Note: This is a research note supplementing the book Unscarcity, now available for purchase. These notes expand on concepts from the main text. Start here or get the book.

M-Pesa: When a Text Message Beats a Bank

The SMS-powered revolution that turned more than 70 million Africans into banking customers without building a single bank

Nairobi in 2006. You’re a young man who has migrated to the city to work construction, sending money home to your mother in a village eight hours away. Your options are bleak: stuff cash into an envelope and hope the bus driver is honest, pay Western Union’s 15% fee (assuming there’s an office anywhere near your mother), or make the trip yourself every few weeks, losing two days of wages in the process.

The country has 37 million people and 575 bank branches. That’s one branch for every 64,000 people. The nearest ATM to your mother? There isn’t one. Kenya has about 600 ATMs in the entire country, and none of them are within walking distance of a subsistence farmer.

Now move to 2026. The same country processes more than $322 billion in mobile money transactions annually, more than twice Kenya’s entire GDP. More than 80% of adults have access to financial services. Your mother buys vegetables from a market vendor who accepts payment via text message. She receives your remittance instantly, on a device that cost less than a single Western Union fee.

No new bank branches were built. No ATM network was constructed. Instead, Kenya simply skipped the infrastructure that wealthy nations spent a century building.

The tool that made this possible was called M-Pesa. And its story is the most important case study in infrastructure leapfrogging that exists.

The Accidental Revolution

M-Pesa was never supposed to revolutionize anything.

In 2003, a British technologist named Nick Hughes was working at Vodafone, trying to figure out how to use mobile phones to help microfinance institutions in developing countries. The idea was simple: let people repay their microloan installments via text message instead of trekking to a loan office. It was a nice efficiency improvement for an existing system.

The UK’s Department for International Development provided funding for a pilot. Safaricom, Kenya’s dominant mobile carrier (40% owned by Vodafone), would test it in the field. By 2005, they were running trials in Nairobi and the neighboring town of Thika.

Then something unexpected happened.

The trial participants weren’t just repaying loans. They were using the system to send money to each other. A construction worker in Nairobi would “repay” a fictional loan, and his mother in the village would “receive” the repayment. The microfinance use case had accidentally become a remittance network.

Susie Lonie, who worked on the M-Pesa development, later recalled that the team recognized something profound: Kenyans didn’t need help with microfinance. They needed a way to move money without physically carrying cash across dangerous roads.

On March 6, 2007, Safaricom officially launched M-Pesa. The name combined “mobile” with “pesa,” the Swahili word for money. Within a month, nearly 20,000 people had signed up. Within a year, over a million. By 2009, 8.3 million Kenyans were using M-Pesa, representing more than 20% of the country’s entire population.

They had stumbled onto the financial infrastructure for a nation that banks had deemed too poor to serve.

How It Actually Works

The remarkable thing about M-Pesa is that it runs on technology from 1992.

The underlying system uses SMS and USSD, the same protocols that power basic text messaging and those annoying carrier menus you access by dialing star codes. No smartphone required. No internet connection. No app store. If your phone can send a text message, it can send money.

The process is almost insultingly simple:

You register at an M-Pesa agent (often a corner shop or kiosk) with your national ID
You receive a PIN and a linked account
To deposit money, you give cash to an agent and receive “e-float” on your phone
To send money, you select the recipient’s phone number and enter your PIN
The recipient receives an SMS confirmation instantly
To withdraw, the recipient visits any agent and exchanges e-float for cash

That covers it. No credit checks, no minimum balances, no monthly fees for basic transactions. The system treats money like airtime, converting cash into digital value and back again at tens of thousands of tiny storefronts across the country.

By March 2026 Safaricom counted 333,011 M-Pesa agents and 40.99 million customers active in the past month in Kenya. To put that in perspective: Kenya has 1,611 bank branches and 2,228 ATMs, by the central bank’s latest count. The M-Pesa agent network is about 87 times larger than the entire formal banking footprint.

The technology was twenty years old; the genius lay elsewhere. It was the recognition that a corner shop owner with a float of cash could do exactly what a bank branch does, for people who would never in their lives walk into a bank.

