Africa's venture record: a scorecard

1. B2B distribution and logistics hold most of the losses

Companies that buy goods and deliver them to small shops (Twiga, Wasoko, MaxAB, Capiter, Alerzo) or run trucks and delivery for them (Sendy, Lori, Kobo360). In every other sector combined, — of — failed or are distressed. Of all the money raised by companies that later failed or became distressed, — went to these two sectors.

So: supplying informal shops and moving their goods is where African venture capital lost the most, in every market that funded it at scale.

2. No company that sold for less than customers already paid has failed

Poor households often pay more than richer ones for the same thing: kerosene costs more per hour of light than solar, renting a motorbike every day costs more than paying it off, and a telecom's transfer fees take a bigger share of a small transfer. A company that sells the same thing for less does not have to create demand. Customers simply move money they already spend. Every one of these companies is still operating.

The opposite case is a company selling goods that informal markets already sell at the lowest price. It has no saving to offer, so it competes on price alone: — of — such companies failed or are distressed.

So: the strongest businesses here replace something customers already pay too much for. Asking customers to find new money for something new is the riskier bet.

3. Asset-heavy companies fail when the market price cannot cover their costs and they lack the pricing power to raise it

Owning trucks, warehouses or power plants means high fixed costs. That works only if the price the company can charge covers them. A price taker cannot choose its price: informal rivals with no fleets or warehouses set it at a level that covers their costs, not the startup's, and if the startup charges more, customers go back to the market stall. It loses money on every sale and has no way to fix it. The one survivor, —, takes the market price too, but as the lowest-cost producer its costs fit under it. Asset-heavy companies that set their own price did far better: — of — failed (—), neither because of its assets.

This is not an argument for asset-light models. Price takers that left the trucks and stock to others did better, but — of — still failed or are distressed, and — have published a profit. Taking someone else's price caps the margin, and a capped margin cannot produce venture-scale returns, however large the company grows. Most African customers are low-income households, so the Amazon playbook of thin margins at huge scale has not worked here: Jumia, the continent's best-known attempt and outside this sample, still loses money every year more than a decade after it launched.

So: the test is not how many assets a company owns but whether it can set a price that covers its costs. Without that pricing power, asset-light or not, it is at best a steady, low-margin business: worth backing with debt or private equity, which seek steady returns, but not with venture capital, which needs a few companies to return many times the money.

4. Selling an essential need does not protect a company when the existing price is already the lowest, especially in food and household goods

Food and household goods are as essential as energy, and demand for them never fell. But informal markets already sold them at the lowest price. That is why "disruption" failed here. Twiga used apps, data and its own trucks and warehouses to cut out the middlemen between farms and market stalls, yet the produce was already sold at a price informal traders had driven to the floor. However much technology a company deployed, there was no saving to offer customers and no margin left for it to capture.

Energy shows the other side. It is just as essential, but what customers already paid for it, kerosene and diesel, was expensive. Solar companies such as d.light, M-KOPA and Sun King could charge less than that and still keep a margin. Of the companies selling an essential need for less than customers already paid, — of — are still operating. Being essential guarantees demand; the price customers already pay decides whether there is money in it.

So: an essential need is an opportunity only where customers currently overpay for it. In food and household goods they rarely do: informal markets have already pushed prices to the floor, and no technology can capture a margin that is not there.

5. Low-income customers can carry a business when they pay in small, frequent amounts

Most customers in this sample are low-income households and individuals, or small traders, with little cash on hand at any one time. They can rarely pay a large sum up front, but they can pay a little every day, week or month, or a small fee on each transaction. Companies built around those payments failed far less often than the rest, where — of — failed or are distressed. Small, regular payments also make revenue predictable, so a company can borrow against it and survive a funding round that comes late. Of the traits that come from how a business is designed, this one separates survivors from failures the most.

So: with mostly low-income customers, how they pay matters as much as what they buy. A business that collects in small, frequent amounts is far more likely to last than one that needs a large single payment.

6. Companies that handle their customers' money survive far more often

A company handles its customers' money when its product processes their payments or lends to them: a wallet, a card machine, a loan, or a phone or motorbike sold on credit. Of the — companies that do, — failed or are distressed. Of the — that do not, — did. The gap holds within each kind of business: software, marketplaces and physical businesses all fail far less often when payments or credit run through them, as the product architecture section below shows.

So: the most useful question to ask of any startup in this market is whether money passes through its product and whether it earns on that flow.

7. Every company with a published profit earns by moving or lending money

They are —. Some are pure finance (payments, lending); the others sell a phone, motorbike or solar kit on credit and earn on the repayments. — of them, —, fund themselves by borrowing against the assets they finance or the loans they have made, and no company in the sample with that kind of debt has failed. No healthy company in distribution, logistics, health, software or consumer apps has published a profit.

So: in this market the product often wins the customer, but the financing makes the money.

8. The highest valuations go to payment networks, but most profits come from lenders

— of them have payments at their core. Only —, —, have published a profit. Meanwhile — of the profitable companies, —, have never been valued at $1bn. Kenya has — unicorns and — profitable companies.

So: a $1bn valuation shows that investors expect growth, not that the company makes money. The quieter lenders are the proven earners.

9. The funding pullback did not cause the failures; it exposed businesses that only survived on investors' money

Many of these companies spent each round covering losses that their own sales could not, so they needed the next round before the last one ran out. That usually took about three years. When investors pulled back in 2022 and 2023, the next round did not come, and — of the — failures ended at exactly that point: a round, loan or rescue that did not arrive, or bills and payroll that could not be paid. For these companies, more money would mostly have delayed the end. The same drought hit every company in the sample, yet every company that sold for less than customers already paid, and every company with a published profit, came through it. Companies whose largest round came in 2023 have only now reached the three-year mark: — of — have failed or become distressed so far.

So: a funding drought is a test, not a cause. Many good companies still need more money before they break even, so the question is not whether a company could survive without its next round. It is whether it is on a credible path to break even: whether each round brings it measurably closer to covering its costs, or only buys more time. When a failure is blamed on a funding drought, it often means the second. Investors had stopped believing they would get their money back, and another round would only have extended the runway to an end that was already coming.

10. About one failure in five combined a governance breakdown with a business that could not pay for itself

• Gro Intelligence (Kenya) sold agricultural and climate data to large companies. In 2024 its board replaced the founder; it could no longer pay staff and closed within months, amid reports that regulators were examining what investors had been told.
• 54gene (Nigeria) built a large biobank of African DNA to sell research access to drug companies. Deals came slowly while its labs and staff were costly; the founder left in 2022 amid leadership turmoil, and it could not raise again.
• Capiter (Egypt) grew in two years to thousands of staff and hundreds of trucks supplying small shops. In 2022 the board removed both founders, one later admitted it owed more than it owned, and buyers walked away amid legal fights. It was liquidated.
• Dash (Ghana) raised one of Africa's largest seed rounds to link mobile-money wallets across countries, but never earned revenue. Audits found its user numbers inflated and tens of millions of dollars missing; the founder was fired.

Governance did not create the weak business underneath; it hid it until there was no time left.

So: judge the people and the business model together. A governance breakdown is often a sign that the business underneath is weaker than it looks.