The Jumia Story: The Amazon of Africa that might be neither
On the twelfth of April, twenty nineteen, a Nigerian e-commerce company called Jumia did something no African technology company had ever done. It listed on the New York Stock Exchange. Its shares opened at fourteen dollars and fifty cents. Inside seven days they touched forty-nine dollars and seventy-seven cents. The market capitalisation crossed three billion dollars. For a week, on paper, Jumia was worth more than every other African tech company combined.
Twenty-eight days later, a short seller in Los Angeles named Andrew Left ended that.
On the ninth of May, twenty nineteen, Left’s firm, Citron Research, published a twenty-two page report. The title was four words. Jumia is a fraud. Left wrote that in eighteen years of running Citron, he had never seen anything this obvious. The stock closed down eighteen point eight percent the same afternoon.
Three months later, Jumia disclosed, in its own quarterly earnings release, that an internal review had uncovered a fraud inside its Nigerian sales-agent network. Some of the orders the company had used to build its IPO prospectus had not been real.
This is the story of a Berlin-incorporated, French-founded, Delaware-listed, Dubai-managed company that was sold to Wall Street as the Amazon of Africa, and of what was left of that promise after fourteen years, three country exits, and roughly two billion dollars of losses.
The Berlin Factory
To understand Jumia, you have to begin in Berlin.
In the early years of the two thousand and tens, a German company called Rocket Internet had developed a business model that was less an internet company and more an internet factory. Rocket was run by three brothers — Marc, Oliver, and Alexander Samwer. Their method was simple. Find a successful American internet business. Copy it. Drop a team of MBA graduates into a region the original had not yet entered. Capitalise the clone aggressively. Scale it for two to five years. Then either take it public or sell it back to the company they had copied.
By twenty twelve, Rocket had run this play more than fifty times. The biggest win was Lazada, a copy of Amazon for South East Asia, founded that same year. Lazada would eventually sell to Alibaba in twenty sixteen for a billion dollars.
The Africa version was next. Rocket called the holding company Africa Internet Group. The model it was copying, again, was Amazon. In May twenty twelve, the holding company was incorporated in Berlin. In June, the first operating brand launched commercially in Lagos. It was, briefly, called Kasuwa, the Hausa word for market. Within weeks, the name had been changed to Jumia.
The team that ran it on the ground was four people.
The first two were French. Jeremy Hodara, twenty-seven years old, a graduate of the HEC Paris business school and a former McKinsey consultant in Paris, Mumbai, and New York. Sacha Poignonnec, twenty-seven years old, also HEC Paris, also McKinsey, also retail. They were the men Rocket had picked to run the African build. Hodara would handle operations. Poignonnec would handle strategy and, eventually, every Wall Street roadshow the company would ever run.
The second two were African. Tunde Kehinde, a Nigerian Harvard MBA who had spent his early career at Procter and Gamble. Raphael Afaedor, a Ghanaian Harvard MBA from the same Procter and Gamble pipeline. They were the men who actually ran Jumia Nigeria in its first eighteen months. They hired the riders. They signed the warehouse lease. They negotiated the first inventory deals.
The first office was a house in Yaba, Lagos. The first warehouse was the same house. The first delivery vehicles were the employees’ personal motorbikes. The first three months of orders were fulfilled by twenty-somethings packing cardboard boxes by hand on the living-room floor.
By the end of twenty thirteen, Jumia Nigeria was doing a thousand orders a day. Jumia Egypt had launched. By the end of twenty fourteen, Jumia Kenya, Jumia Morocco, Jumia Ivory Coast, and Jumia South Africa were live. By the end of twenty fifteen, the company was running operations in eleven countries.
And it was not only e-commerce. Under Africa Internet Group, Rocket had stamped out seventy different operating brands across the continent. Jovago for hotel booking. HelloFood for restaurants. Easy Taxi. Lamudi for real estate. Carmudi for cars. Jumia House. Jumia Travel. Jumia Jobs. The strategy was textbook Rocket. Spray brands. See which one stuck. Build a unicorn out of whatever survived.
