Why African Banks Reject Borrowers They Don’t Understand with Dalumuzi Mhlanga

[00:00:00] [SPEAKER_00] You're listening to the Unlocking Africa podcast.

[00:00:03] [SPEAKER_02] How businesses operate and how individuals generate incomes, they don't quite neatly into the requirements from a credit risk assessment perspective for lenders. Ultimately, what we are looking for is willingness and ability to meet obligations. Any kind of obligation that has to be met within a specified time period or within a deadline is a very helpful signal.

[00:00:28] [SPEAKER_02] The reason we are a credit bureau is we want to be able to provide a single source of truth, a global view on the credit worthiness and affordability of individuals and small businesses.

[00:00:40] [SPEAKER_00] Stay tuned as we bring you inspiring people who are unlocking Africa's economic potential. You're listening to the Unlocking Africa podcast with your host, Terser Adamu.

[00:00:55] [SPEAKER_01] Welcome to the Unlocking Africa podcast where we explore the ideas, innovations and strategies unlocking Africa's economic potential. Today, we are asking why so many Africans remain unable to access credit when financial institutions have money available to lend.

[00:01:12] [SPEAKER_01] My guest today is Dalu Buzin-Sara, founder and CEO of Notto, an alternative credit bureau working with lenders to understand financial behavior that conventional credit records often miss. Dalu, welcome, welcome, welcome to the Unlocking Africa podcast. How are you? Hey, thank you so much, Terser. Really, really great to be on Unlocking Africa. I'm doing all good. How are you? I'm very well, thank you. And it's a pleasure to have you on the podcast.

[00:01:42] [SPEAKER_01] I know we've had this in the plans for quite a while, so it's great that the time has finally arrived. Absolutely, absolutely. I've been counting down. Brilliant, brilliant, brilliant. So as always, I like to start from the beginning. So I was hoping you could give us a introduction into who Dalu is. Sure.

[00:02:02] [SPEAKER_02] So my name is Dalu Buzin-Sara, but I also go by Dalu. I'm founder and CEO of Notto, as you've mentioned. Bored and bred in Zim, but then I've worked and lived across, you know, countries, continents, the US, the UK, but at the moment mostly based in South Africa and building our business with the team from there.

[00:02:26] [SPEAKER_01] Fantastic. As you said, you're based in South Africa. So I thought, what a great place to start the conversation in terms of the work that you do in terms of credit, lending, banking, South Africa. So I read one of the reports which mentioned something about South African banks hold capital and liquidity above, say, their regulatory requirements. But many consumers who actually applied for credit were declined.

[00:02:56] [SPEAKER_01] What would you say explains the gap between that capacity to lend and also the willingness to approve borrowers?

[00:03:04] [SPEAKER_02] Yeah, to your point. We started off in South Africa because the reality is that South Africa has got one of the most developed and most sophisticated credit markets on the continent. And so there's a great deal to learn from that. So we learned a lot in the process and we are licensed and regulated as a credit bureau there.

[00:03:25] [SPEAKER_02] The gap that exists sort of like between just the liquidity that's available and its limited flow to the majority of people, there is sort of a couple of things to explain. One is the informal economy, which is where a lot of the opportunity is.

[00:03:45] [SPEAKER_02] How businesses operate and how individuals generate incomes, they don't quite fit neatly into the requirements from a credit risk assessment perspective for lenders. And there's a couple of reasons for that.

[00:04:02] [SPEAKER_02] One is historically the largest opportunity for lenders has been in the formal sector, you know, formerly employed individuals, formerly registered SMEs, larger businesses, etc. And that opportunity is big in and of itself, right?

[00:04:20] [SPEAKER_02] Now for a lender to make a very clear and decisive move towards lending to the informal sector as it were, requires that there be a very clear opportunity that moves the needle. Move the needle in terms of, you know, the additional revenue, the profitability of that opportunity, etc.

[00:04:41] [SPEAKER_02] And until now, the opportunity hasn't sufficiently moved the needle in the sense that how do they make sure that they are fully capturing the opportunity and, you know, it makes sense, like, you know, that this will drive 10%, 20% growth. So I think that's the first bit.

