How Safaricom and M-Pesa Actually Make Money

The standard description of Safaricom goes something like this: Kenya's largest telecom company, the firm that built M‑Pesa, the mobile money platform that brought financial services to millions of unbanked Africans. The description is accurate and almost entirely useless for understanding how the business actually works.

Safaricom is not primarily a telecommunications company that happened to build a payments app. It is, by revenue, a financial services company that also operates a telecom network. M‑Pesa, the division everyone treats as a feature, is the engine. Voice calls, the service that defined Safaricom's first decade, are now a declining side business. Mobile data, the growth story every telecom analyst focuses on, comes second. M‑Pesa comes first, and it is pulling further ahead every year.

M‑Pesa generates roughly 40 percent of Safaricom's total revenue. It processes transactions worth more than half of Kenya's annual GDP. It holds customer deposits in commercial banks, earns interest on those deposits, and pays customers zero shillings of that interest income. It operates a lending business through a product called Fuliza that disburses hundreds of billions of shillings annually at effective rates that would be considered predatory in any regulated credit market. This article is the account that does not exist anywhere in full.

Safaricom Limited is listed on the Nairobi Securities Exchange. Its shareholders are not who most people assume. The single largest shareholder is Vodacom Group, the South African telecom that is itself majority‑owned by Vodafone UK. Vodacom holds approximately 35 percent of Safaricom. The government of Kenya, through a combination of direct treasury shareholding and a stake held via Telkom Kenya, holds another approximately 35 percent. The remaining roughly 30 percent is publicly traded and widely held, with significant institutional ownership concentrated in Kenyan pension funds.

This ownership structure matters for understanding the company's political position. Safaricom is not simply a private enterprise operating in Kenya. It is a company in which the Kenyan state has a substantial economic interest, and that creates a regulatory dynamic very different from a purely commercial relationship. The government of Kenya has a direct financial incentive for Safaricom to be profitable.

Safaricom holds approximately 64 percent of Kenya's mobile subscriber market, and a significantly higher share of active data users and mobile money accounts. Its nearest competitor, Airtel Kenya, has been consistently unprofitable. The competitive landscape is not a duopoly; it is a monopoly with cosmetic competition around the edges.

Safaricom reports revenue across five primary segments: M‑Pesa, Mobile Data, Voice, SMS, and Fixed‑line and other services. The trajectory of each tells a distinct story about where the company has been and where it is going.

These figures are approximations derived from Safaricom's annual reports and investor presentations. M‑Pesa grows. Data grows. Voice declines. SMS collapses. The company that was once defined by voice minutes is now defined by mobile money, and the gap is widening.

M‑Pesa was launched in 2007 as a joint venture between Safaricom and Vodafone, initially positioned as a loan repayment tool for microfinance borrowers. By 2010, it had more customers than Kenya's entire formal banking sector combined. Today, with over 32 million active users and a network of more than 200,000 agents, it is the financial infrastructure of the Kenyan economy in the most literal sense.

The common understanding of how M‑Pesa makes money is incomplete. Most descriptions identify transaction fees as the primary revenue source. Transaction fees are real and material. They are also only part of the picture. M‑Pesa earns money from five distinct mechanisms, and the one that receives the least public attention generates the most passive income of all.

Revenue Stream 1: Transaction Fees

The most visible revenue source. Every peer‑to‑peer transfer carries a tiered fee based on transaction size. Transfers of KES 100 and below are free. Transfers between KES 101 and KES 500 cost KES 7. The fees scale upward through multiple transaction bands, reaching a maximum of KES 105 for transfers between KES 70,001 and KES 150,000. Cash withdrawals at agents carry separate fees.

Safaricom processes hundreds of millions of transactions every month. In FY2024, total M‑Pesa revenue across all fee‑bearing services exceeded KES 117 billion. Transfer fees alone remain the foundation on which everything else is built.

When a customer deposits cash into M‑Pesa through an agent, the physical cash is collected and deposited into one of multiple designated commercial banks in Kenya. The corresponding e‑money balance is credited to the customer's account. The customer holds a legal claim on that cash. Safaricom holds the cash.

These pooled customer deposits are collectively called the M‑Pesa Trust float. At any given moment, the float represents the total outstanding M‑Pesa balances held by all customers across the country. Estimates suggest the float consistently exceeds KES 100 billion, with peaks above KES 130 billion.

Interest is earned on this float. The funds are deposited into interest‑bearing accounts and government securities by the Trustee. Even at a conservative effective rate of 8 to 10 percent, a float of KES 120 billion generates between KES 9.6 billion and KES 12 billion in annual interest. Under Central Bank of Kenya regulations, this interest does not go to Safaricom's commercial bottom line; instead, it is retained by the Trust to cover administrative costs or channeled into public charitable causes. The individual customers who generated this float earn zero interest. Their standard M‑Pesa balance does not accrue anything.

