The Mobile Money Agent Network: Infrastructure Without Branches
Mobile money is often described as “banking without branches.” But the description is only half-right. While mobile money does not require a brick-and-mortar bank building, it depends on a different kind of physical infrastructure: the agent network. Agents are the human face of mobile money — the local shopkeepers, kiosk operators, and small business owners who let users deposit cash into and withdraw cash from their mobile wallets.
This network of cash-in/cash-out points is arguably the most important piece of infrastructure in mobile-money-enabled financial systems. Without agents, a mobile wallet is just a number on a phone screen. With them, it becomes a gateway to the financial system for people who live miles from the nearest bank branch.
How Agent Networks Scale
The IMF Financial Access Survey tracks the number of registered mobile money agents per 100,000 adults across 55 economies. The density varies by more than three orders of magnitude:
| Rank | Country | Agents per 100,000 adults | Year |
|---|---|---|---|
| 1 | Nigeria | 17,665.9 | 2022 |
| 2 | Thailand | 5,068.8 | 2024 |
| 3 | Zambia | 4,675.8 | 2024 |
| 4 | Ghana | 4,241.2 | 2024 |
| 5 | Eswatini | 3,987.5 | 2024 |
| 6 | Benin | 3,844.5 | 2024 |
| 7 | Malawi | 3,828.5 | 2024 |
| 8 | Tanzania | 3,744.6 | 2024 |
| 9 | Uganda | 3,238.9 | 2024 |
| 10 | Lesotho | 3,140.6 | 2024 |
Nigeria’s figure of 17,666 per 100,000 adults stands out as an outlier — roughly one agent for every six adults. This reflects both the explosive growth of mobile money in Africa’s largest economy and Nigeria’s distinct market structure, where telco-led mobile money services (like MTN Mobile Money and Airtel Money) compete for agent networks alongside traditional bank agents.
Thailand at 5,069 per 100,000 adults is the only non-African economy in the top five, driven by the ubiquity of PromptPay booths and 7-Eleven-based cash-in points that have made mobile payments accessible even in remote rural areas.
At the other end, the bottom of the list tells a different story:
| Rank | Country | Agents per 100,000 adults | Year |
|---|---|---|---|
| 53 | Iraq | 50.8 | 2022 |
| 54 | Bahamas | 42.5 | 2024 |
| 55 | Vietnam | 15.3 | 2024 |
| 56 | Turkiye | 11.4 | 2024 |
| 57 | Mauritania | 7.8 | 2021 |
In these countries, mobile money exists — the accounts are there — but the agent network is too sparse to make it practically useful for cash-in/cash-out. The difference between Nigeria and Mauritania is more than 2,000-fold.
Agents vs. Bank Branches: Two Infrastructures
To understand what agents achieve, compare agent density to bank branch density in the same countries:
| Country | Agents per 100k adults | Bank branches per 100k adults | Ratio |
|---|---|---|---|
| Nigeria | 17,665.9 | 4.4 | 4,015:1 |
| Zambia | 4,675.8 | 2.7 | 1,732:1 |
| Ghana | 4,241.2 | 4.4 | 964:1 |
| Uganda | 3,238.9 | 2.3 | 1,408:1 |
| Tanzania | 3,744.6 | 2.3 | 1,628:1 |
| Thailand | 5,068.8 | 8.3 | 611:1 |
| Kenya | 1,069.2 | 4.5 | 238:1 |
| Turkiye | 11.4 | 14.4 | 0.8:1 |
In Sub-Saharan Africa, mobile money agents outnumber bank branches by ratios ranging from 238:1 in Kenya to more than 4,000:1 in Nigeria. This is what “infrastructure without branches” actually looks like: an agent network that spreads financial access to neighbourhoods and villages no bank branch ever reached.
For perspective on how the branch network is evolving, see our analysis of where banking still means walking in and our global ranking of bank branches and ATMs per capita.
