Nobody outside Africa understands why mobile money won. Let me explain. (a thread, 1/22)

1/ Every few months someone in a San Francisco podcast "discovers" that Kenyans pay for tomatoes with their phones and calls it leapfrogging. It was not magic and it was not an accident. It was a series of boring, specific decisions. Here they are.

2/ M-Pesa launched in Kenya in March 2007, run by Safaricom (Vodafone was a big shareholder). The original pilot was not even about payments. It was a way to repay microfinance loans.

3/ Users ignored the plan and used it to send money to family. That is the first lesson: watch what people actually do with your product, not what your deck says.

4/ Why sending money? Urbanisation. Someone works in Nairobi or Mombasa, their family is in Kisumu or Kakamega. Before M-Pesa you sent cash with a bus driver, or a friend, and hoped. I grew up receiving money that way. Sometimes it arrived.

5/ The killer feature was not the app. There was no app. It was the agent: a shop on your street that converts cash to e-money and back. Cash-in, cash-out. Without agents, mobile money is a wallet you cannot fill.

6/ Western fintech forgets this every time. Your users live in a cash economy. If you cannot touch cash, you are not in the market.

7/ It ran on any phone. SIM toolkit menus and USSD (dial *334# on Safaricom today and you still get the menu). No data plan, no smartphone, no app store account. A 2007 Nokia was enough.

8/ The receipt is an SMS. People screenshot it, forward it, keep it. Your "payment confirmed" page is less trusted than a text message from the network. Design for that.

9/ The regulator let it happen. The Central Bank of Kenya gave Safaricom a letter of no objection instead of forcing it into a banking licence first. Regulate later, watch closely. Not every regulator was that brave.

10/ Cards never had a chance here. Few people had bank accounts, POS terminals were expensive, merchants hated the fees and the settlement delays. A fruit seller does not wait three days for her money.

11/ Then came merchants: Paybill numbers for bills and businesses, Till numbers (Buy Goods) for shops. Your landlord, your school fees, your water bill, your mama mboga. One number, one PIN.

12/ Then credit and savings on top. M-Shwari (with a bank partner) in 2012, Fuliza overdraft in 2019. Mobile money stopped being a pipe and became a bank in everything but name. That part has real problems too, see 18/.

13/ The research. Suri and Jack (2016, Science) estimated access to M-Pesa lifted about 194,000 Kenyan households, roughly 2%, out of extreme poverty, with the biggest effect in households headed by women. Argue with the method if you like, but read it first.

14/ Now the honest part: "mobile money won" is mostly an East African story. It did NOT win everywhere.

15/ Nigeria went bank-led. Regulators kept telcos out for years, and instant bank transfers (NIP) became so good that "send me your account number" is the default. Ask any Lagos developer what they use. It is transfers and POS agents, not a telco wallet.

16/ Francophone West Africa is a third model: Orange Money and MTN MoMo are big, interoperability is patchier. South Africa has strong banks and cards, and M-Pesa actually shut down there.

17/ In Kenya, interoperability between M-Pesa and Airtel Money for person to person transfers came in 2018. Before that, every wallet was its own island. Monopoly power is the price Kenya paid for speed.

18/ The downsides are real: one company sees a huge share of the country's transactions, digital lending on top of wallets has trapped people in small high-cost loans, and when the network has a bad day, the economy has a bad day.

19/ If you build for these markets as a developer: phone number is identity, the SMS is the receipt, the agent is your onboarding, cash-out is a feature, reversals are a support process, and your API will be called over a 2G connection in a matatu.

20/ If you are a founder: do not copy the Kenyan playbook into Lagos and expect the same result. The rails are different. The trust is different. The regulator is different.

21/ And if you are a podcast host: it was not leapfrogging. It was agents, USSD, a patient regulator, and millions of people sending money home.

22/ Sources: Suri and Jack, "The long-run poverty and gender impacts of mobile money", Science 354 (2016). GSMA's yearly State of the Industry Report on Mobile Money. Central Bank of Kenya payment statistics. Safaricom's annual reports. Asante for reading. Questions below, I answer all of them, sawa?