day, and it sets luck aside. In the book's sample week the seat scored 75 of 96 marks, and the weakest column held 14 of 24. The table of twelve calls shows where a weakness recurs, which no single call can show.

The book then turns the method on the reader's own business. A case file of six cards is built at the reader's own size, with two sources for every figure and a note on how far each was read. A sixty-minute drill trains a team with roles, a clock and a sealed page, and is scored on the same four tests. The book puts the cost of the drill for seven staff at under USD 8. The status of every source is stated, and the book says it is education and not advice.

Book 6: Lawyers Draft, Accountants Count, Owners Decide

Handbook 148 is a desk manual for an owner who works with an advocate, an accountant or a tax consultant. It draws the line between what each is trained and licensed to do and what only the owner can decide, which the book names as markets, prices, partners and risk. The fourteen tabs are arranged in four parts, and the book states that it is financial and business education, not legal or accounting advice.

The licence line is taken from the Mainland texts. An advocate is checked against the Roll under the Advocates Act, Cap. 341. An accountant is checked against the register kept by the National Board of Accountants and Auditors under Cap. 286. Before an advocate is instructed, five things can be checked, beginning with the Roll and the current certificate. Zanzibar is kept apart, and points that the book could not read are marked CONFIRM for a Zanzibar advocate.

The working tools are fifteen desk sheets. Sheets 1 and 2 belong to the owner: the matter memo and the decision statement. Sheets 3 to 7 brief an advocate, and sheets 8 to 12 brief an accountant. Sheet 13 serves a matter that needs both. Sheet 14 compares quotes by scope as well as by price. Sheet 15 is the decision log. Six instruction letters and a set of drift files support the sheets.

The worked cases show the point. In one purchase the seller claimed USD 1,150 a month, and the statements supported USD 920. The book uses the gap to show that an accountant's job is to show the record, not to settle the owner's price. Fourteen questions in Tab J4 test whether advice stays inside the adviser's field, and the back matter adds twenty-four practical questions about running an engagement.

The decision log is the last tool. Hanifa's log for 2026, kept for a catering business in Mikocheni, holds a dated page for each decision that stays with the owner. Each entry is written before the result is known and records what was decided, who advised, what was assumed and when to look again. The book makes the point that advice fades within a season, and the owner's reasons fade faster still.

Where a question turns on a law it goes to an advocate admitted where the matter arises. Where it turns on accounts it goes to an accountant on the NBAA register. The book marks the audit independence question CONFIRM, because no rule on it was read for the edition.

Mainland Tanzania and Zanzibar

The series keeps Mainland Tanzania and Zanzibar apart, and the books say so in different ways. Book 3 states that Mainland and Zanzibar are kept apart throughout, and that a rule is printed only as far as it was read. Book 6 states that the law of advocates, companies and tax differs between the two, and it marks Zanzibar points for confirmation with a Zanzibar advocate.

Book 1 is set in Mainland Tanzania only. Its sources note that Zanzibar has its own Employment Act, 2005, so the clause analysis in that book does not carry over to a Zanzibar contract. A reader on Zanzibar should take the contract and the employer's rules to an adviser before applying the book's reasoning.

Book 5 treats the dairy's crises on the Mainland and, separately, in Zanzibar, and it marks each case with the status of the source it rests on. The reader should not read a Mainland answer as a Zanzibar answer, and should check any figure that depends on the island's own rules.

Choosing where to start

The books can be read in number order, which follows a worker's first question, then the first sales, then the running and advising of a business. Book 1 is the natural start for an employee who has an idea and a salary. Book 2 suits a person who is about to receive more money and wants to see where it goes. Book 4 suits a counter owner, and Book 5 suits an owner or manager who wants to rehearse a hard decision before one arrives.

Book 3 suits a reader near a mine, a farm or a building site who is choosing between supplying the work and joining it. Book 6 suits an owner who already works with professional advisers and wants to keep the decisions. Each book stands alone, but all six ask the reader to record first and decide second.

What the series settles

Across the six books the reader ends with a small set of decisions rather than general advice. A payslip is the first measure and the pay date is the first decision. A belief is a claim to be priced. A supply business is a credit book as much as a shop. A gift is a cost with an end date. A crisis is a decision to be rehearsed. A piece of advice is a letter that the owner must still decide upon.

Each book also gives the reader a way to test progress: a re-score date, a weekly bench, a quarterly count, a sixty-minute drill or a decision log reviewed on a stated date. The method does not promise a result. It makes the next decision smaller and easier to check, which is the purpose the six books share.

Figures to confirm before acting

Every money figure in the series is illustrative and worked from stated inputs. Positions are stated at 25 October 2026. Several books carry projections after that date. Book 1, for instance, gives earliest projected runway dates of Friday 26 February 2027 and Monday 31 May 2027, which are worked examples and not records. A reader should take the dated figure and the date that carries it, and treat any later projection as an assumption to be re-scored.

The catalogue description of Book 2 is looser than the handbook. It says the household cut its counted yearly cost from USD 3,690 to USD 1,972 over thirteen weekly benches. The handbook describes USD 3,690 as a yearly baseline and USD 1,972 as a Day 90 run-rate over a short period, and it says that the run-rate says nothing about the year to come. This summary follows the handbook.

The shilling rate is a book assumption, not a market rate. Book 3 uses TZS 2,600 per US dollar, and Book 6 treats TZS 2,600 as equal to USD 1 wherever a shilling figure appears. Anyone converting an amount should confirm the current rate with a licensed bank before relying on it.

Rules carry status marks. Book 3 uses READ, REPORTED, CONFIRM, NOT READ and STOP. Book 6 uses READ, SECOND HAND, HEADING ONLY and CONFIRM. A reader should give most weight to READ items and should confirm the others with the issuer named in the book. The Zanzibar mining point in Book 3 and the audit independence point in Book 6 are the two clearest items to take to the authority before any decision rests on them.