4 ideas
Pilot users instead began sending money to relatives, paying for goods and parking cash overnight, and the team rebuilt the product around one message: 'send money home.' M-Pesa launched nationally in March 2007 and within about four years was used by most Kenyan adults.
M-Pesa worked in Kenya because several conditions held at once. Heavy urban-to-rural migration created steady demand for remittances. The existing options were costly and risky, such as bus drivers carrying cash or slow Post Office transfers. Safaricom held a dominant share of mobile subscribers, so network effects came almost immediately. Copies of the scheme in markets without a dominant operator, remittance demand and a tolerant regulator stalled, even when their technology was similar.
Mobile money only works if people can turn physical cash into electronic value and back again near where they live. M-Pesa solved this by recruiting thousands of existing airtime dealers and small shopkeepers as agents who handle cash-in and cash-out for a commission. These agents replaced bank branches at a fraction of the cost. The density of agents built trust and made the service useful, so the retail network mattered more to success than the phone software.
The Central Bank of Kenya chose not to classify M-Pesa as banking and let it operate under a letter of no objection. Customer funds were held in a trust account at commercial banks, and the Bank regulated after watching how the service actually behaved. That tolerance of a non-bank payment system is presented as a necessary condition that most other countries' regulators did not provide.