The Berlin company that set out to own the world’s takeaways, but never learned to run them as one, agreed on July 16th to be swallowed by Uber. It was 15 years old.
It always preferred the map to the meal. At headquarters, the wall that mattered was the one studded with markets – at its height, some seventy countries. Yemeksepeti in Turkey, talabat across the Gulf, PedidosYa through Latin America, Baemin in Korea, foodpanda from Hong Kong to Karachi, Glovo from Kyiv to Nairobi. To read the wall was to feel the pull of empire.
In its youth this was not foolish. It was, for its moment, rather clever. Food delivery is a stubbornly local business: the merchants, the riders, the payment rails, the regulators and the appetites all differ by the city. Buying up local champions and leaving them largely to themselves – Yemeksepeti to run Turkey, talabat the Gulf, Baemin Korea – was a defensible way to go global in an age when capital was cheap and the world’s delivery markets were still a scatter of unclaimed territory. Its most faithful patron was Prosus, whose own history had taught it that a sufficiently good portfolio could excuse a loosely governed one. Delivery Hero listed in Frankfurt, touched €177 a share in early 2021, and spoke the language of the era – gross merchandise value, total addressable everything – with real fluency.
Then the business it was in quietly changed beneath it. Food delivery stopped being a marketplace that matched hunger to restaurants and became something harder: a logistics and technology operation, won or lost on dispatch, rider density, merchant tools and the daily grind of shaving seconds and cents. Capital, once free, turned expensive. And in the markets that mattered a different kind of competitor appeared – Meituan in China, Coupang Eats in Korea, Grab and Shopee in Southeast Asia – companies whose operating systems had been beaten into shape by years of close-quarters combat at home. They too had grown by acquisition; Grab had swallowed Uber’s own Southeast Asian business. The difference was not that they bought less. It was that they had built an engine capable of running what they bought.
This was the flaw that turned a virtue into a fate. Delivery Hero had assembled its federation before it had built an operating system strong enough to bind it. Headquarters knew how to acquire a champion; what it never mastered was welding a shelf of them into a single company, still less rescuing one that had begun to slip. Its reflexes remained those of an owner rather than an operator. It intervened in products and markets it knew largely through dashboards, while country managers learned, in time, to stop arguing and comply. The federation that had made expansion possible was making integration impossible.
The evidence arrived market by market. In Hong Kong, where Meituan’s Keeta made its first foray outside mainland China in 2023, the newcomer overtook the incumbent foodpanda within a year. In Southeast Asia foodpanda’s share slid from around 22% in 2021 to roughly half that by 2025, overtaken by a food arm bolted onto a shopping app. The company that had collected the most flags was being out-worked beneath too many of them.
The reckoning was dressed, as these things are, as a deal. Uber – which had once fled both Southeast Asia and Hong Kong – offered €41.50 a share, valuing the whole at $14.8bn: less than a quarter of the 2021 peak, and less than the market value of Grab, a company Delivery Hero had once dwarfed. Fifty markets would go to Uber; fourteen others to a New York investment firm charged with finding them, in the release’s own phrase, “long-term homes.” Even the jewels flattered the diagnosis. Korea’s Baemin, the largest single prize, sat in a market worth more than all of Southeast Asia combined. The Gulf platforms, talabat and Hungerstation, were the most profitable of the lot – owing much to favourable market structures, high order values and thin competition, but owing something, too, to the fact that the decentralised model genuinely worked when Delivery Hero owned the undisputed local champion. By then it was good at buying winners, but no longer good at making them.
There were the customary courtesies. Niklas Östberg, the co-founder already on his way out, thanked the team for fifteen years. Uber pledged not to disturb the Berlin headquarters until 2029, and to invest €2bn in Germany. Delivery Hero did not deliver food in Germany.
In the end it was not outbid. It was out-operated. Uber will inherit the apps, the users and the hard-won market positions. Delivery Hero never found a way to turn them into one company. Whether Uber can is now its problem.
