Jumia's Amazon problem: why 'the Amazon of Africa' couldn't find the customers

Jumia is the most-watched consumer-internet story in Africa, and its export lesson runs the opposite way to most teardowns here: this wasn't a great local product that failed to travel. It was a foreign playbook — Amazon's — imported into markets that couldn't yet run it. In April 2019 Jumia became the first African technology company to list on the New York Stock Exchange, and investors treated it like the "Amazon of Africa." Within its first year the stock fell roughly 77%, and the company has posted losses for most of its public life. It has since pulled out of markets like Cameroon and pared its footprint back toward a single-digit number of countries. The tech worked. The substrate underneath it didn't exist yet.

What they did — and the going-global move

Jumia's pitch was clean: e-commerce is inevitable across a continent of 1.4 billion people, so build the marketplace, the logistics, and the payments rail before anyone else and ride the wave. It launched across many African markets at once and raised aggressively on the "Amazon of Africa" framing that global investors instantly understood. That framing is exactly the problem. Amazon is a website on top of thirty years of American substrate — reliable roads and addresses, near-universal card penetration, cheap parcel logistics, high trust in remote payment. Jumia had to build most of that itself, market by market, while burning capital to acquire customers who often had no card and no street address.

The teardown — through the deeper-local lens

Strip Jumia to primitives and the same four local-substrate gaps show up that sink most "the X of Y" plays:

1. Logistics substrate. Building e-commerce at scale where roads, addressing, and parcel networks are underdeveloped is slow and expensive. Jumia had to run its own delivery and last-mile across fragmented geographies — a cost Amazon never carried at launch.
2. Payments substrate. Card penetration and digital-payment rails were thin, so a large share of orders ran on cash-on-delivery — which raises returns, fraud, and working-capital pain, and quietly caps the flywheel.
3. Trust substrate. In low-trust remote-commerce markets, cash-on-delivery isn't a feature; it's a symptom. Customers won't prepay a stranger, so the model absorbs the risk the buyer won't.
4. The pain-and-tariff shape. High import tariffs and limited local sourcing raised landed costs, and fragmented, lower-income demand kept the unit economics stubborn.

None of these are app problems. They're the roads, rails, and trust an e-commerce business sits on — and in Jumia's markets that layer was still being poured.

My operator take

E-commerce is not a website; it is a logistics-payments-trust machine with a storefront bolted on. In markets where that machine already exists, a marketplace is a software problem. In markets where it doesn't, the marketplace is the machine — and you are financing roads, delivery fleets, and payment habits out of retail margins. Jumia's mistake wasn't ambition or execution; it was accepting a label that priced the storefront and ignored the substrate. The teams that win here stop pitching the destination and start underwriting the build.

The playbook takeaway

Before you brand yourself "the Amazon of X," price the substrate you'll have to build, not just the TAM you'll address. Amazon inherited its logistics-payments-trust stack; if yours doesn't exist, you're not running Amazon — you're building the roads and the store at once, on the store's margins. Tag each advantage as portable product or local substrate, then ask the brutal question: can this market's demand fund the substrate build before the capital runs out? It's the same lens the M-Pesa teardown and Localizing trust apply from the other direction.

One honest open question

Jumia may still win the long game — a decade of substrate-building is a moat once it's poured, and a patient owner could compound it. The open question is whether public markets will fund a road-builder on store margins for the years that takes, or whether "the Amazon of Africa" only ever worked as an IPO narrative, not an operating model.