The Leapfrog Effect

Development economists have a term for what happened in Kenya: technological leapfrogging. It’s when a country skips an entire generation of infrastructure by adopting a newer technology directly.

Africa skipped landlines and went straight to mobile phones. Kenya had only 300,000 landlines for 30 million people in 2000. By 2007, it had 11 million mobile subscribers. The economics were simple: building telephone poles across rural Africa costs a fortune. Building cell towers costs a fraction of that.

M-Pesa was the same logic applied to banking. Building bank branches requires real estate, security, staff, vaults, armored cars, regulatory infrastructure, and the kind of capital that no one was willing to invest for customers averaging $2 per day in income. But a corner shopkeeper already existed. A mobile phone network already existed. M-Pesa was duct tape that connected them.

A World Bank analysis of mobile banking put the prospect this way:

“If the services offered through mobile banking services such as M-Pesa are deepened, it is possible that low-income countries could leapfrog branch based banking into mobile banking in similar fashion to their leapfrog over landline telecommunication infrastructure straight to mobile phone technology.”

Kenya proved that a country doesn’t need to build the infrastructure that wealthy nations built. It needs to build something that works.

This distinction matters profoundly. The Western financial system assumed you needed:

Physical branches where people could deposit and withdraw
Expensive verification systems to prevent fraud
Regulatory frameworks designed for large institutions
A minimum viable customer worth serving

M-Pesa proved you could substitute:

A network of existing shopkeepers
A simple PIN system and SMS confirmations
Light-touch regulation (Kenya’s central bank famously said “let them try and we’ll see what happens”)
Customers at any income level, because marginal costs approach zero

The result was a system that achieved 80% financial inclusion in a country where banks had reached barely 20%. Beyond being faster, the leapfrog was more complete.

The Numbers That Matter

Consider the scale of what M-Pesa accomplished.

Before M-Pesa (2006):

18.9% of Kenyans had bank accounts
38.4% were completely excluded from any financial services
Kenya had about 1.6 bank branches per 100,000 people
Sending money home required physical travel or trusting strangers

After M-Pesa (2026):

84% of Kenyans have access to financial services
M-Pesa processed 46.4 billion transactions worth 41.7 trillion Kenyan shillings in the financial year ending March 2026, about 127 million transactions per day on average. (The dollar figure is ~$322 billion at the central bank’s September rate of 129.6 shillings.)
40.99 million one-month-active users in Kenya as of March 2026. M-Pesa’s share of mobile-money subscriptions was 88.8% in June 2026 against Airtel Money’s 11.1%, by the regulator’s count: down from 90.9% a year earlier and from a peak near 99% in 2020. The same report counts 54 million mobile-money subscriptions, more than Kenya has people, and 568,463 registered agents across all providers, down 5.6% in a quarter
92 million customers across the eight African markets where Vodafone runs mobile money (March 2026, up from 66 million two years earlier), roughly 73 million of them on M-Pesa itself, with the service turning 19 in March 2026

MIT and Georgetown economists calculated that M-Pesa lifted 194,000 households, 2% of all Kenyan households, out of poverty. The research also found that 185,000 women moved out of farming and into business or retail work because mobile money made small-scale commerce viable.

The mechanism is straightforward: when you can receive payments instantly and digitally, you can run a business without cash flow gaps. The “mama mboga,” the vegetable vendor at the local market, can track daily sales on her phone, receive payment from customers who don’t carry cash, and immediately pay her own suppliers. Money circulates faster. Businesses that weren’t viable become viable.

There’s a criticism worth acknowledging: transaction fees for small amounts can hit 10% or more, which disproportionately burdens the poorest users. On Safaricom’s own tariff, taking 100 shillings out at an agent costs 11. (In August 2026 the company cut merchant fees, making Buy Goods collections free up to 500 shillings; the withdrawal fee stayed.) And products like Fuliza, M-Pesa’s overdraft facility, have trapped some users in debt cycles; in October 2026 Safaricom extended it to the tills of the smallest merchants, with a limit set by their transaction history and repayment taken automatically from their next sales. Financial inclusion isn’t automatically financial welfare. But the net impact has been overwhelmingly positive, and the fees are still far lower than the alternatives that existed before.

