Tyme: the free account is the funnel, the lending book is the business.

The thesis

Banking has always carried a brutal customer-acquisition problem. Branches, staff, paper, compliance: incumbents spend heavily to open an account, then recover it slowly through fees. Tyme inverted that. It put kiosks inside Pick n Pay and Boxer grocery stores, later TFG clothing stores, and let South Africans open a full bank account in around five minutes, walking out with a personalised Visa debit card. A 2022 World Bank case study put onboarding cost at roughly US$3 per customer through the kiosk, and around US$0.60 through the web. The account itself charges no monthly fee. That combination bought Tyme one of the fastest customer-acquisition curves banking has seen: one in five eligible South Africans held an account within five years of launch.

But an account acquired for three dollars earns nothing by itself. A free account is a funnel, and a funnel is a cost, however cheap. The business is what happens behind it: the deposits that fund lending, the transaction fees that tick over, and above all the credit book. Tyme's most interesting move was recognising it did not have to build that engine. In December 2022 it bought one, paying R1.5 billion for Retail Capital, South Africa's largest fintech funder of small businesses, and bolting a decade-old, proven lending machine onto a two-million-a-year customer funnel.

What they sell, and to whom

Two customers, two products, one balance sheet.

• Mass-market consumers. The zero-monthly-fee account, high-interest fixed deposits (Tyme paid South Africa's best retail rate in 2023), a buy-now-pay-later product (MoreTyme), and salary and grant advances. Distribution rides on partners: kiosks in grocery and clothing stores, cash in and out at more than 170,000 informal spaza-shop tills through the Kazang network. The customer Tyme serves is largely the customer traditional banks priced out.
• Small businesses. Through the Retail Capital division: merchant cash advances. A business with card turnover gets a lump sum, typically within 48 hours, no collateral, and repays a pre-agreed total as a percentage of its daily card takings (or fixed instalments) until the amount clears. Repayment speeds up when trade is good and eases when it is quiet, which is precisely why informal and seasonal businesses accept it.

The revenue engine

Three layers, in ascending order of interest.

• Transaction income. Interchange and fees priced 30 to 50 percent below the incumbents, by Tyme's own account. Deliberately thin: the price gap is the marketing.
• The float. R7.7 billion of deposits at June 2025, up 22 percent in a year, funding the balance sheet the way deposits always have. Cheap, sticky funding is the quiet advantage every retail bank is really built on.
• The lending book, especially the merchant cash advances. By October 2024 the group had advanced a cumulative R12 billion-plus (over US$600 million) to more than 80,000 small businesses, counting Retail Capital's pre-acquisition decade. Note the distinction that matters: cumulative disbursements are the odometer, not the engine. The outstanding book is only around R2 billion at any moment. The same rand goes out, comes back over months, and goes out again.

That recycling is the point most readings miss. In short-cycle lending, the yield on the book is set less by the price of any single advance than by how many times a year the capital turns. Run the arithmetic: a R2 billion book that turns over twice a year writes R4 billion of advances at a given margin; the same book turning once writes half as much at identical pricing. Tyme's disclosed figures imply something between those speeds (roughly R2.5 billion of new disbursements in the year to October 2024 against a book of under R2 billion), and the exact turnover is not published. The mechanic, not the precise number, is the lesson: in this kind of lending, velocity is yield.

The economics, read as an owner

The metric that matters here is not the customer count. It is revenue per active customer, and the word doing the work is "active". Tyme reported 11.6 million SA customers against R3.1 billion of annual revenue at June 2025: roughly R270 per registered customer per year. Capitec, the benchmark for low-cost banking at scale in South Africa, reported 26 million active clients and R16.8 billion of headline earnings (profit, not revenue) in its year to February 2026. That distance is the whole story. Tyme has built the funnel of a giant and, so far, earns like a challenger; the registered-versus-active gap (the company has claimed around 70 percent 30-day activity; independent studies earlier in its life found much lower engagement and low balances) is where those two facts reconcile.

