TymeBank: The Bank That Rented Its Branch Network

The most important South African fintech insight was about supermarkets, not software. This story covers the kiosk model, the retail partnerships, the cost structure, deposit gathering, lending economics, the international expansion and what the model does and does not solve — part of the South Africa Company Stories hub.

Why did digital-only banking struggle before this?

Because a large share of the target market is paid in cash, spends in cash and needs a physical location to convert between cash and digital balances. An app cannot accept a banknote.

Earlier digital banks in emerging markets consistently found that customers opened accounts, used them once and reverted to cash, because the deposit step was inconvenient or expensive at third-party agents.

The other barrier was trust. Opening an account with an institution that has no physical presence requires a leap that many first-time banking customers reasonably decline to make.

What does the kiosk actually do?

It performs the entire account opening process in a few minutes: identity verification against the national population register, biometric capture, account creation and issuance of a personalized debit card printed on the spot.

The physical card matters more than it appears. A customer who walks out holding a card has something tangible, immediately usable and demonstrable to family and merchants, which resolves the trust problem in a way an app download does not.

Placement inside a supermarket also puts the bank where the customer already is on payday, at the moment they are converting wages into groceries — which is precisely when a transactional account is most obviously useful.

Why is the retail partnership so valuable?

Because it delivers physical presence at near-zero marginal cost. Supermarket chains already operate thousands of stores with cash-handling infrastructure, security and staff, and adding banking transactions uses capacity that exists.

For the retailer the arrangement reduces cash in the till, generates fee income and increases footfall from customers coming in specifically to bank. The incentives align without either party subsidizing the other.

The comparison with incumbents is stark. A traditional bank building comparable coverage would spend enormous sums on property, fit-out, staffing and security for branches that would take years to reach break-even.

What is the cost-to-serve advantage?

The absence of branch infrastructure removes the largest fixed cost in retail banking, which allows the bank to serve customers whose balances and transaction volumes would be unprofitable for an incumbent.

That is the real market opening. Low-income customers are not unbankable; they are unprofitable at a cost structure designed around branches, relationship managers and legacy technology.

The corollary is that pricing can be genuinely lower — monthly fees at a fraction of incumbent levels — which is both the acquisition proposition and the reason incumbents cannot simply match it without cannibalizing their own economics.

How do deposits and lending fit together?

Deposits fund lending, and a bank paying attractive rates on savings gathers a stable, low-cost funding base without wholesale market dependence. Competitive savings rates have been a deliberate acquisition tool.

Lending is where profitability ultimately comes from. Transaction fees on low-value accounts cover costs at best; credit is where a bank earns a return on the customer relationship it has built.

The risk is equally clear. Lending to customers with limited formal credit history in a high-unemployment economy requires underwriting that works, and the graveyard of emerging-market lenders is full of businesses that grew a loan book faster than they understood it.

What is merchant lending and why does it matter?

Working capital advances to small businesses — spaza shops, market traders, small retailers — repaid from their transaction flows rather than against collateral or audited accounts.

It works because the lender can see the merchant’s actual sales through the payments they process, which is far better information than a credit bureau file and updates daily rather than annually.

It also addresses a genuine gap. Small merchants are chronically short of working capital, banks will not lend without security, and informal lending is expensive — so a data-driven product at a reasonable rate has real demand.

Why expand into Southeast Asia?

Because the conditions that make the model work exist there too: large populations with limited banking penetration, extensive cash usage, dense modern retail networks and mobile phone ubiquity.

The model transfers more readily than most fintech because its core asset is a partnership template rather than a technology, and retail chains in those markets face the same incentives as South African ones.

The strategic prize is scale. A banking platform amortized across several markets carries a lower cost per customer than one operating in a single country, which is the argument that justifies the capital raised at international valuations.

What are the risks in this model?

Dependence on retail partners is the first. The partnership is commercial rather than owned, and a change in the retailer’s strategy, ownership or terms directly affects the bank’s distribution.

