Table of Contents
ON A WEEKDAY morning in Lagos, somewhere between the time a customer downloads a fintech app and the moment she taps "deposit", her face is photographed, matched against a national identity database, scored for liveness, and either approved or rejected. The whole exchange takes a few seconds. It costs the fintech a few cents. Somewhere in the middle, a company called Smile Identity earns a tiny piece of the transaction. By early 2023, Smile had completed an estimated 50 million such checks across the continent; the figure today is many multiples higher. None of those numbers individually moves a market. Together, they describe one of the most overlooked business models in African technology: data as a product.
The phrase is fashionable but slippery. It does not mean a company that uses data to build something else — a marketplace, a lending app, a logistics platform. It means a company whose product is the data: collected, structured, packaged and sold by the call, the record or the subscription. The economics, when they work, look more like software than services. Gross margins are high. The marginal cost of one more verification, one more credit check, one more sector report is close to zero. The moat widens with every customer, because every customer feeds the dataset.
That, at least, is the theory. Africa's reality is messier — and more interesting. A handful of segments have produced businesses with real recurring revenue and credibly defensible moats. Several others have raised money, launched products and discovered that the buyer they imagined either does not exist or will not pay what the unit economics require. The gap between the two halves of this category is now wider than the gap between fashionable sectors such as fintech and edtech. Investors who treat "African data companies" as a single trade are buying a basket of very different bets.
Tier one: when the regulator does the selling
The clearest commercial success belongs to identity verification. Smile Identity, founded in 2017 and headquartered in Lagos, has raised about $31 million to date, including a $20 million Series B in 2023. Its customers are banks, fintechs, mobile-money operators, gig platforms and, increasingly, sectors as varied as agriculture and e-commerce. In the year before its Series B the company doubled its customer base and tripled revenue. Youverify and Dojah, both also Nigerian, compete in the same lane. Together they have turned a regulatory obligation into a recurring revenue line.
The economics of the segment are flattering for three reasons, and they are worth stating plainly because they explain why this tier works and others do not. First, the buyer is a regulated entity. Banks and fintechs do not run verifications because they want to; they run them because a central bank requires it. Compliance budgets are durable in a way that marketing or R&D budgets are not. Second, the verification feeds a high-frequency operational decision — onboard this customer, yes or no — rather than a strategic one. High frequency means recurring revenue rather than annual contract negotiation. Third, the unit of consumption is small enough to price per call. A fintech with a million new sign-ups can pay a few cents a check without flinching; a few cents a check, multiplied across the continent's onboarding volumes, builds a real business.
Strip any of those three conditions out, and the model gets harder.
Tier two: the credit middle ground
Indicina, also based in Lagos, sits one tier down. The company sells alternative credit-decisioning to lenders that cannot rely on Africa's thinly populated bureau infrastructure. The proposition is sharp: in Nigeria, credit-bureau coverage is roughly 14% of the adult population, leaving the rest invisible to traditional underwriting. Indicina analyses bank statements, applies machine-learning models and returns a score. By 2022 it served around 120 customers including banks, non-bank lenders and fintechs, with names such as Polaris Bank and CreditDirect on its roster; the company said it had helped clients process $3 billion in loan applications and disburse $700 million. Its seed round of $3 million was led by Target Global, with Greycroft participating.
The customer is again a regulated lender. The decision is again operational. But two of the three flattering conditions weaken. The buyer universe is smaller — there are fewer lenders than fintechs — and the value proposition collides with incumbent bureaus and, increasingly, with banks' own internal data-science teams. Pricing pressure follows. Indicina's response, like that of several peers, has been to inch up the value chain: from selling data to selling decisioning, and from selling decisioning to running the entire credit workflow. The economics improve, but the business looks less like a pure data play and more like vertical software. That is a perfectly fine business. It is not the same business.
Tier three: the frontier
The hardest tier is the most ambitious. Stears, founded by four LSE and Oxford graduates in 2017, set out to build a Bloomberg-equivalent for African markets. It began with a $100-a-year consumer subscription that found significant traction among finance-sector employees. In 2022 it raised $3.3 million in seed funding led by MaC Venture Capital, with Serena Ventures participating, bringing total funding to about $4 million. Enterprise customers grew to contribute more than 75% of revenue, up from 45% the year before. By late 2023, the company had concluded that the individual-subscriber business was a distraction and pivoted entirely to B2B intelligence — market forecasts, consumer indices and macroeconomic datasets — for global institutions operating in Africa. The published customer list now includes the United Nations Development Programme, the European Investment Bank and Citibank Nigeria.
The pivot is rational and the customers are blue-chip. But the structural challenge is honest: Stears's buyers are doing strategic, lumpy, low-frequency work — not regulated operational work that recurs every onboarding. Contracts are larger, but they renew annually, not by the API call. The data is bespoke rather than commoditised. The model can be a good business; it is unlikely, on present evidence, to be a software-margin business in the way that identity verification can.
Sagaci Research, headquartered in Kenya, sells consumer-panel data to brands and investors and has been in the market longer than most. Amini, also Kenyan, is building geospatial datasets for agriculture and supply-chain monitoring and has attracted meaningful venture capital. Both are credible companies; neither has yet produced public evidence that the data-as-a-product model in their segments yields repeat enterprise willingness to pay at scale. The frontier tier may still get there. The honest reading of the evidence is that it has not got there yet.
The three-condition investors and operators must test before building a data startup
A pattern emerges. The African data-as-a-product businesses that have crossed into commercial maturity share three features. Their buyer is a regulated entity with a compliance budget. Their data feeds an operational decision taken many times a day, not a strategic decision taken once a year. And their unit of consumption is small enough that pricing per call or per record works. Where all three hold — identity, certain corners of credit — the unit economics look genuinely good. Where any one breaks down, founders find themselves discovering, on their customers' dime, what the buyer will and will not pay for.
For investors, the implication is unsentimental. The next breakouts will probably emerge where the three conditions can be reassembled: KYB (know-your-business) checks for the corporate market, alternative data for insurance underwriting, anti-money-laundering screening as continental regulation tightens. These are all variations on the identity playbook, not departures from it. For founders chasing the frontier — geospatial, market intelligence, consumer panels — the harder honesty is that selling proprietary data to a small number of large institutions is a consulting business in software clothing. It can be a fine business. It is rarely a venture-scale one.
Africa's data merchants are at last building something durable. They are also, quietly, drawing a line between what the continent's enterprise buyers will pay for and what it merely wishes existed. Investors would do well to learn the difference.