In Lagos, Tunde Kehinde and Raphael Afaedor watched all of this play out from the operating side. By November twenty fourteen, neither of them was still at Jumia.
Four Wars at Once
Why the two African co-founders left has never been entirely public. What is on the record is the shape of their departure. They left within weeks of each other. Neither of them appeared on the IPO prospectus five years later as a founder. Neither of them held meaningful equity.
The Nigerian press read it as a quiet acknowledgment that, under Rocket Internet’s standard playbook, the local market-entry team was a class of operators, not a class of equity-holding founders. Kehinde would go on to co-found a logistics company called Africa Courier Express, then a Nigerian lending fintech called Lidya. Afaedor would start a Lagos grocery-delivery business called Supermart. Both still work in the Lagos founder community today.
What they left behind, in twenty fifteen, was a company sprawling across eleven countries and fighting four overlapping wars at once.
The first war was logistics. Lagos, Nairobi, Accra, and Abidjan did not have the addressing infrastructure that Amazon’s model takes for granted. There was no door number on most streets. There was no postcode that would route an order. Jumia’s solution became a hybrid of GPS pins, calls to the customer when the rider was ten minutes out, deliveries to “the kiosk on the corner of,” and, eventually, a network of pickup stations that customers walked to themselves.
The second war was payments. Card penetration across the continent in twenty twelve was under five percent. The default was cash on delivery. The customer paid the rider, in cash, when the goods arrived at the doorstep. This created two compounding problems. Return rates were enormous, because customers could simply refuse the package on the doorstep and the rider had no recourse. And the riders themselves became, every working day, walking targets for theft.
The third war was trust. An internal Jumia Nigeria survey in twenty fourteen found that roughly seven out of every ten first-time customers expected the goods either not to arrive, or to arrive counterfeit. The marketing budget through twenty seventeen was, in effect, a trust-purchase fund.
The fourth war was counterfeit itself. Jumia was a marketplace. It listed third-party sellers. It could not, in any meaningful way, vet tens of thousands of small Nigerian and Egyptian merchants for whether the phone they were shipping was actually the phone the customer had ordered. Counterfeit goods became a permanent reputational liability. Every market the company entered relived the same problem.
While the operating team fought all four wars, Berlin was raising money. By the end of twenty fourteen, Africa Internet Group had raised roughly three hundred million dollars cumulatively, from a syndicate that read like a tour of every category of global capital available to an African startup. Rocket Internet. JP Morgan Chase. MTN. The Swedish telecoms group Millicom. The French insurance group AXA. The British development-finance institution CDC.
In February twenty sixteen, Rocket finally pulled the trigger on consolidation. The seventy-plus African Internet Group brands were folded into a single master brand. From that month, almost everything Rocket owned on the continent was called Jumia.
Two weeks later, in March twenty sixteen, AXA led a three hundred and twenty-six million dollar growth round. Goldman Sachs participated. So did MTN. So did Rocket. The valuation was one point oh five billion dollars. With that round, Jumia became Africa’s first technology unicorn.
Inside the building, the unicorn round was treated as the beginning of an IPO countdown. Inside the building, almost no one used the word profit.
Twelve April to Twenty-One August
For three years, the Berlin head office, the Dubai management team, and the country managing directors in Lagos, Nairobi, and Cairo prepared the company for the New York Stock Exchange.
In March twenty nineteen, Jumia filed its IPO prospectus with the US Securities and Exchange Commission. The ticker would be J M I A. The lead bookrunners were Morgan Stanley and Citigroup.
On the twelfth of April, twenty nineteen, Jumia opened on the New York Stock Exchange at fourteen dollars and fifty cents a share. By the end of the first day, it had closed at twenty-five dollars and forty-six cents. By the close of the fourth trading day, it had reached forty dollars and seven cents. At the intraweek peak, it touched forty-nine dollars and seventy-seven cents. The market capitalisation crossed three billion dollars. Jumia had become the first African-operating technology company ever to list on a major US stock exchange. For about ten days, every African newspaper, every African business television channel, and every Lagos and Nairobi WhatsApp group treated this as the moment African technology had finally arrived.