The honest reading of profitability cuts both ways. Tyme announced that TymeBank reached monthly profitability in December 2023, faster than almost any standalone digital bank on record, in a world where fewer than 5 percent of neobanks make money at all (Simon-Kucher's research). Both of these are also true: the audited statements for the year to 30 June 2025 record a net loss of R218.8 million, and they note that continuing as a going concern beyond October 2026 depends on continued shareholder support, further capital and executing the plan. That is not a hidden scandal; it is the normal cost curve of a bank that keeps buying growth and regulatory capital with shareholder money, and the group raised $250 million from Nubank, M&G and existing holders in December 2024 precisely to keep feeding it. But it is a discipline worth naming: read the milestone against the audited statement. A profitable month is a waypoint. A profitable year, sustained, is the arrival, and that part is not yet done.

What would break or reshape the model

• Credit losses in a strained economy. The merchant book repays out of daily card takings. That makes repayment self-adjusting, and it also means the book breathes with South African small-business health. Loss rates are not disclosed; for an owner, that undisclosed number is the one to ask for first.
• Capital dependency. Lending growth consumes regulatory capital, and the going-concern note makes the dependency explicit. The model works while shareholders keep writing cheques on the way to scale; the 2028-ish New York listing ambition is the planned refinancing of that dependency.
• Rented distribution. The kiosks sit in someone else's stores, cash access runs over someone else's till network, and onboarding leans on the national identity database being up. Every one of those is a partner dependency that a branch network, for all its cost, never had.
• Competition from both ends. Capitec already runs the low-cost playbook with 26 million active clients and a 31 percent return on equity, the incumbents are cutting fees, and specialist merchant funders (Lula, Merchant Capital) contest the SME lending space. Cheap acquisition is an advantage only while it converts to earning customers faster than rivals can respond.

What I'd copy, what I'd avoid

Copy

• Buy the engine if you already own the funnel. The Retail Capital deal attached ten years of lending competence, data and book to a distribution machine that was already running. Building that from zero inside a young bank would have taken years and produced worse loss curves.
• Match repayment to the customer's cash flow. The percentage-of-takings structure is why small merchants who would default on a fixed instalment repay an advance. Design the collection mechanic around how the customer actually earns.
• Treat acquisition cost as the start line, not the victory. Tyme's kiosk economics are a real advantage. The equally real work is activation: moving the acquired customer to deposits, transactions and credit. Cheap CAC moves the problem; it does not solve it.

Avoid

• Reading customer counts as value. Registered is not active, active is not earning. Any business that celebrates signups should be made to show revenue per active user in the same breath.
• Reading cumulative disbursement as a book. "R12 billion advanced" and "R2 billion outstanding" are both true and describe different things. The odometer is not the engine.
• Reading one profitable month as a profitable business. Milestones are marketing. Audited years are accounting. Judge lenders on the second.

On the figures: Customer, deposit, revenue and loss figures for the SA bank come from TymeBank Limited's audited annual financial statements to 30 June 2025 (revenue R3.1bn, net loss R218.8m, deposits R7.7bn, 11.6m customers), including the going-concern wording summarised above. The December 2024 round ($250M at $1.5B, Nubank $150M for ~10%, M&G $50M) is from company statements and press reporting. Onboarding costs (~US$3 kiosk, ~US$0.60 web) are from a 2022 World Bank case study and are dated. The R12bn+ cumulative advance figure to 80,000+ businesses (October 2024) includes Retail Capital's pre-acquisition history; the ~R2bn outstanding book is from reported statements. Book-turnover estimates are my arithmetic from those disclosures, not company numbers, and Retail Capital's pricing, terms and loss rates are not public. Group customer counts (15M+ end-2024, 20M+ cited in 2026 interviews) are company statements. "Fastest to monthly profitability" is the company's framing of its December 2023 milestone. The rebrand to GoTyme Bank completed in early 2026.