[00:04:58] [SPEAKER_02] The second is that, you know, these are large regulated institutions and they have sort of governance requirements, compliance requirements, and some of that constrains the ability to really fully aggressively take this opportunity out. But then the third is that until recently, you know, with digital payments, etc., it's actually been quite difficult to take those credit risk assessments to double check affordability, revenue, etc. But now we have more digital payments that are available.

[00:05:28] [SPEAKER_02] And so now you really can get the data to be able to undertake those credit risk assessments. So I think it's those three things, really, which is until recently, the opportunity hasn't been sufficient to really move the needle so that lenders can pay attention to it. Just like the institutional design and institutional inertia as well, because they are regulated entities. And the third is just the availability of the data to really take credit risk assessments. So I think that's why there's that gap.

[00:05:56] [SPEAKER_01] So you mentioned that when it comes to the credit risk or the assessment that a lot of people don't fit nicely into the credit risk assessments of lenders. So when a lender, say, declines someone because it cannot confidently assess them, how often is that mistaken for evidence that the person is not able to repay?

[00:06:16] [SPEAKER_02] Yeah, I think it happens quite often. Historically, a lot of these transactions have been through cash transactions for people generating their income and being paid in cash. And so that doesn't leave a sufficient digital trail to really validate the flows of income as well as the outflows from an expense perspective.

[00:06:38] [SPEAKER_02] But as I've mentioned, you know, in the last 10, 15 years, there is more and more of what had historically been cash transactions that are being captured digitally. So a lot of those declines really are because the lenders will be dealing with what in our space is called theme file crimes, because it's just not enough evidence as it were.

[00:07:00] [SPEAKER_02] So that evidence is growing because of the increasingly, you know, ubiquitous digital payment platforms and increasing adoption of those. The second is, you know, sometimes a lender will ask for collateral as part of the structure of that credit transaction. But a lot of small businesses, you know, and individuals simply don't have the collateral.

[00:07:23] [SPEAKER_02] Although you can see just from the transaction flows that there is capacity to take up credit and that there is a level of sort of willingness and ability to repay that obligation. But, you know, if a bank or lender has got a very strict credit policy as it relates to collateral, that really prows out a very big part of the market.

[00:07:50] [SPEAKER_02] And then I guess the third is at times, you know, the credit products that are out there aren't, you know, even though somebody might check the various boxes in terms of, you know, requirements and maybe they have collateral. But then maybe there's anality in those transaction flows or, you know, there's just some other kinds of complexity or nuance in the numbers. And the credit products that exist aren't designed to account for some of those nuances.

[00:08:19] [SPEAKER_02] Then that one will see in the informal sector and then, you know, a customer gets declined. So those are sort of like a range of reasons that we see, all of which can be addressed through the alternative credit intelligence solutions that we have built.

[00:08:35] [SPEAKER_01] So you touched on cash payments and also the evidence. So for a small business operating largely in cash, what evidence of financial reliability exists that a credit file doesn't have available?

[00:08:49] [SPEAKER_02] Yeah. So with that, the reality is that for ourselves to do our job as well as we can, we do need digital trail off payments. And so as I've mentioned, quite a number of people are switching from cash or just, you know, increasing the portion of what they earn, you know, moving that away from cash to some like digital payments.

[00:09:11] [SPEAKER_02] So that's the start as it were. But even then, you know, a traditional credit file is going to, depending on the market and their sophistication, is going to either just have two kinds of reporting for credit bureaus, positive reporting and negative reporting.

[00:09:28] [SPEAKER_02] In some markets, there will only be sort of like negative reporting, which is to say a lender pings the credit bureau and they just get a response to say, this person is good to go, not overextended from a debt perspective and, you know, doesn't have overdue accounts. But from a positive reporting perspective, there are bureaus that will sort of like provide information on all the various accounts, how well they're being paid, et cetera.

[00:09:54] [SPEAKER_02] Whether it's positive reporting or negative reporting, the traditional bureaus really largely only report on credit repayment. So an individual or small business first has to access credit for there to be any information on their credit report or credit file. Now, that by definition excludes individuals or small businesses that don't have access to formal credit in the first place.

