3 Million Of Africa’s Cheapest Phones Did Not Sell, The Average Price Rose 41 Dollars And The Reason Is An AI Data Center.
The $100 smartphone is disappearing from Africa, and nothing about the reason is African.
In Lagos, Nigeria, a mechanic explained this month why he had gone back to a handset that does almost nothing. Every time he switched on mobile data, background updates swallowed about N1,000 before he opened a single app.
So he bought a palasa, the feature phone Nigerian retailers cannot keep in stock.
He is a market.
African smartphone shipments fell 7% year on year in the second quarter of 2026, to 17.8 million units. That is the first annual decline since 2023, and Omdia expects the full year to close 26% down.
Look at where the fall sits. The sub $100 tier lost 34% of its shipments, nearly 3 million devices, while the average phone sold in Africa rose $41 in 12 months to $202. Nigeria fell 11%. Kenya fell 15%. Egypt fell 26% after input costs rose 50% since January.
The cause is not on this continent.
Memory is the reason. AI data centers are bidding for the same chips that go into cheap phones, and they pay more. Micron closed its consumer Crucial business in December 2025 to serve data center demand instead.
“Vendors can no longer profitably manufacture $75 smartphones.”
MANISH Pravinkumar, Omdia analyst
Read that as a sentence about Africa. The device that brought the last few hundred million people online was a $75 device, and it is being discontinued by an industry that has found a richer customer.
Africa is paying for the AI build out at the till, and the poorest buyer pays first.
What $100 buys, and what $100 costs
In Nigeria a $100 phone now costs N135,041. The minimum wage is N70,000 a month and has not moved since the 2024 Act. The cheapest smartphone on the shelf is almost 2 months of legal minimum income.
Across Sub Saharan Africa, GSMA puts an entry level device at 87% of monthly income for the poorest 20%.
That is the real barrier, and it is not coverage. 63% of Africans live under a mobile broadband signal they do not use, and most of them do not own a phone to use it with.
The tower is built. The handset is not affordable. Only 1 of those problems has been getting money.
The market did not shrink evenly.
TRANSSION, which makes TECNO, Infinix and itel and still holds 47% of African shipments, fell 14%. Samsung, which sells further up the price ladder, rose 15%. In Nigeria’s installed base Samsung sits at 37.8% and Apple at 21.4%, together 59.2% of phones actually in use, much of that arriving second hand.
The top of the market is fine. The bottom is being priced off the internet.
Who wins, who is squeezed, and the gap
The winners are anyone who finances or refurbishes a device rather than selling one. Device credit, pay as you go bundles, agent networks, certified second hand, repair. Each of those turns a price the customer cannot pay into a payment the customer can.
The squeezed are the budget manufacturers whose entire model was volume at $75, the operators whose data revenue depends on smartphones they do not sell, and the fintech, agritech and commerce founder whose product assumes the customer already owns a screen.
That last one deserves a beat. Every African startup building for the next 100 million users has been assuming those users arrive on their own. This quarter, 3 million of them turned around and went home.
The gap is the balance sheet that stands between the shelf price and the wage.
Somebody already turned the price into a payment.
M-KOPA has deployed N231 billion of credit in Nigeria since 2019, to more than 1 million customers, the fastest of any of its markets. 290,000 of them owned a smartphone for the first time. 52% were borrowing formally for the first time. 33% are women. 77% use the device to earn, and 75% say they earn more since they got it.
It runs on 11,000 agents, at 0.1% turnover.
Sit with the last figure. The hard part here was never the credit model. It was finding the person who signs a customer up in a market with no addresses, and this is a company that solved that and then kept the people who do it.
So the proof exists, it is profitable, and it is 1 company.
That is the arithmetic to take into the rest of your week. 1 million financed customers, against a continent where 63% of the people already under a signal are still not online. The model works. The number of balance sheets running it does NOT.
$100 was never really a price. It was a threshold, and the market that used to clear it has walked away from the counter.
That is the conversation waiting at the Business Week Afrika Summit on 1 and 2 October 2026. The asset financiers, the mobile operators, the device distributors and the development lenders each hold 1 piece of this and are not in the same room, which is why a solved problem runs at the scale of 1 company.
Secure your seat and join the builders closing the gap: https://apps.little.africa/events/105
In a repair shop in Lagos, a mechanic charges a phone that cannot open a single one of the apps built for him.