Africa’s fintech sector is forecast to grow revenue roughly 13-fold, to around $65 billion by 2030 – the fastest rate anywhere in the world, according to Boston Consulting Group. The figure has done what a headline number is supposed to do. Boardrooms across the continent are now debating how to move beyond payments into credit, insurance and cross-border transactions.
Far fewer are asking the question that decides whether any of it works: can the systems we run today actually carry those products?
For a lot of institutions, the honest answer is not yet. And closing that gap is a multi-year infrastructure job, not a product launch. The evidence sits in the same BCG report. Africa accounts for roughly 74% of global mobile money transaction volume, and yet more than half of all lending on the continent still runs through informal channels.
“That gap tells you where the opportunity is”, says Tertius De Bruin, a consultant at Johannesburg-based fintech consultancy Elenjical Solution.
“But it also tells you the infrastructure isn’t automatically ready to support it. Africa’s done an impressive job building payment infrastructure, but financial depth is a different challenge entirely.”
Credit is not payments at scale
Part of the reason the groundwork gets skipped is an assumption that moving from payments into lending is a question of scale: more features, more rules, the same basic idea. It isn’t.
A payment is a transaction between two parties – confirm the funds are there, move them, settle. Credit and insurance require decisions before any money moves: who qualifies, what the risk looks like, who is authorised to approve it. And they create obligations long after it does: monitoring, collections, settlement, regulatory reporting.
Those are different system requirements, not bigger ones and they need to be worked out before a new product is promised to customers.
What breaks first
Scale offers less protection than institutions expect.
Large banks and insurers have years of data, established regulator relationships and loyal customers, the kind of advantages a smaller player would kill for. In De Bruin’s experience, that same history routinely slows them down: platforms accumulated over years and patched repeatedly rather than properly replaced.
“It’s more friction building up over time,” he says.
“One system holds the customer data, another manages the product, another handles risk, and somewhere in between there’s a manual process stitching it all together. At low volumes you don’t notice. Once the business scales, those manual controls turn into bottlenecks, and eventually into real risk.”
That slow accumulation, rather than any single failure, is usually what brings an institution to a consultancy in the first place – not to tear the system out, but finally to understand what it can and cannot support. It is also why that risk and controls have to be designed in from day one rather than bolted on once a product is already live.
“The architecture itself decides what data you have access to and how decisions get made,” De Bruin explains.
“If you didn’t build the right controls into the workflow from the start, going back and fixing it later gets expensive fast.”
Cross-border ambitions raise the bar again. PAPSS, the Pan-African Payment and Settlement System, gives a sense of what is involved: coordinated participation across banks, pre-funding, standardised messaging, sanctions screening, settlement across several currencies and regulatory regimes, and a strong audit trail running through all of it.
Modernise alongside the business, don’t start over
Faced with all this, the tempting response is to scrap the existing estate and rebuild from zero. De Bruin says that almost never works out. The more reliable route is to modernise alongside the current system rather than against it – isolate the components that can be improved, test them properly, then migrate workloads across in stages.
“You can’t just switch off the system you’re running on to go build something new,” he says.
“The two need to run side by side while you migrate, carefully. It’s not unusual for that whole process to take around two years.”
That timeline is the strategic point. Institutions that wait for the revenue to arrive before they start will be two years behind the ones that started already.
Where to start
The first job is an audit rather than a procurement decision. What can the current estate carry, what has to be replaced, what can be left alone.
No new platform answers that question, and neither does an AI tool. De Bruin’s view is that the work is achievable but rarely a solo effort.
It needs a partner who knows the business as well as the technology, and who can modernise the plumbing without putting the business running on it at risk.