[00:10:23] [SPEAKER_02] And that's why alternative data is super valuable. And we get that from telcos, mobile money operators, payments, fintechs. So it's like the very core of what we provide, which is a view of, you know, transactions, inflows, outflows, their predictability, you know, whether somebody is meeting their rental obligations in full and on time, utilities, you know, paying supplies, et cetera. That wouldn't necessarily show up in the credit file.

[00:10:51] [SPEAKER_02] And this is what makes it unique because we can then get people onto the credit ladder using our alternative system.

[00:10:58] [SPEAKER_01] That raises the question in terms of what lenders are missing. So if you look at the financial behaviors that exist, I know you just mentioned that you examine signals such as rent, utility payments, mobile money activity. So which behaviors would you say have proved most useful in predicting repayment of clients, customers?

[00:11:20] [SPEAKER_02] Sure. There's two sides we look at. One is the inflow side just to, you know, assess the regularity of income as well as, you know, the capacity for, you know, taking up credit from an inflow perspective. And then from an inflow perspective, we ultimately what we are looking for is willingness and ability to meet obligations.

[00:11:45] [SPEAKER_02] Any kind of obligation that has to be met within a specified time period or within a deadline is a very helpful signal. So, you know, we will score for both individuals and small businesses, you know, for individuals, are they paying rent in full and on time? Are they paying, you know, the kids' school fees, utilities? And then you just look at other expenses and, you know, how regular that expenditure is and, you know, the timing of when it's made.

[00:12:15] [SPEAKER_02] With small businesses, we'll look at, you know, do you pay yourself or your employees in full and on time, you know, suppliers? And the various sort of like operational expenditure that goes with running a business. And with that, we get at least an initial view and an initial sense to some level of accuracy in terms of willingness and ability to repay.

[00:12:37] [SPEAKER_02] Of course, there's a lot of technicality around it in terms of getting to a level of confidence that also a lender would also support. And then, as I say, it gets super technical. And once they're comfortable with the process we've gone through and its output, then they're able to, you know, take up the opportunity to make a move and start lending to the base of eligible customers that we would have surfaced for them.

[00:13:02] [SPEAKER_01] And in order to do this, there's a huge amount of data. I think I read that you've analyzed more than a billion transaction records. So how do you turn that volume of data into, say, a decision a bank's credit team can understand and defend or make a decision on?

[00:13:20] [SPEAKER_02] Yeah. So a couple of things. The first is credit risk teams, a lot of lenders build their own scoring models. To the extent that we can show them that we have followed the same process that they ordinarily do for building traditional scoring models. It gives them comfort. What we have is above board.

[00:13:47] [SPEAKER_02] So, for example, you know, one of the things that we make sure we demonstrate is that the five C's of credit. And, you know, that's what lenders want to look at. And those are character, capacity, capital, collateral conditions in terms of macroeconomic conditions. And we demonstrate how the traditional version of that is covered in our alternative models. So I think that's the first thing.

[00:14:12] [SPEAKER_02] Lenders don't think we are coming up with something that we can't explain or which is too far off what they are used to. So that's the first bit from a process perspective. And then once we've demonstrated that, of course, we will show them, you know, how we process the data, you know, the features that we develop, how we build the models, validate them, test them, et cetera.

[00:14:36] [SPEAKER_02] So from that process, once we build comfort with that, the next bit then is delivery generating the proof points from a predictive accuracy perspective. So something called area under the curve, which is sort of like a key metric, build comfort with that. And then when we start off with the small governed pilot and then with that, we track default rates, NPLs and various sort of like quality metrics.

[00:15:05] [SPEAKER_02] And when we move from sort of like a process that they're comfortable with, then we can then figure out, you know, how to scale and roll this out and go to market in an aggressive way. So it's really working very closely hand in hand with the lenders, the credit risk teams, and then being able to get in under the hood and have, you know, models that we can explain that regulators can order.

