Digital Lending Platforms

Why small loans stopped paying for themselves at a branch counter

The cost of putting a loan on the book barely moves with its size. The same application, the same checks, the same officer time and the same committee go into a small facility as into a large one. That cost is absorbed on a large loan and not on a small one, which is why so many institutions quietly decline small borrowers or price them in a way that sends them elsewhere.

The demand did not go anywhere. A trader restocking, a farmer buying inputs before a harvest, a salaried customer bridging a gap before payday. Small amounts, short terms, needed quickly. Wallet-based lending reached those customers first and taught the market what to expect: apply from a phone, get an answer in minutes, repay out of the same balance. Any institution entering that market now is judged against that experience.

What follows from it differs by institution. A bank is usually defending a retail relationship it already has and wants to lend into it at a cost that works. A SACCO often lends to its members already, on paper, with membership and payroll deduction doing much of the underwriting, so the gain is speed and control rather than reach. For a microfinance institution the whole book is small tickets, which makes the cost of processing the business model itself. A telco or fintech starts with distribution and needs the credit discipline around it.

The products stay yours

We do not arrive with a loan product. The platform runs the credit your institution designs and prices: salary-backed and check-off lending, group and solidarity lending, SME and working capital facilities, asset finance, overdrafts and revolving limits, and the short-term consumer credit that competes directly with wallet products.

Those products behave differently. Eligibility, tenor, pricing, security, guarantors, the shape of repayment (equal instalments, a balloon, a schedule aligned to a harvest) and what happens when a payment is missed vary by product and often by segment. Most institutions also expect the list to grow: launch one, learn from it, add the next. We scope on that assumption rather than on one product frozen at launch.

Credit assessment stays where it belongs. Who qualifies, what the scorecard rewards, where the cut-offs sit, which applications may be decided without a person looking at them, who may override and up to what limit, and what a declined customer is told, are decisions your risk function owns and signs off. We build to the policy you give us. We do not bring one with us, and we will not tell you who to lend to. Where that policy exists as practice rather than as a document, writing it down is part of getting started.

What a borrower does, and what your staff do

For the customer, a loan is one journey. They apply from wherever they already deal with you, see what they qualify for and on what terms, see the full cost before accepting anything, accept, and receive the money. Then they repay: on schedule, in part, ahead of time against a settlement figure, or in a single payment that closes the facility early.

A loan is more than its repayment schedule. Customers ask for a top-up, renew at the end of a term, ask to restructure when circumstances change, and are contacted before a payment falls due and again when one is missed. Handling those moments well is most of what separates a book that performs from one that does not.

Your staff have a parallel set of journeys: preparing and checking an application, working the cases policy sends to a human, approving within delegated mandates, authorising disbursement, working a list of accounts in arrears, restructures, recoveries and write-offs, and the reporting your credit committee and finance team read. Group lending adds the field side: meetings, attendance and the group guarantee. Which products and journeys go live first is decided with you, and most institutions are better served by taking one product to one segment first.

Where lending meets the rest of your business

Lending is a money movement at both ends. Loans are paid out and collected over the same payment arrangements you use for everything else: account transfers, mobile money wallets, agents, standing instructions, payroll deduction. Digital lending leans on them harder than most products, because for the customer a loan is largely the experience of receiving and repaying money.

It also depends on identity. You cannot lend to someone you have not properly identified, and the person verified when the account was opened should be the person on the loan agreement, without a second round of documents. Institutions that treat onboarding and lending as unrelated projects tend to pay for it later, in friction or in risk.

Whatever the customer sees, the loan is an asset on your balance sheet, provisioned and reported under the rules your regulator sets, so this is a finance and risk programme as much as a customer one. We build payments and onboarding platforms too, so lending can sit beside what you run today or be part of a wider programme.

What has to be settled before the first loan goes out

Much of what decides the timeline is not software. Credit policy and scorecards have to be signed off by risk. Product terms, pricing and fees have to be agreed, and they have to be ones you can publish and stand behind. Approval mandates have to be set. Those decisions usually involve more than one committee, and they are the most common reason a lending programme sits still.

Credit reference arrangements are yours to hold. Tanzania has a credit reference system, and the subscription, the terms and the basis on which you search and report are contracts between you and the bureau. Alongside that sits data protection. Tanzania has a Personal Data Protection Act, and the obligations it places on you when personal information is used in deciding whether someone gets credit are worth a legal view early rather than after the product is designed.

You need a position on automated decisions, and a settled collections process. How far may a credit decision be made without a person involved, for which products and amounts, and what is a customer told when the answer is no. Who contacts a borrower in arrears and at what point, when a case escalates to recovery or to legal, and when a loan is written off. Most institutions have a process on paper that differs from what happens in practice, and a digital book grows faster than a branch one. Deciding these is yours. Supporting them is ours.

Working with us

Discovery comes first: the book you have and how it is worked today, the products you want to run digitally, the policy behind them, and where the money will move. That produces a scope and a delivery plan, and the plan is yours whether or not we go on to build it.

Through delivery you see working software in an environment you control rather than status reports, and the source code and documentation are yours from the first commit. We will usually argue for a pilot on one product with a contained group of customers, because credit is unforgiving of assumptions nobody tested with real money. After launch we support the platform and train your people to take on as much of it as they want, and nothing in the arrangement locks you in.