Within Proximity

On 2 September 2026, Uber switched off Nigeria. Twelve years after launching in Lagos, it emailed riders, shut the app the same day, and left. Uganda went dark at the same time. Tanzania had gone in January. Côte d'Ivoire the September before. Four African exits in twelve months.¹

My gut reaction was simple. Africa is not consequential to Uber. I would have wagered that the whole continent was about 1 percent of Uber's gross bookings. The numbers say that wager was generous. Uber did $193 billion in gross bookings in 2025.² The entire Nigerian ride-hailing market, every app combined, is worth around $450 million.³ If Uber had owned all of it, Nigeria would have been about 0.2 percent of the business. It owned a fraction of that.

Uber didn't leave because it ran out of money. It grew bookings 22 percent last quarter and passed $10 billion in annual free cash flow for the first time.⁴ It left because Nigeria was too small to matter inside a global portfolio. Market share was shrinking, the macro was unstable, and policy kept shifting. Nobody at headquarters could justify the next dollar.

So the more useful question is what it takes for a market like Nigeria to earn priority inside a global portfolio, and why that's so hard from a distance.

A seat at the table

The easy explanation is policy. Levies, data-sharing rules, airport bans, licensing. Nigeria has all of them. Every market has them, New York included. What decides the outcome is how much flexibility and presence an operator has to be in the room when they're written.

Look at New York. In 2015 the city proposed capping the number of ride-hail vehicles. Uber fought it in public and in City Hall, and the cap was dropped. When caps and a driver minimum wage came back in 2018, Uber negotiated inside them rather than walking away.⁵ The outcome was a middle ground. Government got what it needed. The business kept operating. Riders kept a service that, in a city like New York, is close to essential.

Now look at Lagos. In 2020 the state announced a 10 percent levy on every e-hailing trip. After operators sat down with the governor, it became a flat ₦20 per trip, housed in a Road Improvement Fund.⁶ That is what engagement buys. Four years later, the state asked for real-time trip data through an API, and Uber raised privacy concerns. The standoff disrupted drivers for about a month before the two sides reached a truce.⁷ Both positions were legitimate. A faster, earlier conversation might have saved everyone that month.

Notice what the government actually wanted. A wider fiscal base. Money for roads. Data to plan where to deploy resources. These are good intentions. African governments generally have them. What they often lack is a set of precedents to learn from. Most have few comparable cases and limited scope to experiment.

That is exactly what a global operator can bring. If you run in seventy countries, you know how a per-trip levy played out in one city and a commission cap in another. You can put the options on the table, with the outcomes, so the government is not legislating blind. That is a real advantage. But it takes people, time and senior attention to deploy, and those are allocated by the same portfolio logic as capital. A market that is a fraction of a percent of the business rarely gets enough of them.

Primary data

Good engagement starts before the meeting. It starts with primary data.

Primary data means not relying on what two or three people, three degrees removed from the problem, told you. It means speaking directly to the people on the ground and seeing what is actually happening. It has three layers.

Ground truth on the economics. How is inflation actually moving prices? What does a driver pay for fuel this week, and what does a rider accept as a fare? In Lagos, petrol went from about ₦800 a litre in late February to around ₦1,300 by mid-April this year.⁸ That's roughly a dollar a litre at today's exchange rate, and a monthly report can't keep up with prices moving that fast. Reading the market from the ground is the subject of The First 72 Hours in a New City.

Policy as it actually plays out. On paper, electric vehicles in some African markets carry no import duty. In reality, clearing those vehicles from the port can take three, four, five months. For all that time the capital is tied up and the vehicles earn nothing. You only find that out by talking to the people clearing them. It's part of why cheap cars are not the whole story.

A map of who holds influence. The ministry with the right title is not always the one that moves reform. In some countries, the investment promotion agency has more leverage than anyone else in government. If you know that, that is where you go. If you do not, you will spend years believing you are engaging the right people, or using the right data, when you are not.

Two more things matter, and both are about people. You need talent on the ground that is adaptable and flexible enough to build real rapport with the right stakeholders. And senior leadership has to get on the plane, as often as it takes, to engage at the very highest level. Nothing tells a government that a market is a priority like the people who make the decisions showing up in person.

Proximity

All of this comes down to one thing: proximity. If you run a business from San Francisco, or from Amsterdam, across six or seven time zones, you are far from the problems. Your data is secondhand and your relationships are thin.

Uber has made this trade before. Southeast Asia was very far away, and Grab was a very strong competitor. Ceding the market to Grab was an easy management decision. The same happened with Didi in China and Yandex in Russia and Central Asia. And there is a real logic to it. Who understands the lay of the land better than a business that was built there? I wrote about how those retreats played out in The Second Player Problem.

Latin America looks different, but only at first glance. Uber has been bullish in the very large markets: Brazil, Mexico, and Argentina. Brazil is its biggest market outside the United States, and Uber says Mexico led its international growth last quarter.⁹ In the smaller markets, Colombia, Peru, and the rest, it has been much more relaxed.