[00:15:31] [SPEAKER_02] And those become the foundations for us to be able to really operate at scale and make the most of the opportunity.

[00:15:38] [SPEAKER_01] Thank you for sharing that. So from the information I've read, you described the decision gap as a central problem. So at what point would you say useful data fails to become a lending decision?

[00:15:52] [SPEAKER_02] There's the secret that you've mentioned. There's useful data and then there's a lending decision that needs to be made. You know, at times what we find is there are other risks that sort of need to be taken into account. Like the useful data is not enough because scoring is one bit of the sort of like broader process, right? The other bits, you know, we have to think about the portfolio management.

[00:16:21] [SPEAKER_02] You know, once you have the useful data, you've got this book. How you manage this portfolio is quite different from how you manage a traditional book. You have to be thinking about early warning signals with that. You have to think about collections, you know, if things start to go left. So we get to a point where the data is useful. We have scored customers with an eligible base.

[00:16:46] [SPEAKER_02] Then now it's like, okay, how do you actually execute on this in terms of, you know, how you will disperse, how you will manage the portfolio, how you will collect. And then other elements to think about, you know, you have to rework the credit strategy and credit policy. So we've worked with our lenders where we've ticked all the various boxes from a process perspective. We've got a pilot that's worked, et cetera.

[00:17:12] [SPEAKER_02] Then now you have to go back and rework the credit policy so that the credit policy supports what we would have seen and uncovered in terms of the opportunity. So it gets to a point where, like, that sounds like a fair bit of work. And it is. And then the lender has to then make the decision to say, given the return that we're likely to see here, you know, what's the operational capacity and appetite to really double down on this.

[00:17:39] [SPEAKER_02] I think we find some lenders are very happy to move quickly and run with it. Others might be a bit slower. But, you know, I think this is an inevitable space that we are entering. And then the ones who have moved quicker hopefully will make the most of that first move advantage.

[00:17:55] [SPEAKER_01] The potential is clear. And from what you've just explained, recognizing useful data and putting it to work across, say, a bank are very different things. So many banks already know alternative data exists. What would you say prevents them from using it in their everyday lending decisions?

[00:18:18] [SPEAKER_02] I think one of the big things I think that I alluded to earlier, maybe I can unpack a little bit more. Is the opportunity to pursue that comes with alternative data, is it large enough for them to really double down on? And that's just like one thing for them to consider, given the risks that come with it.

[00:18:39] [SPEAKER_02] And so at times the opportunity is there, but it comes at an operational cost and a level of risk taking that the lender just might not be comfortable with at that particular time. Right. Maybe down the line, but at the time, maybe not. And remember, these are regulated entities as well. So they have to think about the compliance side of things. And, you know, if they go down this route, what additional regulation compliance burden are they taking on by going down this route?

[00:19:09] [SPEAKER_02] And so one has to think about that. And we also have to think about how regulators and policy can create frameworks that make a decision for lenders to go down this path a little better, as it were. And I think those are some of the key things, really, which is, is this big and exciting enough, given the risks that come with it? And are we can do this without overextending ourselves from a regulatory compliance perspective?

[00:19:34] [SPEAKER_02] I think that's what slows down, you know, some lenders, even though they see that alternative data is available and there's a real opportunity in London.

[00:19:43] [SPEAKER_01] So would you say that it makes more sense for, say, a bank to partner with the likes of Notto than to actually adopt this alternative lending data framework?

[00:19:54] [SPEAKER_02] If you think about it, I think, from my view, it's usually helpful for different players to really focus on their core business, their core strengths and, you know, their daily bread and butter, as it were. And I think we've been in the game long enough and really understood the data, the opportunity across multiple markets.

[00:20:18] [SPEAKER_02] There's a lot of learning, differentiated insight and really capacity and expertise that we bring to the table, which might take, you know, lenders a little bit more time to get. I mean, like on our team, we have engineers who take off the infrastructure side of things, you know, the various work integrations. And a lot of them have worked in some of the biggest banks across the continent.