What Agents Actually Do
Agents perform a deceptively simple function: they convert physical cash into digital value and back again. A customer hands cash to an agent, who credits the customer’s mobile wallet. Another customer wants cash in hand and asks the agent to debit their wallet in exchange for banknotes. These two operations — cash-in and cash-out — make mobile money useful for everything from paying school fees to buying groceries to receiving remittances.
The economics are straightforward but thin: agents earn a commission on each transaction, typically 0.5–3% of the value. In high-volume markets like Ghana and Tanzania, a busy agent might process hundreds of transactions per day, making the business viable even at low margins. In low-volume markets like Vietnam or Turkiye, the sparse network may itself be a symptom of insufficient transaction volume to sustain the agent business model.
The relationship between agent density and transaction activity is mutually reinforcing, as our Global Mobile Money Report 2026 notes: more agents attract more users, which generates more transactions, which supports more agents. Countries that have crossed a certain density threshold — roughly 1,000 agents per 100,000 adults — tend to see a virtuous cycle of network expansion.
Where Agent Density Meets Account Ownership
Agent density correlates with — but does not perfectly predict — mobile money account adoption. Compare the top agent-density countries with their account-ownership rates from the Global Findex:
| Country | Agents per 100k | Mobile money accounts (% of adults) |
|---|---|---|
| Nigeria | 17,665.9 | 32.8% |
| Ghana | 4,241.2 | 78.3% |
| Zambia | 4,675.8 | 69.3% |
| Uganda | 3,238.9 | 67.7% |
| Tanzania | 3,744.6 | 52.9% |
| Thailand | 5,068.8 | 41.7% |
| Kenya | 1,069.2 | 87.5% |
The relationship is not linear. Kenya, the world leader in mobile money account ownership at 87.5% of adults, has only 1,069 agents per 100,000 adults — near the middle of the pack. This suggests that as a market matures, the agent network can become more efficient (fewer agents processing more transactions each) rather than continuing to grow in raw numbers.
For the full ranking of mobile money account ownership, see our article on the top 10 mobile money countries.
The Global Distribution
Of the 55 economies with agent data, the median sits at roughly 850 agents per 100,000 adults. The top quartile exceeds 2,000; the bottom quartile falls below 150. Geographically, the distribution is heavily weighted toward Sub-Saharan Africa, which accounts for 13 of the top 15 positions. No country in Europe or North America registers a meaningful agent density — not because mobile money doesn’t exist there, but because those markets rely on card rails, bank transfers, and digital wallets that operate without a cash-in/cash-out agent layer.
The gap between Australia and Nigeria is not a failure of Australian fintech. It is a reflection of infrastructure path-dependency: where bank branches and ATMs are already dense, the business case for agent networks never arises. For a look at how the world pays for things through different rails, see our comparison of credit cards vs. mobile money.
Why Agent Networks Matter for Financial Inclusion
The agent network is arguably the single most important infrastructure for extending financial access to the unbanked. The World Bank’s Global Findex data shows that the unbanked disproportionately live in rural areas where bank branches are rare or absent. In those same areas, small shops and kiosks are everywhere — and converting them into financial access points is far cheaper and faster than building bank branches.
Countries that have deliberately built agent networks as part of their financial inclusion strategy — Ghana, Tanzania, Zambia — have seen mobile money account ownership rise rapidly alongside agent density. Countries where regulation has made it difficult to register or compensate agents — Turkiye, Mauritania — show low agent density and correspondingly low mobile money adoption.
For a broader look at how different financial access gaps interact, see our analysis of the digital payments divide and the online bill payment gap.
Sources & Method
Agent density figures come from the IMF Financial Access Survey (FAS), which collects annual data on the number of registered mobile money agents from central banks and financial regulators in 55 economies. The metric is expressed per 100,000 adults for comparability across countries. Data ranges from 2020 to 2024 depending on the country’s reporting lag.
Mobile money account ownership figures (where cited) come from the World Bank Global Findex 2024 survey wave. Bank branch density uses the World Bank’s WDI indicator FB.CBK.BRCH.P5 (commercial bank branches per 100,000 adults).
All numbers cited are the latest available observations. No projections were used. For details on our data processing methods and calculations, see the methodology page.