In this exclusive interview, Moses Akampurira, Digital Products Lead at Letshego Uganda, gives us a front-row seat into how this breakthrough is unfolding on the ground. Powered by real-time behavioural data such as mobile money activity, airtime usage, and phone patterns—combined with traditional financial records—gnuGrid’s Mobile Credit Score provides a dynamic, AI-driven risk profile for borrowers who were previously invisible to the formal credit system. For Letshego, this isn’t just about smarter lending—it’s about building a more inclusive and disciplined financial ecosystem, one borrower at a time.
To start us off, could you give us an overview of the financial solutions you provide, particularly those tailored to underserved and financially excluded communities?
At Letshego, we offer loans—mainly microloans—to a wide range of customers, including small business owners and salaried individuals like civil servants or employees in various organisations. In addition, we offer digital loan products specifically targeted at underserved communities, especially in rural areas and semi-urban parts of Uganda.
Our main goal at Letshego Uganda is to serve the underserved, especially individuals who have historically been excluded from the formal financial banking system. Financial inclusion is at the heart of what we do and we want to ensure everyone has a fair chance at accessing credit, no matter where they live or what their background is.
You’ve spoken about serving the underserved—often referred to as the ‘last mile,’ especially in fintech circles. How does mobile-based credit scoring help you assess and serve individuals who lack a formal financial track record?
You’ll realise that the majority of Ugandans still don’t have access to formal financial services. While many actively use mobile money, they leave behind very few formal financial records. So, when it comes to credit, most financial institutions and banks face a real challenge in accurately assessing the financial position and risk level of such customers.
That’s where gnuGrid comes in—it helps bridge that gap by offering alternative credit scoring data drawn from various mobile network operators (MNOs). By analysing how people use their mobile phones and mobile money services—including the loans they access through different mobile money platforms—gnuGrid provides valuable insights. This alternative data helps financial institutions like Letshego assess customer risk more accurately and make informed lending decisions for individuals who were previously excluded from the formal financial system.
Help us paint a picture—what was the process like before gnuGrid came in? How were people applying for loans, and what data or systems were you relying on to assess their creditworthiness before introducing this technology?
Before gnuGrid, we were mostly relying on data from customers we had previously lent to, as well as Credit Reference Bureau (CRB) reports. However, that information was limited—it excluded a large number of potential borrowers we would have liked to serve.
That’s why we started looking for a credit scoring partner. GnuGrid brought in both the expertise and the technology to unlock new data sources. With their support, we’ve been able to assess risk more accurately and extend credit to people we previously couldn’t serve.
Since then, we’ve seen a noticeable increase in the number of customers we’re able to reach, and we’ve improved our ability to lend responsibly. We’re now giving credit to people who are actually capable of repaying, because we’re able to rank and segment risk better than we could before we had gnuGrid on board.
So, as you’ve explained, gnuGrid has helped you assess risk more accurately. But before that, did you experience losses or deal with non-performing loans due to limited data or reliance on subjective decision-making? Were there cases where a lack of credit history led to poor lending outcomes?
Yes, that’s for sure. When you don’t have a proper credit scoring system in place, you’re forced to rely on individuals to assess and advise on lending decisions—and that often comes with bias. Sometimes, it’s a staff member recommending a friend or relative for a loan, and you’re depending on word of mouth rather than data. Even at the managerial level, decisions were often made subjectively, which increased the risk of lending to the wrong customers.
But with technology like gnuGrid’s credit scoring engine, it changes everything. It allows us to assess risk more objectively and accurately. I can confidently say that many of the loans we gave out in the past—especially those based on individual judgment—might not have been approved if we had used credit scoring at the time.
Since partnering with gnuGrid, we’ve seen our non-performing loans drop from about 15% to below 10%. It’s been close to a year now, and the trend is very encouraging. The more we rank customers using this engine and integrate it with other data sources, the more we reduce our credit exposure. We’re targeting a drop to below 5% by the end of this year.
Can you speak to the impact of credit scoring on your loan approval rates? How have those figures changed since integrating gnuGrid’s solution?
We’ve seen a massive improvement. Our loan approval rates have increased by over 300% since we began using gnuGrid’s credit scoring, which relies on alternative mobile data.
Credit scoring technology gives us access to a much larger customer base in a very short period, without the need for human agents to physically assess each applicant. The system can instantly analyse millions of potential borrowers and identify those who qualify for credit. As a result, we’ve seen a significant rise in both loan applications and approvals within a relatively short timeframe.
So far, we’ve heard a lot about the benefits to Letshego as a lender. But are customers also seeing immediate gains? Given that loan pricing is based on risk, and now that your risk exposure is reducing, are you passing some of those benefits—like lower interest rates—on to the clients? Ultimately, how is this contributing to broader financial inclusion and access for the entire stakeholder ecosystem?
We’re currently adopting a model called risk-based pricing. What this means is that we determine the interest rate based on the customer’s risk profile. In simple terms, the better your repayment history, the more favorable your loan terms are. Customers who consistently meet their loan obligations will begin to enjoy lower interest rates than those who don’t.
This approach directly benefits the customer by bringing down the cost of credit, something that has historically been a major barrier to borrowing for many people. As more customers repay on time, our non-performing loans (NPLs) also go down, which in turn allows us to pass even more benefits to the borrower. So it’s a win-win: responsible customers get cheaper loans, and as a financial institution, we manage risk more effectively.
Beyond the drop in non-performing loans, have you been able to establish a clear correlation between credit scores and repayment behaviour? For instance, do you have data showing that a customer with a certain score has a specific likelihood of repayment?
Yes. Every customer gets a score and a risk ranking. For example, if you lend to ‘Moses’, the system might indicate that he has an 80% likelihood of repayment. So, chances are that ‘Moses’ with a repayment likelihood of 80% is actually a good payer, and the reverse is true for Customer B of 40%.