[00:20:43] [SPEAKER_02] On our credit decision science team, the team that does the models, we have, you know, act series who work hand-in-hand with data scientists to build the models. And they have also worked in some of the largest banks on the continent. And they know what, you know, a credit risk team is looking for, what a chief risk officer wants to see in a model before, you know, we can really fully deploy.

[00:21:08] [SPEAKER_02] So there's a level of focus that we bring from that perspective because it's our bread and butter. But, you know, we bring what a bank would want to see even in their own teams and more because of sort of our breadth and depth of experience.

[00:21:25] [SPEAKER_01] So what would you say a typical, say, relationship or partnership between Notto and an established financial institution looks like?

[00:21:35] [SPEAKER_02] Yeah. What we do essentially is, in the most ideal sense, we are an end-to-end partner for their alternative credit strategy and portfolio. And that means we source the data from alternative data providers. And, you know, we figure out the commercials that work for the lender ourselves and the alternative data provider or providers.

[00:22:02] [SPEAKER_02] And after sourcing the data, we build the infrastructure within both the alternative data provider and the lender so that data can flow consistently. Because what we do is we don't just build credit scores once off. We literally refresh those credit scores on a daily basis so that there is no stale data that has been used to make a lending decision.

[00:22:29] [SPEAKER_02] And we then, after having sourced the data, built the infrastructure, built the credit scores, we actually actively support the management of that loan portfolio. You know, we're looking at both the just like general business metrics, which is how much has been dispersed, how much has been repaid, what's the uptake within eligible base?

[00:22:54] [SPEAKER_02] You know, how are consumers, small businesses interacting with the product and using the product? And how do we refine that? We look at the credit quality metrics and those are reported on, even on a daily basis, right? Just depending on what our partner pretends. We can report on it weekly. We surface, you know, early warning signals on an active basis so you don't have to wait until a book has deteriorated before you know what needs to be done. And finally, we have an AI agent.

[00:23:22] [SPEAKER_02] We call it an auto agent that is just working in the background and surfacing recommendations, et cetera. So that's an ideal relationship really is perfect. An end-to-end, deeply embedded partnership can provide bespoke tailored services as well. That's the ideal kind of partnership that we have with some of our customers.

[00:23:42] [SPEAKER_01] And you are a licensed and regulated alternative credit bureau. So how does that status change the conversations you can have with established financial institutions?

[00:23:54] [SPEAKER_02] Well, I think for one, we are held to an incredibly high standard. So we are licensed in South Africa by the National Credit Regulator. And there's incredibly high compliance standards that they have for credit bureaus. We have to go through audits every year. There's reporting that we have to do.

[00:24:14] [SPEAKER_02] So what that does is it gives a certain level of assurance in terms of our systems and processes as it relates to accuracy of the data that we use, data integrity in general, data protection, cybersecurity, making sure that the privacy, data privacy and the consumer is protected and that we're only operating or rather building our models using consented data.

[00:24:41] [SPEAKER_02] So we're held to that very, very high standard. Now, we're not licensed every market. But because of the Rigo with which we have built our platform initially in South Africa, that means every other market in which we partner or enter or have customers, we bring that same level of Rigo and the highest levels of compliance standards.

[00:25:06] [SPEAKER_02] Even when we don't have a regulatory body looking out for us in the market where we're not licensed. But it's just in our DNA to be compliant and to hold those high standards. And all of that flows into the performance of our models. And, you know, ultimately what really matters is NPLs. Not ultimately, but like a big part of what matters is NPLs. And our NPLs have consistently been well below 4%. And yes, our models do great.

[00:25:33] [SPEAKER_02] But then another layer is, you know, this Rigo with which we've built the system and the standards which we hold ourselves to as we enter into partnerships.

[00:25:43] [SPEAKER_01] Thank you. So if we move from the models and the bank's internal processes to the results, I know you've supported more than $2 million in monthly loan scoring and disbursement activity. What have you learned from the real lending outcomes that you could not have learned from building the models alone?

[00:26:03] [SPEAKER_02] Interestingly, in just the past few months, we have grown that number of monthly dispersed by, you know, several multiples. So it's a much bigger number. And part of what drives that exponential growth is precisely what you pointed out, which is it's one thing to have, you know, really fancy scoring models. And it's another to drive performance.