Why M-Pesa Worked in Kenya (And Failed Elsewhere)

M-Pesa’s success wasn’t inevitable. The same technology flopped spectacularly in other markets.

Vodafone launched M-Pesa in India in 2013. By 2019 it had shut down, unable to compete with existing banking infrastructure and facing regulatory resistance. South Africa’s launch in 2010 was discontinued in 2016, unable to dislodge established banks (MTN closed its own South African mobile-money service the same year). Commercial launches in Romania and Albania were wound up in 2017.

Researchers have identified several factors that made Kenya unique:

1. The infrastructure gap was a feature, not a bug.
In countries with extensive bank branch networks, M-Pesa competed against entrenched incumbents. In Kenya, there was nothing to compete against. The unbanked population wasn’t being served by anyone, so there was no one to resist disruption.

2. Regulatory permissiveness.
Kenya’s central bank took a hands-off approach, allowing Safaricom to experiment before imposing strict rules. Other regulators demanded full banking licenses, extensive capital requirements, and compliance frameworks that made the economics unworkable.

3. Safaricom’s market dominance.
Safaricom had roughly 80% of Kenya’s mobile market when M-Pesa launched. That meant network effects kicked in immediately: everyone you wanted to send money to was already on Safaricom’s network. In more fragmented markets, interoperability problems plagued adoption.

4. Trust in agents.
Kenya had a pre-existing culture of informal financial services: rotating savings groups, trusted shopkeepers who would hold money, community-based lending. M-Pesa formalized networks that already existed informally.

The lesson for future leapfroggers: the technology is the easy part. The hard part is the ecosystem: regulatory environment, market structure, existing alternatives (or lack thereof), and cultural trust networks.

The Global Spread

Despite the failures in some markets, M-Pesa has expanded dramatically across Africa and beyond:

Tanzania (2008): Now M-Pesa’s second-largest market, with mobile money deeply integrated into daily life
Democratic Republic of Congo (2012): Operating in a country where formal banking barely exists
Mozambique (2013): Helping connect one of the world’s poorest populations to financial services
Egypt (2013): Competing in the Arab world’s largest market
Ghana (2015): Launched into a competitive mobile money market; Vodafone sold the operator to Telecel in 2023
Ethiopia (2023): The newest and potentially largest market, with 10.8 million registered customers within 16 months of launch and 29 million by June 2026

Ethiopia’s launch is particularly interesting. Unlike Kenya in 2007, Ethiopia has 32 banks and more than 12,000 branches. Yet cash still accounts for 99% of payment operations. The country is attempting a different kind of leapfrog: not from no infrastructure to digital, but from inefficient infrastructure to digital. Only about half of Ethiopians have any financial account, despite all those bank branches.

The growth curve backs that up: Safaricom’s FY2026 results, published in May 2026, show M-Pesa Ethiopia’s active subscriber base reaching 5.2 million, up 119% year over year, processing 442 million transactions worth roughly $221 million for the year ending March 2026. Three months later the quarterly update had it at 5.69 million, up 131%, with nearly 90,000 active merchants.

The pattern suggests that the leapfrog opportunity isn’t limited to countries with zero infrastructure. It’s available anywhere that existing infrastructure is failing to serve people.

2025-2026: The Super App Era

M-Pesa isn’t standing still. Recent developments point toward a platform strategy far beyond simple money transfer:

Blockchain Integration (December 2025 → 2026)

M-Pesa Africa announced a partnership with the ADI Foundation, a UAE-backed blockchain project, to integrate distributed ledger technology into its platform. The deal was formalized in January 2026, naming eight markets for the rollout: Kenya, the DRC, Egypt, Ethiopia, Ghana, Lesotho, Mozambique, and Tanzania. The plan brings M-Pesa’s 60-plus million monthly users “onchain” through a dirham-backed stablecoin, aimed first at cross-border settlement for Abu Dhabi-based firms, with a start “estimated for early 2026.” That window has passed. The dirham stablecoin went live on ADI’s own chain in February and reached shop tills in the UAE in a September pilot, but as of late October 2026 no launch had been announced in any M-Pesa market. By March 2026, ADI Foundation named M-Pesa alongside Mastercard, BlackRock, and Franklin Templeton among its flagship institutional partners as it added Chainlink’s cross-chain infrastructure to the network.

