Stop selling salary loans to people without salaries.

Look closely at the repayment architecture of almost every mainstream commercial bank loan and digital credit product in East Africa and you will spot an unspoken relic of the colonial civil service: the assumption that human beings earn money in predictable, equal slices on the 28th day of every month. It is a design fossil, preserved in core banking systems first written for salaried markets and imported wholesale. When a lender imposes a rigid 30-day bullet repayment or fixed monthly amortisation on an economy where 83.8 per cent of workers earn outside formal payrolls, the product design itself manufactures arrears. Lenders are selling a salaried debt schedule to people who earn by the trip, the crate and the market day, and this is not a uniquely Kenyan defect: it is the default architecture of consumer credit across the Global South, which is what makes fixing it a competitive weapon rather than a compliance chore.

Instant digital credit has undeniably solved a real liquidity gap. In a July 2021 study published in the Journal of Development Economics, economists Tavneet Suri, Prashant Bharadwaj and William Jack analysed administrative records and household surveys covering M-Shwari borrowers in Kenya, finding that access to fast digital loans made households 6.3 percentage points less likely to forgo essential medical or educational expenses during negative shocks. Emergency bridge liquidity works when a child falls sick on a Sunday night. When that same short-term, fixed-date structure is used as permanent working capital by a self-employed trader whose income fluctuates daily, convenience mutates into a compounding debt trap.

Deducting the cost of staying on the road before sizing the instalment.

Automated credit engines love scoring gross mobile-money inflows because wallet credits are easy to count. The gross inflow hides the relentless daily cash burn required to keep an informal enterprise open.

Step onto the kerb beside a boda boda rider in Kisumu collecting KSh2,000 in passenger fares across a twelve-hour shift. A naive scoring model sees KSh60,000 in monthly gross cash flow and approves a loan requiring a KSh18,000 repayment on the 30th. Now audit the rider's actual daily cash book.

Out of the KSh2,000 gross take, petrol swallows KSh650, the daily remittance to the motorcycle owner takes KSh400, stage welfare and municipal fees take KSh50, and a maintenance reserve for tyres, brake pads and engine oil requires at least KSh100. Only KSh800 remains at dusk to feed the rider's family and pay rent, and that KSh800 is not a guaranteed annuity: three days of heavy rain, a police crackdown or a punctured tyre drops daily net income to zero while the bike remittance and the household food bill keep running. Treating gross fares as disposable salary guarantees that when the fixed end-of-month debit hits, the rider either defaults or borrows from a second app to pay the first.

Pricing cash-flow flexibility on both sides of the balance sheet.

Matching repayments to how informal workers earn requires rethinking both product rules and treasury funding. A merchant or rider who struggles to assemble a KSh15,000 lump sum at month-end can often service KSh150 deducted automatically from daily till settlements, provided the schedule includes breathing room for dry days.

Designing that flexibility demands honest pricing rather than marketing slogans. When a lender offers a flexible repayment holiday that quietly capitalises penalty interest and rollover fees into the principal, a two-week illness can double the effective annual rate. Deducting a fixed 8 per cent slice of verified daily merchant till receipts, paired with three penalty-free pause days per quarter, delivers cleaner affordability than an expensive cash-flow model that misreads seasonal dips. A chatbot sending cheerful budgeting tips over SMS cannot fix an underwater daily margin, and an automated collections sweep that drains 100 per cent of a trader's wallet the moment a supplier payment lands will drive the merchant off digital rails and back into paper cash, destroying the very data trail the lender depends on.

Inspecting the shadow loan behind the clean repayment record.

Banks and fintechs keep renewing mis-timed 30-day loans for informal workers because the internal ledger shows the previous loan was paid on time. As the May 2021 Digital Credit Market Inquiry conducted by the Competition Authority of Kenya and Innovations for Poverty Action demonstrated, high repayment rates in short-term digital lending frequently mask multi-app borrowing across fragmented wallets and unlicensed lenders sitting outside a single credit bureau view. A borrower who takes out Loan B on Tuesday at 10.00 am to clear Loan A at 10.05 am looks like a prime, low-risk repeat customer inside Lender A's database, right up until her entire credit pyramid collapses. The industry's new single-portal credit information sharing architecture will expose some of that stacking; it will not change the product design that caused it.

In the first half of 2027, product and credit chiefs should replace calendar-date lending with turnover-linked amortisation across a controlled six-month cohort of 5,000 self-employed borrowers, comparing a traditional 30-day bullet loan against a daily till-percentage deduction with three fee-free skip days and frontline authority to grant verified 14-day illness or asset-repair moratoria. Tracking net portfolio yield after collection costs, alongside digital retention and multi-lender borrowing rates, will settle the argument quickly. Stop forcing a daily earner to counterfeit a monthly payslip just to access working capital.