[00:26:27] [SPEAKER_02] And I think some of the things to drive performance and really learn about, you know, what's working well and what's not working so well about the product. One of the things that has been fascinating has been for the eligible customers and those who continue to repay and do well, they really drive down NPLs tremendously. And interestingly, we have one product where, so let's say a customer has got a $100 limit.

[00:26:57] [SPEAKER_02] And then they first come and they take $50. And then before they've finished paying off the $50, they come and take up another $30. Those kinds of customers we have learned are at higher risk of, you know, driving NPLs. But then the customer who comes and takes the full $100, repays it, et cetera, drives sort of like lesser defaults than NPLs. You know, that's a learning that we have. And I guess maybe it's a question of, you know, managing limits.

[00:27:26] [SPEAKER_02] But it's a learning that, you know, now we can apply as we go. We knew we were building a very interesting product that, you know, customers will take up. We've seen the repeat borrower rates literally almost touch 90%. I think now it's about 87% of customers who keep coming back. And so it's a sticky product that customers love and are finding helpful. On the consumer side, you know, we have good activity.

[00:27:55] [SPEAKER_02] And another thing maybe is, you know, as we build those products, there are nuances in terms of the product construct itself. But then there are also nuances in terms of how you manage the portfolio. For example, how you define an NPL for a 30-day product is very different from how you define it from sort of like a general five-year product. And we need our lenders to understand that, regulators to understand that.

[00:28:21] [SPEAKER_02] So I think the process of actively managing the book is you'll just have surprising insights in terms of, you know, what drives uptake, what drives poor sort of like deteriorating credit performance. You know, what metrics you actually need to measure and, you know, have a more nuanced understanding of how you measure those.

[00:28:43] [SPEAKER_02] And every day, like, you know, we're learning something new just because of the scale at which we are operating and the nature at which we are aggressively growing the monthly disbursements on our platform.

[00:28:54] [SPEAKER_01] Fantastic. So if we take it a few steps back, I know that you've mentioned some acronyms such as NPL. So for people who are listening who aren't quite familiar with those terms, what is NPL? Oh, sure, sure.

[00:29:07] [SPEAKER_02] Sorry. I mean, this is, I live and breed credit. We're all guilty of doing it. In my mind, everybody knows what an NPL is. So NPL is a non-performing loan. And a key metric is the non-performing loan ratio, which is essentially, let me first define a non-performing loan. A non-performing loan is a loan, whether in full or in part or a portion of a loan, that is 91 days overdue.

[00:29:38] [SPEAKER_02] Right. So you had an obligation that was due 31st of March. Right. By now, you are 90 plus days. We're 90 plus days past the 31st of March. That now is a non-performing loan, which is the portion of it that hasn't been paid is non-performing. And it's sort of like a way of helping lenders manage their risk and say, theoretically, you're saying you're not going to be able to recover that amount. So therefore, it's non-performing.

[00:30:07] [SPEAKER_02] And then the non-performing loan ratio is you take the number or rather the amount of loans and that are a non-performing status that are 90 plus days overdue. And then you divide it by your outstanding principal. So you divide it by the capital that has been deployed. And that gives you an NPL ratio. So that's a definition. I can go into the nuances of complexities for how that looks like for 30-day products, 5-year products. But for now, I'll leave it at that.

[00:30:37] [SPEAKER_01] No problem. So would you say there are certain aspects or elements within the assessment process that could mean approving more borrowers or improving the NPL ratio?

[00:30:50] [SPEAKER_02] Yeah, yeah, absolutely. There's the actual assessment. And one of the things that maybe we don't talk a lot about is collections. So the ultimate NPL ratio is the function of identifying the good borrowers. But a fair bit of it is, you know, being actually able to collect for those who might have defaulted.

[00:31:12] [SPEAKER_02] I think it's important to get those and obviously collect ethically without some of the challenges we've seen in terms of, you know, unethical collections. You know, people calling parents, brothers, cousins, workmates to recover debts from someone who would have borrowed. What more we can do to sort of like attract more good borrowers and reduce NPLs? I really think it just comes from the experience of lending itself.