Twenty-eight days later, a short seller named Andrew Left ended that.
Citron Research, run by Left out of Los Angeles, published a twenty-two page report on the ninth of May, twenty nineteen. The title was direct. Jumia is a fraud. The report made four claims.
One, that an investor presentation Jumia had circulated in late twenty eighteen reported six hundred thousand more active customers and ten thousand more active merchants than the IPO prospectus itself reported for the same year.
Two, that the active-customer metric was systematically padded with inactive accounts.
Three, that forty-one percent of Jumia’s orders, by the company’s own prospectus disclosures, were either not delivered, cancelled, or returned.
And four, that the IPO existed primarily to give Rocket Internet and MTN an exit, not to fund the business.
The stock closed down eighteen point eight percent that afternoon. Jumia issued a press release the same day. It stood by its prospectus.
Three months later, in the second-quarter earnings release on the twenty-first of August, twenty nineteen, Jumia disclosed something else. An internal review had uncovered what the company called improper sales practices inside JForce, its independent sales-agent network in Nigeria. Some JForce agents had colluded with internal Jumia employees to game the commission system. They had placed orders that were never real, generated cash on the synthetic transactions, and then cancelled the orders. The fraud had touched roughly two percent of full-year twenty eighteen merchandise volume, and roughly four percent of first-quarter twenty nineteen merchandise volume. The same first-quarter volume the company had used as the headline number on the IPO roadshow only weeks earlier.
The financial impact, the company stressed, was zero, because cancelled orders do not book revenue. The reputational impact was not zero. US class-action lawsuits followed within days. Citron’s headline thesis, that the metrics had been padded, had been partially confirmed by Jumia’s own accounting team.
By April twenty twenty, Rocket Internet had sold its remaining stake. By October twenty twenty, MTN had sold its eighteen point nine percent stake for about one hundred and forty-two million dollars. Within eighteen months of the IPO, the two foundational shareholders had walked out the door.
The Scale They Had to Give Up
For about a year, the pandemic gave Jumia a stay of execution.
Through twenty twenty and into early twenty twenty-one, e-commerce volumes lifted everywhere on the continent. The work-from-home shift, the lockdown closures of physical stores, the substitution of online ordering for the corner shop — all of it was tailwind. Jumia’s share price, which had collapsed to single digits after the JForce disclosure, began to rally. By February twenty twenty-one, the stock traded intraday at sixty-nine dollars a share. The market capitalisation reached seven point four billion dollars. It was the all-time high. It would never be approached again.
In March twenty twenty-one, Jumia tapped the public markets for three hundred and forty-one million dollars through an at-the-market equity sale. In twenty twenty-two, it tapped them again for another two hundred and thirty million. Both rounds were heavy with dilution. Through twenty twenty-two, the share price fell almost eighty percent. The pandemic e-commerce trade unwound. The global technology rout took everything down with it. Jumia Pay, which had been pitched to Wall Street as a standalone fintech franchise inside the e-commerce platform, was losing momentum. Losses continued. The runway shortened.
On the seventh of November, twenty twenty-two, the supervisory board accepted the resignations of Jeremy Hodara and Sacha Poignonnec. The two co-founders, who had run the company for ten years across two continents and through one IPO and one fraud disclosure, were out the same morning. The stock fell another fourteen percent that day.
The board appointed an acting chief executive. His name was Francis Dufay. He had run Jumia Ivory Coast since twenty fourteen. He was a McKinsey alumnus, French born, Kellogg MBA, and he lived in Abidjan. In February twenty twenty-three the board confirmed him permanently. His mandate was unambiguous. Reach profitability, or close the company.
In the next twelve months, Dufay did things the founders had not been willing to do.
He laid off nine hundred staff, roughly twenty percent of the workforce. He closed the Dubai head office, which had run Jumia’s central group functions for almost a decade, and required that more than half of senior management physically relocate to an African country. He himself remained in Abidjan. Antoine Maillet-Mezeray, the new chief financial officer, ran the cost programme from Casablanca.
