What Kenya’s M-Pesa reveals about where innovation comes from

The taxi stops at the junction on Ngong Road. The driver quotes the fare. No card terminal appears. No bank app opens. He holds out his phone — the M-Pesa agent number already displayed — and waits while the passenger taps through the transfer. The money moves in approximately four seconds. The driver’s phone receives the confirmation SMS before the passenger has pocketed theirs. The transaction is complete, verified, and recorded in a financial system that reaches deeper into daily Kenyan economic life than the payment infrastructure of most European countries.

This is not a developing-world approximation of a developed-world system. This is the system.

The idea was simple, as all consequential ideas are after the fact—Kenyans living and working in cities needed to send money home to their families in rural areas. The formal banking system served 19% of Kenyan adults in 2006. For the other 81%, the options were to carry cash on a bus, entrust it to a driver making the journey, or make the journey yourself. Each option was expensive, slow, and often dangerous. The banks had looked at this problem and concluded it was not their problem to solve.

Nick Hughes, a Vodafone executive working on the DFID-funded Financial Deepening Challenge Fund, reached a different conclusion. He had noticed that Kenyan microfinance borrowers were using Safaricom airtime — prepaid mobile credit — as an informal currency, transferring it between phones as a form of value exchange. If people were already using airtime as money, Hughes reasoned, you could build a money-transfer service on top of the airtime infrastructure that already reached every corner of the country.

On March 6, 2007, Safaricom launched M-Pesa. The initial business plan projected 350,000 customers in year one. M-Pesa hit 1.2 million. A month after launch, it had 19,671 active users. By November of that year, 1,041,522. The product had found a market that the formal financial system had told itself did not exist.

By 2024, Kenya’s financial inclusion rate had reached 84.8% of adults — up from 26.7% in 2006, the year before M-Pesa launched. The platform processed Sh38.3 trillion in transaction value in the financial year ending March 2025, handling roughly 2,600 transactions per second. What the banks said was not commercially viable turned out to be one of the most commercially successful financial products of the past two decades.

THE GEOGRAPHY OF NECESSITY

The standard account of M-Pesa’s success attributes it to technological ingenuity. This is true but incomplete. M-Pesa was not technologically sophisticated. It ran on basic SMS and USSD technology that worked on any phone, including the simplest handsets available in Kenya in 2007. The technological constraint was, in retrospect, the product’s primary competitive advantage: by building on infrastructure that already reached the entire country rather than on smartphone technology that reached only a small fraction of it, M-Pesa achieved the network density that made it useful from its first year of operation.

The more interesting explanation for M-Pesa’s success is geographic and economic. Kenya’s specific combination of high mobile penetration, low bank penetration, significant rural-urban migration, and a regulatory environment at the Central Bank of Kenya that was willing to permit an experiment the banking system opposed — these were the conditions that made M-Pesa possible. The same technology deployed in a country with universal banking access and a hostile regulator would have produced nothing. The technology was generic. The conditions were specific.

This is the place intelligence argument applied to innovation rather than to craft or culture: the most significant innovation does not necessarily come from the place with the most resources. It comes from the place where the problem is most acute, the constraints are most specific, and the regulatory environment is willing to permit a solution the incumbents would prefer to block. Kenya in 2007 was that place for mobile money, and the innovation it produced has since been adapted across sub-Saharan Africa, South Asia, and Southeast Asia — spreading from the periphery to the center rather than the other way around.

Silicon Valley builds for the next thing. Nairobi was built for the absent thing. The difference in starting position produced a different kind of innovation — one that is less about adding features to a system that already works and more about building a system where no system existed.

WHAT THE M-PESA INFRASTRUCTURE BUILT

M-Pesa’s most significant contribution to Nairobi’s technology ecosystem was not the product itself but the infrastructure layer it created. By establishing a payment rail that reached the entire Kenyan population — through a network that had grown to more than 300,000 agents by November 2024 — M-Pesa laid the foundation on which a generation of subsequent companies could build.

Tala, founded in 2011, uses mobile data patterns to extend micro-credit to borrowers with no formal credit history, disbursing loans via M-Pesa. Twiga Foods, founded in 2014, uses M-Pesa to pay farmers in a supply chain that connects smallholder farmers to urban retailers, eliminating intermediaries who had previously captured most of the margin. Apollo Agriculture uses satellite imagery, machine learning, and M-Pesa disbursement to provide smallholder farmers with credit, insurance, and inputs. Each of these companies is a layer built on the M-Pesa foundation — commercially viable because the payment infrastructure already exists and because that infrastructure has the reach and trust that a new entrant could not build from scratch.

Microsoft invested $1 billion in a data center in Nairobi, establishing its African cloud computing infrastructure. Google, IBM, and the African Development Center of Microsoft have all established regional headquarters in the city. The attraction is not primarily Kenya’s domestic market, which is significant but not exceptional by global standards. It is the combination of the M-Pesa infrastructure, the developer ecosystem it has created, and the specific problem-solving culture that eighteen years of building on constrained infrastructure has produced.

THE SILICON SAVANNAH AND ITS LIMITS

Nairobi’s technology ecosystem has been called the Silicon Savannah since approximately 2010, when the iHub — a co-working space and incubator established in the Ngong Road district — became the meeting point for the city’s developer community. The name is partly aspirational and partly accurate. Nairobi is unambiguously the leading technology hub in East Africa and one of the four most significant on the continent, alongside Lagos, Cape Town, and Cairo.

The limits are structural. Kenya’s domestic market of 55 million people is not large enough to sustain technology companies at scale without regional expansion, and regional expansion across East Africa requires navigating 14 different regulatory environments, 14 different currency systems, and infrastructure quality that varies enormously between Nairobi and the secondary cities of neighboring countries. The companies that have succeeded at scale — M-Pesa, Tala, Africa’s Talking — have done so by expanding across the continent rather than deepening within Kenya.

The 2025 to 2026 consolidation wave confirms this dynamic. Sixty-seven deals in 2025 alone, with Nairobi-based companies acquiring competitors and adjacent businesses across East and West Africa. The M-Pesa infrastructure is now being rebuilt at the architectural level: Safaricom’s ‘Fintech 2.0’ upgrade in September 2025 moved M-Pesa to a cloud-native microservices foundation, expanding peak transaction capacity from 4,400 per second to a new baseline of 6,000, with a theoretical ceiling of 12,000. The platform that started on SMS is now a continental-scale financial infrastructure.

WHAT THIS REVEALS ABOUT PLACE

The standard innovation geography argument says that the best technology comes from the places with the most capital, the best universities, and the most concentrated talent. This argument correctly describes where most technology investment flows. It does not correctly describe where all significant innovation originates.

M-Pesa originated from a specific geographic and economic constraint: a country where mobile penetration outpaced bank penetration, where rural-urban migration created a distinct money-transfer problem, and where a forward-thinking regulator was willing to permit an experiment. The innovation was downstream of the constraint. Remove the constraint, and you remove the innovation — because a country with universal banking access and a hostile regulator would never have needed or permitted M-Pesa.

The traveler who pays for their taxi with M-Pesa is not experiencing a workaround. They are experiencing the standard. The developed world has spent the subsequent 18 years trying to replicate what Nairobi built in 2007 and has not yet fully succeeded. The direction of influence in this domain runs outward from Nairobi.