[00:31:42] [SPEAKER_02] One is the more you learn and the more you learn from borrowers, the book, the better you retrain. You actually go back and retrain the model on the back of that. So it's not like we have one model that's just like static. You know, we're always retraining our models. In fact, we set up what I call champion challenger models, right? So we build a model. It has a particular level of predictive accuracy. And then that becomes the champion model.

[00:32:10] [SPEAKER_02] But we are always running the challenger model in the background using the data that we are seeing. And if the challenger model starts performing better than the champion, it then becomes a champion model. We're not always just like sitting with one model. We always have the champion challenger running in the background. And then at intervals, we will actually retrain, you know, the models based on the real sort of like non-performance data that we would have gathered. I think that's one of the key things really.

[00:32:38] [SPEAKER_02] And I think another is use data on sort of like credit performance to inform the product construct itself. Because when we construct credit products, there are certain assumptions that we have, hypotheses that we make. And some of those are invalidated or validated once loans are actually in people's hands and small businesses' hands.

[00:32:59] [SPEAKER_02] And then we basically reiterate and say, OK, we're not just going to change the model, but we will actually rethink the construct of this product in terms of, you know, pricing, 10 years, the various terms that come with it, et cetera. So this iterative process basically gets us to a point where we build the model so that we, you know, we attract, we can better and better identify good borrowers, which in turn leads to NOAA NPLs as well.

[00:33:26] [SPEAKER_01] So if you look at the model in terms of iterating or changing the model, if that's even possible, I know that you operate in South Africa and Zimbabwe. What has expanding across those markets told you about, say, the limits of a single Pan-African credit model? Is that possible? Or do we need to adjust the model depending on the different markets that you operate in?

[00:33:49] [SPEAKER_02] Yeah, I think we absolutely have to change our model depending on the different markets because the structure of those markets is quite different. I mean, South Africa does have, you know, a fairly informal economic base. I'll give you some examples of nuances. So in South Africa, quite a number of people who in the segment that we support and build credit scores for receive social grants from the government.

[00:34:17] [SPEAKER_02] But, you know, they don't admit it, but over and above the social grants, they've got other sources of income. So we have to think about that dynamic. And when they receive social grants, they will receive a social grant, let's say, in one bank account that they've opened. And then they will receive their income from the entrepreneurial revenues in another account.

[00:34:42] [SPEAKER_02] But maybe that account is in somebody's name, you know, because they don't want the social grants department to notice that they're making, you know, more money elsewhere. So there's a new website you have to do, which is very South Africa specific because other markets don't have social grants. And then you go to Zerm where you've got two currencies, you know, you've got the U.S. donor and you've got the local currency that's called ZIG. You know, and people are sort of like transacting between the two currencies.

[00:35:11] [SPEAKER_02] But you want to have as complete a view of, you know, their transaction inflows, outflows, et cetera. And Zerm is much more informal than South Africa is. I'm actually in Zambia right now. And what things look like in Zambia are, you know, quite different from, you know, Zimbabwe and South Africa individually and combined.

[00:35:36] [SPEAKER_02] So absolutely, whenever we enter a new market, we literally sort of like start from as much of a tabula rasa point of departure as possible. Because we know the danger of just, you know, applying what we have learned elsewhere blindly. Because, you know, we really can, you know, get into introduce serious risks by doing that.

[00:36:05] [SPEAKER_01] So keeping to the theme of operating across multiple markets, I know Nato aims to enable more than 100 million credit scores across five African markets by 2030. So what infrastructure or institutional change would you say has to happen first for that ambition to be realized?

[00:36:25] [SPEAKER_02] Yeah, I think the single most important change that we need to see in order to really fully harness this opportunity is getting to a point where alternative data providers share data amongst themselves, right?

