Mobile money did what bank branches never could: it put a transactional account in the hands of hundreds of millions of Africans, over a feature phone, through a human agent on a street corner. Understand the agent, the float and interoperability, and you understand the rail.
A telco-issued e-money account on a phone, backed by float in a trust account. It banked a continent the branch network never reached.
Mobile money is a stored-value e-money account operated over a mobile phone — originally over USSD on any handset, increasingly via apps — typically issued by a mobile network operator (MNO) or its licensed subsidiary rather than a bank. The customer holds e-money: a digital claim, backed 1:1 by funds the issuer holds in trust at a regulated bank.
Kenya’s M-Pesa (Safaricom) is the archetype and still dominates East Africa — in 2025 it moved transaction value equivalent to more than half of Kenya’s GDP. MTN MoMo and Airtel Money are the pan-continental players, between them spanning dozens of markets and hundreds of millions of accounts. Across Africa there are well over 800 million registered mobile-money accounts.
The genius of the model is the agent network. An agent — a shopkeeper, a kiosk, a fuel station — holds both physical cash and an e-money balance. To cash in, you hand the agent cash and they send you e-money; to cash out, the reverse. The agent’s own working balance is their float, and a dense, well-incentivised, well-rebalanced agent network is the single biggest determinant of whether a mobile-money service actually works for people far from a city.
For every unit of e-money in circulation, the issuer must hold matching real funds in a regulated trust account. This is what protects customers if the issuer fails — and what regulators scrutinise.
Liquidity at the agent — enough cash to pay out, enough e-money to sell — is the lived experience of the service. Agent rebalancing logistics are the hidden core of the business.
Smartphone penetration is rising but not universal. USSD access is why mobile money reaches rural and low-income users that app-only rails miss.
Historically wallets were closed loops — you could not pay an Airtel wallet from an MTN one. Mandated interoperability (Tanzania 2014, Ghana 2018) and instant-rail integration are dismantling the silos.
M-Pesa owns East Africa; MTN MoMo has the widest reach; Airtel Money is the challenger; Orange Money owns Francophone markets. Footprints overlap and shift.
| Provider | Anchor markets | Position |
|---|---|---|
| M-Pesa (Safaricom / Vodacom) | Kenya, Tanzania, DRC, Egypt, Mozambique | Dominant in East Africa; the original and deepest network |
| MTN MoMo | Ghana, Uganda, Nigeria, Rwanda, Côte d’Ivoire + | Largest pan-African footprint; spun out as a fintech |
| Airtel Money | Across 14+ markets in East/Central/West Africa | Fast-growing; ~44.6m users by Mar 2025, rebuilding share vs M-Pesa |
| Orange Money | Francophone West & Central Africa | Dominant in the Sahel and Francophone corridors |
Open above; the build-vs-partner call is for members.
Treat the agent network and interoperability as your two hardest variables. A market with dense agents and mandated interoperability (Ghana, Tanzania, Kenya) is a very different integration from a fragmented, closed-loop one. And design for USSD where reach matters — an app-only integration quietly excludes the customers mobile money exists to serve.
You will almost never out-build an incumbent’s agent network. The pragmatic play is to interoperate — plug into MoMo, M-Pesa and Airtel rails and let them carry the last mile — rather than stand up a wallet from zero. Float-in-trust and e-money licensing are the regulatory cost of issuing yourself; weigh that before deciding to.
Underestimating agent liquidity logistics is the classic failure: a beautiful app on top of an agent network that routinely has no cash to pay out is a dead product. The float and the agents are the business.