The African fintech sector attracts billions in investment while producing a generation of product managers that are shaping the future of finance. But with the rise of AI, a new dilemma is surfacing. It is one that many AI product managers are not fully prepared for.
Right now, AI is helping shape product strategy, automation, and personalization across global tech industries. This has caused many African AI product managers to work with a playbook that was not designed for Africa.
A Playbook That Doesn’t Fit
The global playbook built perhaps in Silicon Valley assumes that certain infrastructures already exist. It assumes that there is stable electricity, widespread smartphone adoption, strong data infrastructure, and digitally literate users. It also assumes access to clean, structured datasets and clear regulatory frameworks. The most crucial assumption it makes is that the society it is being used in is one that is 100% digital-first.
But in many parts of Africa, that is really not the case.
So if we continue to try to force this global AI playbook into the African fintech sector, we risk creating financial exclusion and not solving it.
The Misfit of Global AI in Africa Context
There’s this certain belief that is unspoken but apparent. It is the belief that whatever works in the West will eventually work somewhere else. That all that society needs to do is to create infrastructure improvements and give it time. This is the default approach that guides many venture-backed African startups.
Why build or make improvements for USSD when smartphones continue to be the future?
Why collect offline data or analog data when AI excels with digital outputs?
In 2024, the US-based Digital Currency Group (DCG), venture capital firm made 12 new investments across six African countries. Yet many of these fintech startups are designed as if they operate in environments that mirror the U.S. or European markets. So what we have is AI models that are trained on Western data sets being deployed into the African market.
This unfortunately cannot be the way forward. AI systems require large volumes of high-quality data to function effectively. But many African fintech markets are still heavily cash-based, and a large portion of the population remains unbanked or underbanked.
Take Lagos, the “fintech capital” of Africa for example. It has had successes like the $110mn funding round for fintech group Moniepoint giving it a valuation of at least
$1bn.
This same city suffers from issues like recurring power outages and inconsistent data regulation. According to the Financial Times, you might be able to reach millions quickly in Lagos but the lack of infrastructure, loose data protection laws, still exist.
So when an AI product in this kind of society depends on continuous internet access, it’s not just an inconvenience it’s a product risk.
In this setting, it is important that every feature you ship must answer certain questions:
Will this work offline? What will break the system? How do we build trust?
The Rise of Hybrids
In response to the need for AI, some African fintechs are creating their own rules by embedding it with human agents to reach every part of the market.
For example, there’s M-KOPA, a pay-as-you-go solar and smartphone provider operating in Kenya and Uganda. The company uses AI to predict creditworthiness and customer turnover.
But most importantly, it uses a distributed network of field agents to bridge the trust gap by explaining the tech in local languages. These field agents also help verify users, explain the repayment terms, and collect contextual data that algorithms can’t reach.
Similarly, Wave, is a mobile money provider in West Africa that uses QR codes with field agents support. These field agents help users activate accounts, resolve disputes, and learn the system. This helps bridge the trust gap that AI-only interfaces can’t yet solve.
In these hybrid models, AI is just simply a copilot that helps the human systems instead of automating them. This helps cement the fact that the infrastructure that matters most in African fintech spaces is relational.
The Ultimate Product Strategy: Localization
Perhaps the biggest mistake AI product managers make in Africa is underestimating the power of localization.
One great example is Safaricom’s approach to credit scoring. Instead of relying on traditional methods like payslips or credit histories which many Kenyans simply don’t have. Safaricom uses alternative data such as airtime usage, mobile money activity, and transaction patterns. This helps create a more inclusive and responsible lending model that gives financial access to various people who have been previously locked out of the system.
Safaricom also recently introduced solutions like Pochi la Biashara, to help micro-traders manage business finances more effectively. There’s also Taasisi Till Loans which gives small businesses access to working capital without paperwork or collateral, simply by analyzing their transaction history.
These aren’t just features, they’re a small but critical shift from enabling one-time transactions to building a long-term financial ecosystem that works. By designing with local behaviors and constraints in mind, AI product managers move from building for an idealized user to building for actual users.
What the African Fintech Playbook Needs
The dilemma facing the African AI product manager isn’t just about the lack of funding, it’s about resisting the temptation to build things for the European market while living in Accra or Lagos.
AI when used in the African fintech market must be grounded in local knowledge. Hence to move forward, African AI product managers must shift their mindset.
They must start designing for offline-first environments, not cloud-first. Unfortunately, within Africa exist regions where internet access is unreliable and power outages are routine. So products must function without requiring constant connectivity. It’s important that products have offline compatibility.
But infrastructure is only part of the equation. AI product managers must also recognize trust and cultural insight as an integral part of their solutions. Financial behavior in most African contexts is shaped by cultural norms and informal lending. To ignore these factors would be to choose to widen the gap between the product and its users.
That’s why AI solutions can’t work in isolation. Human support systems must be built into AI flows, not as a fallback but as an integral part of the design. There is a need for field agents and community ambassadors. These humans help make the system work.
And finally, analog signals like seasonal trends and community feedback must be respected and integrated into product strategy. They might not be captured in a dataset or modeled in a dashboard, but they’re vital in understanding how your users actually live, spend, and save.
-Article written by Joshua Jordan-Akintoye