How should we read this? My hunch is that when the bulk of your business sits in the United States, you allocate capital purely on the economics. You go where the quick wins are and where the market is big enough to move the group number. São Paulo and Mexico City move it. Lagos does not. Nigeria was never going to clear that bar from that distance.

The challenger

The obvious objection is Bolt. It is run from Tallinn, about as far from Lagos as Uber's European base. Yet Bolt stayed, and by most estimates holds around two-thirds of Nigeria's ride-hailing market.³ So does proximity really explain anything?

I think it does, once you look at how much Africa matters to each company. Africa has always mattered far more to Bolt than it ever did to Uber. For a long stretch it was a meaningful part of Bolt's business, not a footnote. That changes how much attention a management team gives a market.

But look at how Bolt built its African footprint. Its strategy from the start was to be the challenger to Uber across the continent: the player 10 to 15 percent cheaper than the incumbent, wherever the incumbent was. In Nigeria it entered in 2016 charging drivers a lower commission than Uber, and drivers moved.¹⁰ The strategy depended on Uber. When Uber stopped expanding rapidly across Africa, so did Bolt.

The numbers since then point the same way. Bolt's revenue growth has slowed to 14 percent a year, against 20 percent-plus at Uber and Grab.¹¹ Its recent expansion has pointed at Europe; it is now in 26 of the EU's 27 member states. In 2023 it closed Bolt Food in South Africa and Nigeria.¹² To be fair, Bolt has publicly reaffirmed its commitment to Nigeria since Uber left.¹³

The same allocation logic that took Uber out of Nigeria applies to any operator whose main business sits somewhere else. If Africa's share of Bolt's business keeps falling, Bolt will face the same questions.

Where it goes

None of this means African ride-hail is broken. The business is there. It is one of the most mathematically balanced marketplace businesses ever built: a rider who needs to move, a driver who needs to earn, and a price that clears both. People across the continent want good mobility services, and they will keep wanting them whoever owns the app. Nigerian e-hailing alone is projected to roughly double by 2031.³

The next couple of years will be rough from an investment perspective, though. Capital that once flowed into African mobility on the strength of global names will be harder to raise now that the global names are leaving.

I expect Uber to hold on where it has a reason to, and the clearest reason is tourism. Uber's remaining African markets are South Africa, Egypt, Kenya, Ghana and Morocco, with Egypt reinforced by Uber's ownership of Careem's ride-hailing business.¹⁴ These are, broadly, the continent's tourist economies. Morocco received about 19.8 million visitors in 2025 and Egypt about 19 million.¹⁵ Kenya took in 2.4 million in 2024. A traveller arriving from London or New York opens the app they already have. For Uber, those markets are a service to its existing customers, not a bet on a new one. That makes strategic sense.

The pattern is not perfect, and it is worth being honest about where it bends. Three-quarters of South Africa's arrivals come from elsewhere in Africa, so South Africa is an Uber market as much for its own urban scale as for its tourists.¹⁶ Ghana is the one to watch. If the tourism logic holds, it is the next candidate for review.

Everywhere else, the void will be filled. Local players will step in. So will international operators who have chosen to make emerging markets their core business. That is where the real opportunity sits: operators for whom Africa is the core business, not a small part of it, have a genuine chance to consolidate the market Uber has left behind.

Scaling without burning capital

The operators who win the next few years will be the ones who figure out how to scale without burning too much capital. A lot of capital has already been burnt in Africa, by Uber and by Bolt.

Every marketplace subsidises early, to get riders to try the service and drivers to stay. That's fine. The trouble starts when you subsidise past the point where the math works. My test is simple. If you are spending a dollar to buy every dollar of GMV, you are not building a market. You are giving the service away, and the cohorts will break down the moment you stop.

Nigeria shows what that looks like. Trip data from roughly 20,000 Uber rides in Nigeria this year found that on trips under about 20 km, Uber paid drivers more than riders paid Uber, by as much as 23 percent on some distance bands.⁸ That's $1.23 spent for every dollar of GMV. And it ran in a market where an Uber driver completed about four trips a day, against nearly ten in South Africa. Uber raised its per-kilometre fare 71 percent between January and July, and the gap still did not close.

A capital-efficient operator does the opposite. It gets primary data and puts structures in place to keep getting it. It knows how much it can subsidise sustainably and stops there. It expands through levers that don't burn cash: partnerships, such as with the companies that finance drivers' vehicles, stakeholder engagement, and the right talent on the ground. And its senior management shows the market matters by turning up.

None of this costs much cash. It costs attention, and a market that small inside a global business rarely gets enough of it. Uber's exit says more about portfolio structure than about Africa. Ride-hail in markets like Nigeria rewards operators who are close enough to see what's happening on the ground, engage government early and subsidise with discipline. Those operators will build the market Uber left behind.