On the fourteenth of December, twenty twenty-three, Jumia announced that Jumia Food, the restaurant and grocery delivery business it had been operating across seven African markets, would close completely by the end of the year. Dufay told TechCrunch the unit was structurally unprofitable against well-capitalised competitors like Glovo, Bolt Food, and Yango. The same month, Jumia announced that it would exit South Africa and Tunisia. In February twenty twenty-six, the same playbook reached Algeria. The portfolio shrank from eleven countries to eight.
Behind the shrinkage, the unit economics finally began to move.
In twenty twenty-five, Jumia processed thirty-five point five million orders, up thirty-one percent year over year. Gross merchandise value reached eight hundred and eighteen million dollars, up fourteen percent. Revenue was one hundred and eighty-eight million dollars, up thirteen percent. The annual net loss was sixty-one point five million, down thirty-eight percent. The runway, with seventy-eight million dollars of liquidity, was no longer running out.
And most of that growth was coming from the place no one inside the Berlin head office had ever expected. Not Lagos. Not Cairo. Not Nairobi. Rural Nigeria.
The Quiet Reclamation
In the second edition of a report Jumia Nigeria published in late twenty twenty-five, titled E-Commerce in Rural Areas, the company disclosed that sixty percent of its Nigerian orders now came from outside the major metropolitan areas. A year earlier the figure had been fifty.
The mechanism behind that number is the JForce network — the same independent sales-agent network that, in twenty nineteen, had been the source of the fraud disclosure that almost killed the company. JForce now has about thirty-two thousand active agents across the continent. They sit in villages and small towns. They help customers who cannot read order forms or trust an online checkout to place orders. They take cash. They earn a commission. The network that almost destroyed Jumia in twenty nineteen has become, in twenty twenty-six, the central pillar of its survival.
Around the agents sits a physical network of more than three hundred and fifty pickup stations in Nigeria alone, sixty-seven third-party logistics partners, and a thin layer of Jumia’s own riders in the largest cities. Jumia Pay, no longer pitched as a standalone fintech, has been quietly repositioned as the in-app checkout for the marketplace, processing about fifty-six million dollars in payment volume per quarter, roughly twenty-nine percent of total platform value.
In April twenty twenty-six, Jumia reported first-quarter revenue of fifty point six million dollars, up thirty-nine percent year over year. The adjusted operating loss narrowed. Dufay reaffirmed the guidance the company had given at the start of the year. Adjusted operating breakeven by the fourth quarter of twenty twenty-six. Full-year profitability in twenty twenty-seven.
The market, finally, has started to listen. Over the trailing twelve months to June twenty twenty-six, the stock is up nearly one hundred and ninety percent. The market capitalisation has crossed back above nine hundred million dollars. It is still ninety percent below the February twenty twenty-one peak. The cumulative losses since twenty twelve are still roughly two point two billion dollars. The German holding company, the New York listing, and the Berlin registered office have not moved. But the chief executive lives in Africa. The Dubai office is closed. And the company is, for the first time in fourteen years, attempting to be one thing rather than seventy.
That is where the Amazon of Africa stands today. Neither Amazon, nor unambiguously African, but, for the first time, on the edge of being something it was never engineered to be — a profitable, pan-African internet company headquartered, in operational practice if not in legal form, on the continent it sells to. The empire was built on a German factory floor, sold to Wall Street, and reclaimed, slowly, in rural Nigeria. This is Asili Africa. Every empire has an origin.
Key Takeaways
- The Berlin Factory. To understand Jumia, you have to begin in Berlin.
- Twelve April to Twenty-One August. For three years, the Berlin head office, the Dubai management team, and the country managing directors in Lagos, Nairobi, and Cairo prepared the company for the New York Stock Exchange.
- The Scale They Had to Give Up. For about a year, the pandemic gave Jumia a stay of execution.
- The Quiet Reclamation. In the second edition of a report Jumia Nigeria published in late twenty twenty-five, titled E-Commerce in Rural Areas, the company disclosed that sixty percent of its Nigerian orders now came from outside the major metropolitan areas.
Also available on YouTube — search “Asili Africa” or subscribe to our channel.