Every mobile money operator that holds customer float generates interest on it under strict regulatory oversight. The scale of M‑Pesa's float, driven by Kenya's uniquely high adoption rates, simply makes this capital pool unusually large. No other mobile money operation anywhere runs a customer trust float at this scale relative to its home economy.

Launched in January 2019 in partnership with KCB Group and NCBA Bank, Fuliza is M‑Pesa's automatic overdraft facility. When a customer has insufficient balance to complete a transaction, Fuliza automatically extends a short‑term credit line. No application is required. Repayment is automatic: the next time any funds enter the account, Fuliza deducts what it is owed before the customer can access the balance.

Fuliza charges a daily facility fee rather than an interest rate. For amounts between KES 501 and KES 1,000, the fee is KES 10 per day. If a customer borrows KES 1,000 and takes seven days to repay, the total fee is KES 70. Annualized, this represents an effective interest rate in the range of 350 to 1,000 percent per annum. It is structured as a facility fee rather than interest to stay outside the Microfinance Act.

In FY2024, Fuliza disbursed over KES 800 billion in cumulative loans. The credit risk management works through the automatic deduction mechanism. When M‑Pesa receives any incoming funds, Fuliza is repaid before the customer can access anything. The default rate is structurally lower than conventional lending precisely because repayment is automated and non‑negotiable.

M‑Shwari is a savings and loan product integrated directly into the M‑Pesa interface, operated in partnership with NCBA Bank. Customers can move funds from their M‑Pesa wallet into an M‑Shwari savings account, where they earn a modest interest rate. They can also access M‑Shwari loans against their savings history and M‑Pesa transaction behavior. M‑Shwari charges a one‑time 7.5 percent facilitation fee on any loan, regardless of the loan's duration. Annualized, this represents 90 percent per annum. As with Fuliza, it is structured as a facilitation fee rather than interest.

M‑Shwari's credit scoring model relies entirely on M‑Pesa behavioral data: transaction frequency, average balance, tenure, and defaults on Fuliza or previous M‑Shwari loans. This proprietary credit scoring model, built entirely from data that Safaricom owns, is one of the company's most strategically valuable assets.

Lipa na M‑Pesa is the merchant payment product. Businesses register for a Till Number or Paybill Number. Customers pay merchants by entering the number in their M‑Pesa menu, or by scanning a QR code. Safaricom charges merchants between 0.5 and 1 percent of the transaction value, sitting in the position that Visa or Mastercard occupies in conventional card markets. As of 2024, Safaricom reports over 600,000 active till and paybill numbers, including major utilities, government entities, and large retailers.

There are over 200,000 registered M‑Pesa agents in Kenya. This network is not owned by Safaricom. Agents are independent contractors who provide cash‑in and cash‑out services. They earn commissions, and in FY2024 agent and channel costs exceeded KES 20 billion.

The critical structural point is who bears the operational risk. Agents must hold physical cash and electronic M‑Pesa float simultaneously. Safaricom does not provide this capital. Agents fund their own float entirely. They bear the operational risk of holding cash, the regulatory compliance obligations, and the reputational risk when transactions fail. From Safaricom's perspective, this arrangement is extraordinarily efficient: a nationwide financial distribution network covering 200,000 locations without the capital expenditure of branches.

In 2021, a consortium led by Safaricom won a license to operate Ethiopia's second mobile network. Ethiopia has a population of approximately 120 million people, largely unbanked and almost entirely underserved by digital financial services. It is the obvious candidate for M‑Pesa expansion after Kenya.

In practice, the Ethiopia operation has been extraordinarily difficult. Infrastructure rollout requires massive capital expenditure. The Tigray conflict disrupted operations. The Ethiopian birr has weakened sharply against the US dollar. Regulatory friction has complicated spectrum allocation and M‑Pesa licensing timelines. As of Safaricom's most recent financial disclosures, the Ethiopia operation remains loss‑making, with cumulative losses in the billions of Kenya shillings.

The investment thesis remains structurally valid. A market of 120 million people with almost no formal financial services access is exactly where M‑Pesa's model should eventually work. But the timeline has extended well beyond early projections, and the losses are ongoing. Whether and when that bet pays off is the central uncertainty in the Safaricom investment case today.

The first is interoperability regulation. Full mobile money interoperability would reduce the value of M‑Pesa's network advantage. The Central Bank of Kenya has pushed for genuine interoperability, and more aggressive mandates could change the landscape. The second is bank‑owned mobile products. KCB and Equity Bank have invested heavily in their own mobile banking platforms, limiting M‑Pesa's ability to expand into high‑value financial services. The third is the Ethiopia drag. If the Ethiopia operation continues to generate material losses beyond the market's tolerance, investor sentiment will erode. The fourth is currency exposure. Safaricom earns virtually all its revenue in Kenyan shillings, while its costs include significant US dollar‑denominated items. As the KES weakens, cost pressures compound.