Every few years, a new initiative promises to "connect" African payments. A new switch, a new protocol, a new regional platform. Build the infrastructure, and adoption will follow.

It hasn't followed. In a new paper, Beyond the Build, I argue that this technology-first diagnosis is fundamentally wrong, and that getting it right matters enormously.

Thirty-three systems. Zero breakouts.

Africa already has 33 domestic instant payment systems and 3 regional ones. That's not an infrastructure deficit. Yet not one has achieved the breakout adoption seen in India or Brazil. Only one is classified as inclusive. Most can't talk to each other.

The cost is staggering. Sending money to Sub-Saharan Africa costs 8–8.5% on average, nearly triple the UN's 3% target. A $200 transfer from South Africa to Malawi can eat $48 in fees. Low-value cross-border payments lose up to 30%. In total, routing intra-African payments through correspondent banks in Europe and the US drains an estimated $5 billion from the continent annually, not because the technology doesn't exist, but because existing systems are too fragmented to route payments directly.

The clearest evidence that formal rails are failing? Stablecoin usage across Sub-Saharan Africa now represents roughly 6.7% of GDP, one of the highest rates globally. That's not crypto enthusiasm. It's millions of people deciding formal systems are too expensive and too slow, and routing around them. Every dollar flowing through USDT is a dollar that could flow through African payment systems — if those systems offered comparable value.

The real barriers are commercial, not technical

I identify three commercial frictions that throttle adoption: pricing, accessibility, and addressing.

Even newer systems charge fees that kill adoption. South Africa's Payshap charges seven rand per transaction, enough to make it uncompetitive with cash. On accessibility, systems that banks bury deep inside their apps predictably fail; the UK's Paym is the cautionary tale. Brazil's PIX is the counter-example: the central bank mandated that the PIX button appear in the top half of every banking app's home screen. That single regulatory decision — requiring no new technology — was one of the most consequential design choices in modern payments. And on addressing, systems requiring 16-digit account numbers and sort codes can't compete with proxy-based systems where you send to a phone number. Mobile money's 1.4 billion accounts, India's 20 billion monthly UPI transactions, Brazil's 5.7 billion monthly PIX transactions: all run on proxies.

Free is not a nice-to-have. It's the ignition mechanism.

Person-to-person payments must be free. Not cheap. Free. Every system that achieved genuine scale, UPI, PIX, PayNow, PromptPay, Swish, Venmo, Alipay, WeChat Pay, made this choice.

India's experience is the clearest proof. IMPS launched in 2010 with per-transaction fees and saw modest adoption. UPI launched in 2016 with identical technology but free P2P transfers. It now processes over 20 billion transactions per month. Same rails. Different business model. Radically different outcome.

"Zero-fee" doesn't mean "zero-cost" — infrastructure costs must be recovered somewhere. India's UPI monetization remains a live tension. The question for African markets is whether cross-subsidy models work in economies with thinner margins and less state capacity. I argue the answer isn't guaranteed but isn't hopeless either. M-Pesa proved it: free P2P transfers funded by cash-out fees, merchant payments, and adjacent services like credit. The principle holds that someone pays, but it needn't be the end-user making a simple transfer. Designing who pays is the core innovation required.

A sequenced reform agenda

The paper proposes a deliberate sequence, because political capital is finite.

Start tomorrow with the lowest-cost intervention: mandate prominent app placement and proxy-based addressing within existing systems. No new infrastructure, no subsidies — just regulatory will.

Then tackle domestic pricing by mandating zero-fee P2P transfers at the national level, where a single regulator holds authority.

Only then take on cross-border interlinking, the genuinely hard problem requiring coordination across sovereign regulators and diverse legal frameworks.

And critically, acknowledge who loses. Banks profit from fragmentation. Transaction fees are visible revenue; the diffuse benefits of an inclusive ecosystem are not. The path forward requires either compensating incumbents by letting them capture value elsewhere, as Brazil did, or building coalitions of fintechs, mobile operators, and merchants as political counterweights.

The bottom line

Africa's payments problem was never about technology. Thirty-three instant payment systems are proof of that. The demand is proven by mobile money's success and the stablecoin migration happening whether regulators engage with it or not. What's missing is business model courage,  the willingness to make transfers free, the regulatory muscle to mandate good UX, and the political strategy to overcome incumbent resistance.

The blueprints are clear. The cost of inaction is immense.

Read the full paper →