Credit is the second. Growth in lending to a customer base with thin credit histories can look excellent for several years and then deteriorate quickly if unemployment rises or the underwriting was fitted to a benign period.

Regulatory capital is the third. Banking is a capital-intensive business, growth consumes capital, and a bank scaling rapidly across multiple jurisdictions needs shareholders willing to keep funding it until profitability arrives.

What did incumbents do in response?

Launched low-cost digital accounts of their own, reduced fees on entry-level products and expanded partnerships with retailers for cash transactions — a rational competitive response that also compressed their own revenue.

They face a structural difficulty. An incumbent matching a challenger’s pricing cannibalizes profitable existing customers while retaining the branch network the challenger never built.

The likely outcome is that entry-level banking becomes a low-margin utility in South Africa, competition moves to credit and value-added services, and the branch network becomes a legacy cost rather than an advantage.

What is the lesson?

That the barrier to banking the underserved was distribution cost, not technology or demand. Once the distribution problem was solved with someone else’s infrastructure, millions of customers arrived quickly.

The second lesson is that digital does not mean physical-free in a cash economy. The successful model is hybrid, and the businesses that insisted on app-only distribution learned this expensively.

The third is about partnership design. An arrangement where both parties gain independently — the retailer reducing cash and earning fees, the bank gaining presence — is far more durable than one requiring goodwill from either side.

What does customer acquisition cost look like?

Dramatically lower than incumbent banking. A kiosk in an existing store, staffed by an ambassador rather than a banker, converts walk-past traffic into accounts at a fraction of what branch-based acquisition costs per customer.

The comparison that matters is fully loaded. An incumbent’s acquisition cost includes branch rent, security, staff and technology amortized across the accounts opened there, which is why entry-level accounts have historically been loss-making for them.

The risk in cheap acquisition is dormancy. Accounts opened easily are also abandoned easily, so the meaningful metric is active accounts with regular transactions rather than the cumulative number opened.

How does a bank make low-balance accounts profitable?

Through transaction volume rather than balances. Interchange on card purchases, fees on selected transactions and the spread on deposits held collectively all generate revenue that individually looks trivial and aggregates across millions of accounts.

Cost discipline does the rest. Automated servicing, no branch overhead and a modern core banking platform without decades of accumulated legacy systems together produce a cost per account an incumbent cannot approach.

Credit is the eventual profit engine, which is why every low-cost transactional bank ultimately becomes a lender, and why the quality of that transition determines whether the model works.

What role does the reward programme play?

A significant one, because a bank distributed through supermarkets can offer rewards the retailer already runs, giving customers a benefit that costs the bank far less than a cash incentive would.

It also reinforces the partnership. A customer earning grocery rewards through their bank account shops more at the partner, which is exactly the outcome that keeps the retailer committed to the arrangement.

The limitation is that rewards attract customers who optimize for them, and building an account base on incentives risks acquiring users who leave when a competitor offers more.

Why is South African banking so concentrated?

Because four large groups have held the overwhelming majority of retail deposits for decades, protected by branch networks, payment system access, capital requirements and customers who rarely switch.

Regulatory reform lowered some of those barriers by licensing new banks, opening payment infrastructure and requiring clearer fee disclosure, which is what made challenger entry possible at all.

Concentration has not ended, and it has changed shape: incumbents retain the profitable customers while challengers take the entry-level market, which shifts where the competition happens rather than removing it.

What does financial inclusion actually require?

More than an account. A customer is included when they can receive income, store it safely, pay for things, transfer money to family and access credit at a price that reflects their actual risk.

Account ownership statistics overstate progress because many accounts are opened, used once to receive a grant or wage and then emptied, with the customer reverting to cash immediately.

The meaningful measure is active usage across several product types, which requires the account to be cheaper and more convenient than cash at every step — a much higher bar than opening it.

Frequently Asked Questions