Fintech challenges in South Africa: Credit-risk, data, and consumer trust
40,000 loans sharks (known locally as mashonisas) exist across South Africa. Let that number sit in your brain for a second.
It’s not the number that resonates with me; it’s what the number means. It means that for every one of those people, someone did not get approved for credit by a regulated institution. It means that those mashonisas are charging up to 50% interest. Per month. Per month! And yet somehow, consumers go to them because, more likely than not, they’ll never get credit anywhere else.
I’ve seen fintechs try to solve this for the last several years. Many fail. The few that don’t fail, solve for reality, not a fragmented version of other markets. They find solutions based on empirical findings from what’s available on the ground. The case in point today is digital short term loan lender Wonga South Africa, who now process over triple the short term loans volume compared to before its product and tech stack reform amid financial regulation overhauls in 2015. It’s not unique because they’re successful. It’s unique because they’re the only ones I’ve seen document the way through the minefield to get successful.
The data desert dilemma
What people fail to understand is that sometimes people are un-trackable in emerging markets for credit. In South Africa, a huge 40% of adults are essentially un-scorable; they can’t get credit through legitimate means. Not because they’re risky borrowers; they just have ‘no data’ as defined by traditional analytical standards.
No credit history. No formal job (if they do have one, it’s under the table with cash in hand). No fixed address sometimes. When banks scan these applications to predict risks, they see worrisome blanks where information should go. Risk models created for Western consumer systems get blown out of the water when looking through the lens of a cash worker who’s banked with stokvels and hasn’t set foot in a formal bank or found themselves on an official bank ledger.
Brett van Aswegen, the CEO who took over Wonga during its transition period ten years ago, asserts that if you solve problems for customers, you keep them. This should be obvious; yet in the emerging lender space, problems are never truly solved – they digitize the wreckage of exclusion and call it efficiency.
Yet the smart ones start finding data from previously unturned stones. How often do people use their phones? They must pay their bills somehow. How do M-Pesa transactions work in Kenya? Utility bills? Informal contracts? Based on the lack of traditional data, they’re now proxies for creditworthiness. Wonga changed its entire tech stack to make sense of alternative data input through machine learning algorithms. Not because it’s a fun thing to do; because the traditional approach leaves millions vulnerable and underserved.
Trust takes bytes to create
You know what kills digital lending uptake faster than bad user experience? Fear.
I sat through a focus group in Johannesburg where a loan platform app was previewed (several years ago). The aesthetics suggested legible forms and potentially instant approval. Half the room thought it was a scam – a joke – too good to be true. Why? Because it was too easy. Misinformation is such a common occurrence in this economy that ease breeds suspicion. People were accustomed to in-person agreements, meetings, someone telling them what documents they needed.
According to The Bank for International Settlements, globally 21% of women engage with fintech resources compared to 29% of men; it’s worse in developing nations than developed ones. Why? Because it’s more than access – it’s understanding and trust with agencies they’d never interacted with before. If they’ve been treated poorly or turned into indentured servants every time they’ve borrowed, why does it make sense to go digital when it seems safer not to?
The companies who get ahead without the slick aesthetics appeal through education first – product second. Wonga’s proprietary Money Academy has reached more than 300,000 South Africans through campaigns for financial literacy awareness – for free, no strings attached, no markup for immediate sale. They teach people how credit works, how to budget, how to avoid predatory lending practices. Others might call this a long-term value-added marketing strategy; I call it the only thing effective with this population since everyone else is working on borrowed trust.
Language also plays a critical role: there are eleven official languages in South Africa; every loan document needs translation into at least two languages so every borrower knows what they’re doing in-depth. Sounds easy? Good luck translating compound interest into Zulu when half the financial terms don’t have direct equivalents. Companies that view translation merely as a compliance box check fail; those who genuinely seek linguistic access, who test their explanations with real users – those are the trust builders.
Regulatory confusion meets industry disruption
Of course there exists the National Credit Act in South Africa, that seeks to protect against predatory lending – but only enforced through courts and fines. There’s an interest cap in place; no automatic increases are allowed; debt counseling is required so any subsequent options would be more beneficial than initially seeking loans elsewhere.
Yet 40,000 illegal, unregulated lenders exist beyond the scope of this framework.
What regulators fail to understand is that you cannot regulate people out of poverty. When legitimate short-term loans become impossible, informal ones don’t disappear – they smell blood in the water… they become more predatory. This means that the regulatory framework needs balance: too strict, and people go underground to illegal short term cash-based lenders; too loose, and regulated lenders become loan sharks with HTML-ready and attractive sites.
The fintech that survives needs to work with regulators instead of around them. When van Aswegen’s first day as CEO started with regulators at the door, he could have resisted and manoeuvred around the restrictions. Instead, Wonga created proactive measures which cut its income by two-thirds overnight – forcing the company to understand its lessons first – and forced stabilization by reworking the model instead of keeping the ‘churn-and-burn’ mentality that ultimately took down the brand’s parent company in the UK.
In chaotic regulatory environments, compliance is a competitive advantage. While others spin their wheels trying to loophole their way out of compliance, companies that embrace transparency can move faster, knowing down the line they can expand internationally since their model won’t be crushed at every regulatory turn.
The fraud factor no one talks about
Identity fraud in developing markets doesn’t look like identity theft fraud in the West – it’s messier, more personal, often involving family or trusted associates. In village settings where someone approaches for a loan from rural Eastern Cape – how do you know they’re legitimate? Government IDs present new systems still rolling out – but they’re too young in development to see extensive reliability; biometric systems help – but there’s no foolproof guarantee.
Innovation isn’t in prevention – it’s in recovery. Traditional financial institutions write off fraud as part of the business plan; fintechs do not have this luxury in developing worlds – they need systems that can figure out unusual activity without flagging complaints on legitimate transactions from those who don’t fit mapped out Westernized credit patterns; collection efforts help when the fraudster is the applicant’s nephew using their ID with quasi-consent.
Building for reality – not perfection
The fintechs that succeed are those who do not have the best algorithms or Series A funding – but those who can validate failure AND success based on empirical findings and research – not broken angles from developed worlds – but entirely different ecosystems requiring different solutions.
Approach it as though exploitation was part of the credit game before you began. Weave education into your product as something valuable without fear of it biting you back with outcome metrics later on down the line. Use whatever information you can get your hands on legally and ethically, since conventional credit scores are luxury items across Africa.
Translate your assumptions and expectations into easily digestible language in applicable local languages instead of just technical translation, especially when seen in real life without guaranteed comprehension.
Work with regulators even if it hurts your current program now but keeps it safer down the line when others go bankrupt fighting regulatory compliance efforts.
And finally – solve real problems. The last thing people need is another goddamn loan app – they need credit solutions that won’t ruin their lives; they need financial tools that track inconsistent income streams, informal employment and cash-based considerations; they need legitimate companies to think of them as viable consumers instead of risks.
The numbers are staggering – Sub-Saharan Africa’s mobile money provision topped $832 billion in 2022 – financial inclusion is getting better but millions are left outside formal systems they should be part of.
The fintechs who get this right will profit but also operate with ethical consideration that transforms how billions approach financing for life. Every loan shark exists because formal financing failed someone at some point – and your job isn’t to digitize that failure – but actually fix it.