[00:36:42] [SPEAKER_02] So getting to a point where telecores, payments, syntax, some mobile money operators share data across, you know, with each other and across the boundaries that exist between them. Because today, each of these entities have got a lot of really rich data. But that data is limited to what they can see. Ultimately, the reason we are really a credit bureau is we want to be able to provide a single source of truth,

[00:37:11] [SPEAKER_02] a global view on the credit worthiness and affordability of individuals and small businesses. And that's really most effective when we're able to see data across a number of partners. So, one, we need to build the commercial models that make it make sense, or at least the commercial rationale that makes it make sense for alternative data providers to share data,

[00:37:37] [SPEAKER_02] you know, build up the protocols for that, and then enable this single source of truth. And we know there's precedent for this because this is what the bureaus do, right? Whenever one, whenever a consumer small business signs up for, particularly consumers, signs for credit agreements, they consent that the lender will send data to a credit bureau. And when the credit bureau gets that data, it basically aggregates the data across, you know, the various data providers.

[00:38:01] [SPEAKER_02] We will need that for alternative data providers as well to just really elevate what's possible. But, you know, we also need policymakers, regulators to build out the frameworks that enable this. That's from the data sharing itself, but as well, obviously, protecting the consumer. So, making sure that, you know, it's consented, there's consent processes, data privacy protection, and all of that.

[00:38:27] [SPEAKER_02] But if there's one thing that I will sort of like point and say, you know, this is what we need in the next, you know, four or five years, is getting to a point where we're able to share data amongst alternative data providers, and that we have the necessary regulatory frameworks to support that and protect the consumer, obviously.

[00:38:48] [SPEAKER_01] Fantastic. So, you kind of shared where you hope to see this alternative credit lending or data space in the next five years. What contribution do you hope Noto will be making to this space in the next five to ten years?

[00:39:03] [SPEAKER_02] What we have done as a licensed and regulated paper is we really are building the infrastructure and the intelligence layer that will support this ambition and this hope that we have for where our space is going to go. We have infrastructure that works, is robust, is reliable, operates scale.

[00:39:27] [SPEAKER_02] And from a credit intelligence perspective, we've got models that provide, you know, really high, you know, predictive accuracy to support, you know, low NPLs down the line. So, we see ourselves as a really core part of this from an infrastructure, technology, and intelligence perspective. As that data becomes available, it will flow to us as a credit bureau and as a scoring engine. And we will expose that to lenders.

[00:39:57] [SPEAKER_02] And as I said, you know, just provide that end-to-end, getting the data all the way to portfolio management, collections, et cetera. So, we really see our role as central to this effort from an infrastructure, technology, and intelligence perspective.

[00:40:12] [SPEAKER_01] Quote of the week. Is there, say, a quote, an affirmation, philosophy, or even an African proverb that has helped you in times are challenging or want times are good through this journey of building Noto?

[00:40:26] [SPEAKER_02] Absolutely. It is a quote by Frantz Fanon. And my team here has this. Every opportunity I have to say it, I share it. And if I'm walking along the streets, if anybody will care to listen, I will share. So, it's a quote by Frantz Fanon. It says,

[00:40:50] [SPEAKER_02] And I really believe that our generational mission is to drive Africa's economic prosperity. And credit is an incredibly key killer for enabling that. And when all is said and done, if our generation delivers that, we will have fulfilled our mission.

[00:41:12] [SPEAKER_01] I agree. Fantastic, Dalil. What a fitting way to close today's conversation. Thank you for joining me. And also, thank you for helping me understand what it would take to make more of Africa's economic activity visible to lenders. We know in the UK, Europe, wherever you live, sometimes we take credit and being able to access banking and funding for granted. But you're providing a service that is much needed on the continent.

[00:41:40] [SPEAKER_01] And so, well done to you. And I look forward to seeing how things progress and grow in the near future.

[00:41:46] [SPEAKER_02] Great. Thank you so much. I really appreciate it. Thank you again for having me. This was an incredibly wonderful conversation.

[00:41:54] [SPEAKER_01] Thank you to everyone who has listened and stayed tuned to the podcast. If you've enjoyed this episode, please subscribe, share or tell a friend about it. You can also rate, review us in Apple Podcasts or wherever you download your podcast. Thank you and see you next week for the Unlocking Africa podcast.