Inside Uganda’s Auditor General’s Report: The Good, the Bad and the UglyFrom our experience, these scores have been quite accurate. The rankings we’ve received have consistently reflected actual repayment behaviour. As a lender, you can choose to extend credit even to customers with lower scores, but we encourage those individuals to pay back their loans on time, so they can steadily improve their scores.
The better their repayment behaviour, the higher their score becomes. This not only increases how much they can borrow over time, but also allows them to enjoy better loan terms, such as lower interest rates. So, it’s a continuous improvement cycle—borrow, repay on time, build your score, and unlock greater financial benefits.
Let’s talk about data sharing. How important is it for industry players to embrace shared credit scoring systems? And what role does collective data—like that from mobile usage or alternative sources—play in strengthening responsible lending across the financial sector?
Generally, data sharing is very helpful, especially if all players in the market are using a common credit scoring system. In the long run, it benefits everyone. When you have access to alternative data from different sources, you’re able to aggregate that information and make better lending decisions, compared to relying on just a few limited data points.
We believe all players in the financial sector should embrace credit scoring and leverage alternative data, like the mobile data provided through platforms such as gnuGrid. This would help reduce non-performing loans across the board and shape the future of lending in a more sustainable way.
It’s not just about Letshego. As an industry, we need to build a lending culture where customers understand that borrowing comes with the responsibility to repay. If people know their behaviour is being tracked through a credit score—and that repaying loans improves their standing—they’ll be more disciplined. In the end, that benefits all financial institutions, because we’re all operating in the same ecosystem. So, when one lender benefits from better behaviour, we all do.
Are you seeing interest or potential to replicate this success in other Letshego markets? Given that gnuGrid is likely to expand, do you believe their credit scoring model would work just as effectively in other countries where Letshego operates, like Botswana or elsewhere?
Letshego currently operates in 11 markets across Africa. In East Africa, we’re in Uganda, Kenya, Tanzania, and Rwanda. In West Africa, we operate in Ghana and Nigeria. And in the south, we’re present in Botswana, Namibia, Eswatini, and a few other countries.
If gnuGrid ever decided to expand into any of these markets, we would absolutely recommend them. We’ve seen firsthand the impact of their credit scoring solution here in Uganda, and we believe it would be easy for other Letshego markets to adopt it as well. We’re already a case study of how it works, and that experience can help drive adoption across the group.
Looking at credit behaviour in our communities—where many people have little or inconsistent credit histories—do you think Uganda’s challenges are unique? Or are these similar to what you see in other markets? How does Uganda compare, and can this kind of credit scoring solution work just as effectively elsewhere?
I wouldn’t say that Uganda’s credit behaviour is unique. These patterns actually cut across most of the markets we operate in. But it’s a process—people need time to get used to new systems and technologies. Once they understand that you don’t need physical collateral to access credit, that all you need is a good score, things will begin to shift.
As more people become aware that a strong credit score can give them access to credit—without the traditional hurdles like guarantors, collateral, or long assessment processes—they’ll start embracing it. It saves time, offers lower interest rates, and removes many of the barriers people have faced in the past.
The future is bright, but we as industry players must take the lead in sensitising our communities. Once people understand the benefits of credit scoring and how it works in their favour, they will adopt it naturally.
You’ve mentioned that credit scoring is helping bring down risk—has this translated into lower interest rates for customers? And as a result, are you seeing more people coming on board and happier customers now that they’re being assessed more fairly?
That’s exactly why we made the decision to adopt what we call risk-based pricing. With this model, interest rates are determined based on how the credit score ranks a customer’s risk level. If you’re ranked as low risk, you’ll receive the lowest interest rates. For those with higher risk scores, the rates won’t be excessively high, but they’ll be slightly higher than those for customers with stronger repayment histories.
This approach encourages responsible financial behaviour. Customers who consistently pay back their loans on time will earn better scores—and in turn, qualify for cheaper credit. Over time, this builds a culture of discipline and rewards those who borrow responsibly.
I can confidently say that in the near future, more customers with good scores will start enjoying significantly lower interest rates across all our products. We’ve also committed to digitising most of our offerings to make the entire customer journey more seamless and accessible.
Around September last year, you launched LetsGo PESA in partnership with Airtel Money and gnuGrid. How has the market uptake been so far? What have you observed in terms of customer adoption and response?
We launched LetsGo PESA about a year ago in partnership with Airtel Money and gnuGrid. The uptake so far has been encouraging, especially considering that this is a new product in the market. It was important for us to approach it carefully and deliberately, so we started with a controlled rollout, allowing us to closely monitor user behaviour, gather feedback, and understand how customers were interacting with the service.
Through this phased approach, we’ve picked up a number of valuable lessons. We’ve learned what’s working well and what areas need improvement, particularly around pricing, user experience, and accessibility. Based on these insights, we’ve been working on refining the product to ensure it delivers even more value, especially for customers who demonstrate responsible borrowing behaviour. For instance, one of the key enhancements we’re making is adopting a model that rewards good repayment history with more favourable terms, such as lower interest rates. This aligns with our broader commitment to risk-based pricing and financial inclusion.
We are now at a point where we feel confident about scaling the product further. In fact, within the next month—or even sooner—we plan to roll out the enhanced version of LetsGo PESA at full scale. Together with our partners at gnuGrid and Airtel Money, we’re implementing all the necessary upgrades and product improvements so that it not only reaches more people but does so in a way that is cost-effective and impactful.
What customers should look forward to is a significantly better product—smarter, more affordable, and fully digital—designed to meet their needs with less friction and more flexibility. We believe this next phase will not only deepen adoption but also reinforce our mission of driving inclusive finance across Uganda.