Sitoyo Lopokoiyit, who ran M-Pesa Africa until he left for Absa in March 2026, described it as tapping into “new technologies and how these can transform financial services.”

Super App Transformation

Safaricom is rebuilding M-Pesa’s technology architecture to support what it calls “Fintech 2.0”:

Offline mode: Transactions without data bundles or internet connectivity
NFC payments: Tap-to-pay functionality using mobile phones
Wallet sharing: Multiple users can access the same account for specific purposes
Transaction capacity: The core platform migration completed in September 2025, lifting capacity from 4,500 to 6,000 transactions per second at launch, with the cloud-native architecture designed to scale as high as 12,000 tps as demand grows
Business loans: Direct lending to small traders from 1,000 to 250,000 Kenyan shillings

The super app model, pioneered by WeChat in China and Grab in Southeast Asia, bundles financial services, commerce, transport, and communication into a single platform. M-Pesa is positioning itself as Africa’s equivalent: the operating system for daily life.

What M-Pesa Teaches About Leapfrogging

The M-Pesa story carries a specific claim about infrastructure well beyond mobile money:

You don’t have to build what came before. You only have to build what works.

Kenya didn’t need 10,000 bank branches. Kenya needed a way for people to store and transfer value safely. The assumption that those two requirements demanded the same solution was wrong.

This matters enormously for the Free Zone model described in Unscarcity. The question isn’t “how do we build the entire welfare state infrastructure of a wealthy nation?” The question is “what do people actually need, and what’s the shortest path to providing it?”

The parallels are striking:

Challenge	Traditional Approach	M-Pesa Approach
Banking access	Build expensive branches	Use existing shopkeepers
Identity verification	Government-issued cards, credit checks	Phone number + PIN
Security	Vaults, armored cars, guards	SMS confirmations, distributed float
Scale economics	Minimum viable customer size	Marginal cost approaches zero

Now imagine applying the same logic to other infrastructure:

Challenge	Traditional Approach	Leapfrog Approach
Power distribution	National grid, substations	Solar microgrids + battery storage
Healthcare	Hospitals, clinics, specialists	AI diagnostics + mobile health workers
Education	Schools, teachers, curricula	AI tutors + practical apprenticeship
Food distribution	Supermarkets, supply chains	Vertical farms + automated delivery

The insight isn’t that technology magically solves problems. The insight is that the absence of legacy infrastructure can be an advantage. When you don’t have to maintain, upgrade, or replace existing systems, you can build directly for the endpoint.

Wanjiku, the character in Unscarcity who grows up with M-Pesa as a given, understands this instinctively. By her 80s, she’s watching her grandchildren experience a second leapfrog: abundance infrastructure built from scratch because scarcity infrastructure was never worth preserving.

The Deeper Lesson

The part that should make policy makers uncomfortable is this: M-Pesa succeeded largely because the Kenyan government got out of the way.

The Central Bank of Kenya’s initial response to M-Pesa was “let them try and we’ll see what happens.” They could have demanded full banking licenses. They could have insisted on extensive capital requirements. They could have protected the established (and politically connected) banking sector from disruptive competition.

They didn’t. And the result was a financial revolution.

Nineteen years on, the state is writing the rules it skipped, and arguing over who owns the rails. On September 21, 2026 the Treasury and the central bank published a draft National Payment System Bill that would set minimum capital for e-money issuers, require customer funds to sit in trust accounts, and let the central bank order providers to interoperate. Six days earlier a three-judge High Court bench had declared unconstitutional the government’s sale of 15% of Safaricom to Vodacom, a KSh 204 billion deal completed in June that had taken Vodacom to 55%, and ordered the shares returned “to the ownership of the Government of Kenya on behalf of the people of Kenya.” The government and Vodacom appealed, and on October 13 the Court of Appeal declined to suspend the judgment while it hears them, “for lack of merit.” It will take the appeal itself on a fast track, and in the meantime, Vodacom says, the shares stay registered to Vodafone Kenya, the subsidiary that holds them, and the purchase money stays where it was paid. The comment period on the payment bill closed on October 9. Infrastructure that starts as an experiment ends as something a country goes to court over.

