Cash is not the past. Across much of Africa it is still the default settlement asset — trusted, anonymous, accepted everywhere, and free at the point of use. Every digital scheme that wants to win has to solve the boring, expensive problem of turning cash into value and back again.
Cash-in/cash-out is the on-ramp and off-ramp between physical money and digital value. Get it wrong and nothing else matters.
Cash is physical legal tender — notes and coins. Cash-in / cash-out (CICO) is the act of converting cash into electronic value (cash-in: you hand over notes, your wallet or account is credited) and back out again (cash-out: you withdraw notes against your balance). CICO is the hinge the entire digital-payments story turns on.
In most of Sub-Saharan Africa the question is not “card or wallet?” — it is “how does a person who is paid in cash, by an employer who holds cash, get value into a system at all?” Until that answer is cheap, near, and trusted, digital money stays a thin layer on top of a cash economy.
The actors that do the converting — corner-shop agents, mobile-money kiosks, ATMs, bank branches, retailers running cash-back — are collectively the CICO network. They are physical, they need float, and they are the most under-appreciated piece of payments infrastructure on the continent.
There is no single CICO rail. There is a patchwork, and its weakest link — usually rural agent liquidity — sets the real ceiling on adoption.
| Channel | How value moves | Reach | Economics / friction |
|---|---|---|---|
| MM agent | Customer hands cash to agent; agent debits own float, credits customer wallet (or reverse) | Densest network in Africa — ~30m registered mobile-money agents globally in 2025, most active growth in East Africa | Agent earns a commission; needs working float & e-value on hand; rural liquidity is the constant problem |
| ATM | Card or cardless code authorises a withdrawal against a bank/wallet balance | Urban-skewed; capital-heavy to deploy and refill | High per-unit cost; cash replenishment & security dominate; declining in some mature markets |
| Bank branch / teller | Deposit or withdrawal over the counter | Sparse outside cities | Highest cost-to-serve; the channel financial inclusion was built to replace |
| Retail / merchant cash-back | Shopper buys goods and draws cash, or pays in cash that is digitised at till | Wherever organised retail reaches | Piggybacks on existing footfall; limited by the till’s own cash position |
No fee, no balance enquiry, no failed transaction. For someone earning a few dollars a day, a 2–5% cash-out fee on every transaction is a real, repeated tax that digital often cannot beat.
No network, no power, no smartphone, no KYC. In a country where connectivity and grid power are intermittent, cash never declines.
Every counterparty takes it. Digital acceptance is patchy — the person you owe may not be on your network or any network.
No data trail, no platform that can freeze it. For the informal economy — the majority of economic activity in many markets — that is a feature, not a bug.
If you are paid in cash by an employer who is paid in cash, digitising means an extra step and an extra fee before you can spend. Cash-in only the value you must.
People trust the agent they can see, in the shop they know. Trust in a balance on a screen is earned slowly, and lost fast after one bad experience.
It is tempting to treat cash use as a problem to be scolded away. It is not — cash persists because it is genuinely good at things digital still does badly in much of the continent.
Printing, minting, distributing, securing, counting, reconciling, and destroying notes is a continuous, capital-intensive operation for central and commercial banks. None of it shows up on the spender’s receipt.
Moving physical money means armoured vehicles, guards, insurance, and risk. In South Africa, cash-in-transit heists are a recurring, violent feature of the security landscape.
Cash leaves no record. No transaction history means no credit score, no underwriting, no targeted product — the informal majority stays invisible to formal finance.
Untraceable value is friendlier to tax evasion, corruption, and illicit flows than auditable digital value — one of the few unambiguously public-interest arguments for digitisation.
Cash is free at the point of use and costly everywhere else. The cost simply sits with people other than the payer — central banks, commercial banks, retailers, and ultimately the economy at large.
A wallet you cannot cash out of is a trap, not a service. If the nearest agent has no e-float (can’t take your cash-in) or no physical cash (can’t pay your cash-out), the service has failed — and the user blames the brand, not the agent. Rural liquidity rebalancing is the unglamorous core operational problem of mobile money.
Many users cash-in only to immediately cash-out somewhere else — the float never circulates as digital value. You have paid for two CICO events and digitised nothing. Real progress is measured by value that stays digital, not by registered accounts.
Price the off-ramp wrong and users rationally hoard cash, defeating the purpose. The off-ramp fee is the single most behaviour-defining number in the whole model.
“X million accounts” is a vanity metric if most are dormant or cash-in-cash-out conduits. Watch 30/90-day active users and value retained digitally.
Agents handle other people’s money and trust. Direct-deposit scams, fake reversal messages, and over-charging on cash-out are common; weak agent oversight erodes the trust the whole network runs on.
Start with the off-ramp, not the app. Map agent density and liquidity in your target geographies before you build features. A beautiful wallet with no reliable cash-out within walking distance is dead on arrival.
Price the cash-out to keep value digital. Every percentage point on cash-out pushes users back to notes. The goal is to make digital value worth keeping — merchant acceptance, P2P, bill pay — so the off-ramp is used less, not to tax it harder.
Treat agents as infrastructure, not a channel. Commission economics, float financing, and liquidity rebalancing decide whether your network survives in low-density areas. This is an operations and treasury problem dressed up as a product problem.
For the smallest, most frequent, in-person transactions among unbanked counterparties with no reliable connectivity, cash is often genuinely the lowest-friction, lowest-cost option. Forcing digital here destroys value.
Cost of getting it wrong: over-investing in acceptance where there is no liquidity, or pricing the off-ramp so high that you train users to distrust digital value. Both produce impressive registration numbers and a dead network. The expensive mistake is mistaking sign-ups for circulation.