The complete history of M-Pesa (2007 — 2026)

M-Pesa is the most successful mobile-money product in history. It moves more value annually than the Kenyan banking sector, reaches almost every adult in the country, and has spawned a generation of African fintech that started by asking the same question: what if the phone is the bank?

Two decades after the idea was sketched, it's easy to forget that the original product was supposed to be something completely different — and almost didn't happen at all.

The origin: a microfinance experiment that pivoted

In 2003, Nick Hughes — then a Vodafone executive working on corporate social responsibility — won a £1 million matching grant from the UK Department for International Development (DFID) to design a mobile-based system that would help microfinance customers in Kenya repay loans more cheaply. The team that built the prototype was small, and the bet was straightforward: if money could move over SMS, the cost of disbursing and collecting microloans would collapse.

Safaricom, the Kenyan mobile operator (then 60% owned by Vodafone), ran a small pilot in 2005 with Faulu Kenya — the same Faulu microfinance bank that today is listed on paybillke. The pilot worked. But the more interesting finding was buried in the data: pilot users were sending money to each other, not just repaying loans. They were sending it to relatives. To pay rent. To buy goods.

The team made a decision that would define the next two decades of African fintech: drop the microfinance use case, launch as a person-to-person money-transfer product, and let customers tell them what to build next.

March 2007: the launch

M-Pesa launched commercially on 6 March 2007 with a single product: Send Money. A registered customer could deposit cash with a Safaricom agent, send money via SMS to any other Kenyan with a Safaricom line, and the recipient could withdraw cash at another agent. That was it.

Three things made it work where everyone else had failed. First, Safaricom already had over 70% market share in Kenya, so the network was effectively pre-built. Second, the agent network was incentivised on transaction volume rather than customer signups, which created a flywheel: more agents made the product more useful, more usefulness drove more transactions, more transactions paid more agents. Third, regulators allowed Safaricom to operate the scheme as an extension of its telecoms licence rather than a banking licence, giving it flexibility traditional banks didn't have.

2007–2012: the growth phase

The numbers in the first five years were absurd. By the end of 2007, M-Pesa had 1.2 million registered customers. By the end of 2008, it was 5 million. By 2010 it was over 13 million — more than the formal banking population of Kenya at the time. By 2012, M-Pesa was processing more transactions per day than Western Union processed globally per year.

This was the period where the phrase “Lipa na M-Pesa” — “Pay with M-Pesa” — entered Kenyan vocabulary. Bills were paid by M-Pesa. Goods were sold for M-Pesa. Salaries were sent by M-Pesa. KPLC — the Kenya Power and Lighting Company — was an early adopter and remains one of the highest-volume paybills in the country.

2012: M-Pesa becomes banking infrastructure

In November 2012, Safaricom launched M-Shwari in partnership with what was then the Commercial Bank of Africa (now NCBA). M-Shwari let M-Pesa customers save and borrow directly through the M-Pesa menu — open a savings account, earn interest, take a small loan based on M-Pesa history, all without ever visiting a bank.

This was the moment M-Pesa stopped being a payments product and became the default consumer bank for most Kenyans. By 2015, M-Shwari had over 12 million customers — making CBA instantly one of the largest retail banks in Africa, despite never opening a new branch.

KCB followed in 2015 with KCB-M-Pesa, structured similarly. By 2020, every major Kenyan bank had its own M-Pesa-integrated savings/lending product, and most retail banking in Kenya — measured by transaction count — happened over the M-Pesa USSD menu.

The Paybill / Lipa na M-Pesa expansion

The Paybill product (B2C billing) and Buy Goods / Till product (retail) were rolled out from 2009 onwards. The structural difference matters: Paybill is paid by the customer, Till is paid by the merchant. That single design choice is why a coffee at a café charges your card but is free on M-Pesa Till — your bank pays the card networks, the café pays Safaricom.

By 2019, paybill numbers had become the de-facto rails for nearly every regular payment Kenyans make. Utilities, schools, government services, betting, lending — if you owe a Kenyan business money, there's a paybill. The paybillke directory tracks 200+ of them, with the top 50 re-verified weekly.

2019: Fuliza and the overdraft economy

In January 2019, Safaricom and NCBA launched Fuliza — a continuous M-Pesa overdraft. If you tried to send money you didn't have, Fuliza would cover the gap, then repay itself the moment money entered your wallet. Within a month, Fuliza was the most-used credit product in Kenyan history. Within a year, it had disbursed over KES 81 billion.

Fuliza was a structural shift: M-Pesa was no longer just moving money, it was creating money — short-term credit at scale, accessed in seconds, repaid automatically. Critics worried about the debt-trap implications. Defenders pointed out it was cheaper than the alternative (digital lenders charging 30%+ APR). Both were partly right.

The regulatory journey

M-Pesa's regulatory history is its own essay, but the key beats:

• 2007–2010: regulated as a telecoms-adjacent service, light supervision.
• 2011: the National Payment System Act formalised CBK oversight of mobile-money providers.
• 2018: 12% excise duty introduced on M-Pesa fees (later raised to 15%).
• 2022: CBK and Safaricom unbundled M-Pesa into a separate legal entity, partly to enable cross-border services and partly to reassure regulators about systemic concentration.
• 2023–2024: ongoing debate over interoperability with other mobile-money networks and bank rails. Pesalink and the eventual full integration with Airtel Money finally unlocked send-across-network.

The competitors who matter (and the ones who didn't)

For most of M-Pesa's history, competition was theoretical. Airtel Money launched in 2009 and has been the consistent number-two ever since, but with a fraction of the market share. T-Kash (Telkom) is essentially a rounding error. Equitel (Equity Bank's MVNO play) was the most credible challenger but has stalled.

The interesting development is that Airtel Money is now genuinely useful for paybill payments — most major Kenyan paybills accept both. We cover the comparison in detail in our M-Pesa vs Airtel Money guide.

Where M-Pesa is now (2026)

As of early 2026, M-Pesa moves over KES 39 trillion in annual transactions across more than 70 million active customers (Kenya plus other markets including Tanzania, Mozambique, DRC, Lesotho, Egypt, and Ethiopia in pilot). It's still operationally part of Safaricom but runs as a distinct product organisation, M-Pesa Africa, headquartered in Nairobi.

The next-generation challenges are interesting: cross-border interoperability with non-Vodafone markets, competition from Hustler Fund and other government credit programmes, and the slow encroachment of stablecoin rails for diaspora remittance. Whether M-Pesa adapts to crypto remittance or fights it is one of the more consequential strategic questions in African fintech.

Why this matters for paybillke

Every paybill in our directory exists because of decisions made in M-Pesa's first decade. The Paybill numbering convention, the account-format expectations, the way fees are bracketed by transaction size, the BCLB-licensed-only constraint on betting paybills, the zero-rated government paybills via eCitizen — none of those happened by accident. They happened because Safaricom, CBK, KRA, and a handful of regulators made specific choices over nearly two decades.

Understanding the history changes how you read the directory. A paybill isn't a number — it's a settlement contract, an excise-duty bracket, a regulatory licence, a merchant-acquirer agreement, all rolled into seven digits.