The argument here isn’t for deregulation in general. It’s for regulatory humility in the face of genuine innovation. The regulators who killed M-Pesa in India and South Africa weren’t evil; they were applying reasonable frameworks developed for traditional banking. But those frameworks assumed you needed to protect consumers from institutions that looked like banks. M-Pesa didn’t look like a bank. It looked like a phone service that moved money.

For Free Zones, the equivalent insight is: don’t try to regulate abundance infrastructure using scarcity frameworks. The rules that govern welfare programs, means-tested benefits, and poverty alleviation are designed for rationing limited resources. If resources aren’t limited, those rules become obstacles rather than protections.

M-Pesa shows what happens when you let infrastructure evolve to fit actual needs rather than forcing needs to fit existing infrastructure. The results are messy, imperfect, and sometimes exploitative (those transaction fees really do burden the poorest users). But they’re also transformative in ways that tidy regulatory frameworks never achieved.

Connection to the Unscarcity Vision

M-Pesa is, in many ways, the origin story for the Free Zone model.

It proved that:

Infrastructure can skip generations. You don’t need to build what wealthy nations built. You need to build what works.
Legacy absence is an advantage. The places with the least existing infrastructure are often best positioned for leapfrog innovations.
Distribution networks matter more than central facilities. M-Pesa’s 333,000 agents beat 1,611 bank branches because they’re where people actually are.
Technology is the easy part. The hard part is ecosystem: regulation, trust networks, market structure, and the willingness to let new approaches prove themselves.
Inclusion scales when marginal costs approach zero. Banks couldn’t profitably serve customers with $2/day income. M-Pesa could, because adding a new customer costs almost nothing.

The clearest demonstration is already running. The world’s largest basic-income experiment travels on M-Pesa: since 2018 GiveDirectly has paid about 23,000 people across 195 Kenyan villages $22.50 a month, some on a twelve-year commitment, and every transfer arrives by M-Pesa. No benefits office, no checks, no caseworker deciding who deserves it. The payment rail was already in everyone’s pocket.

The Free Zone article references M-Pesa directly when describing the Leapfrog Effect. Nairobi is proposed as one of the first Free Zone pilot locations precisely because the city already understands what infrastructure leapfrogging looks like.

What M-Pesa did for financial services, the Abundant Foundation aims to do for survival itself: housing, food, healthcare, education, energy. The mechanisms will differ and the scale is vastly larger, but the core insight holds. When existing infrastructure is failing people, the answer isn’t to fix the existing infrastructure. The answer is to build something that actually works.

Kenya’s corner shops became banks. Perhaps tomorrow’s corner shops become everything else.

References

Primary Sources

Safaricom: The M-Pesa Wonder - How It All Begun
Safaricom: M-Pesa - 17 Years of Transforming Lives
Wikipedia: M-Pesa
Safaricom: FY2026 results booklet (agents, active customers, transaction value and volume)
Communications Authority of Kenya: Sector Statistics Report, April-June 2026
Central Bank of Kenya: 2025 Bank Supervision Annual Report and the 2006 report
McKinsey: Driven by Purpose - 15 Years of M-Pesa’s Evolution

Research Papers

Leapfrog Technology

Recent Developments

Statistics

Ownership and Regulation (2026)

Kenya Law: High Court judgment on the Safaricom divestiture, 15 September 2026
Tech-ish: Vodacom Now Controls Safaricom: The KES 244.5 Billion Deal Is Done, But the Court Case Isn’t
Central Bank of Kenya: Draft National Payment System Policy and Bill, 2026 (comments closed October 9, 2026)
The Standard: Court of Appeal rejects State, Vodacom bid to suspend ruling nullifying Sh204.3b Safaricom sale (October 13, 2026) and Techpoint: Vodacom keeps Safaricom shares as Kenyan court fast-tracks acquisition appeal (October 19, 2026)

How M-Pesa Banking Skipped Banks: Kenya's $322B SMS Revolution