Consumer lending is a key lever for boosting economic activity in Nigeria. Yet, with retail lending still accounting for less than 10% of GDP, the sector remains largely untapped. In response, innovators are stepping up with bold strategies to close the credit gap, and leading that charge is Carbon, one of Nigeria’s pioneering digital lending platforms.
Founded in 2012 by brothers Chijioke and Ngozi Dozie, Carbon has evolved into a household name with standout offerings like Carbon Zero, one of the country’s most widely used Buy Now, Pay Later (BNPL) services.
In this exclusive conversation with BusinessDay’s David Olujinmi, the two Dozie brothers unpack the journey so far, offering sharp insights into the state of consumer credit in Nigeria. They speak candidly about the sector’s toughest challenges, the potential of AI and machine learning, and how Carbon is building a smarter, more responsible credit system for the future.
Let’s start with how Carbon began?
Chijioke: The concept of Carbon as it is now started in 2012. At the time, we looked at the market and saw that consumer credit was almost unavailable in Nigeria. At the time, retail credit to GDP was about 6 percent. Meanwhile, in other countries like Kenya and South Africa, it was much higher at 30 percent.
Back then, it was nearly impossible to access a loan from a Nigerian bank. In contrast, my experience abroad was completely different; I was issued a credit card at 18 and secured a mortgage by the age of 26, even as a non-citizen. That stark difference really highlighted for us that something fundamental needed to change in Nigeria’s financial system. We knew that if the future we envisioned, in which everyone could access credit, was going to become a reality, we had to act immediately. That’s why we started.
So, who came up with the idea first? Or was it more of a moment where you both just sat down one day and said, “Let’s do this together”?
Chijioke: In my previous role in private equity, I came across a company called Bayport that left a lasting impression. They offered payroll loans to mine workers in Zambia and Ghana. During due diligence, I visited some mining sites and saw men with axes collecting loans as small as $120; despite having steady incomes, they couldn’t get loans from traditional banks like Standard Chartered.
That experience opened my eyes to the credit gap for people who are technically bankable but excluded. What also stood out was that a group of 30-year-old South Africans had set up the business in Zambia and Ghana back in 2003–2004, giving loans to teachers and mine workers. I’ve always been inspired by that kind of bold, practical thinking, building real solutions in underserved markets.
“We knew that if the future we envisioned, in which everyone could access credit, was going to become a reality, we had to act immediately.”
It’s been nearly 13 years since this journey began. Looking at where things stand today, how would you assess the current state of the consumer lending space in Nigeria?
Ngozi: Honestly, we’re still at ground zero. There’s been a lot of momentum. Yes, we pioneered mobile lending, and many other players have come in since then. So, penetration is better than it used to be. But when you compare it to more developed markets, it’s clear how far we still have to go.
In those markets, most people have access to credit cards, they can finance short-term purchases at reasonable rates, and young people can get mortgages to buy homes. That kind of financial infrastructure just doesn’t exist here in Nigeria, at least not at scale.
And it’s not because there’s no demand. If someone walked up to the average Nigerian and said, ‘Hey, I’ll give you a 30-year mortgage at 16 percent interest,’ they’d jump at it. People want access to credit; they just don’t have the options.
Some consumer lending platforms might push back on your assessment. What do you think is still that missing factor?
Chijioke: The biggest gap in the market is still real, transparent credit scoring. A lot of lenders, even though licensed, use misleading tactics. You’ll see ads like, ‘Borrow N100,000, repay N101,000 in 90 days,’ but most people never get those terms.
Many of these platforms aren’t built with the customer in mind. That’s where we’re different. Carbon has always been mission-driven. I’ve shared before, at 26, living in South Africa on a Nigerian passport, I got a mortgage. I want young Nigerians to have access like that.
That’s why we offer free credit reports in our app. Even if we can’t give you N5 million, a bank like GTBank could use that history to offer you something. It’s about building visibility and trust.
Sadly, most lenders don’t report good behaviour to credit bureaus, only defaults. They’re not trying to build long-term value. Some even harass customers, sending messages to contacts or WhatsApp groups. Their model is simple: lend high, collect aggressively, and profit even if only half repay.
That’s not the kind of ecosystem we want to be part of.
Since Carbon acquired Vella Finance, how has that integration gone so far?
Ngozi: So first, it’s important to clarify that Carbon Microfinance Bank didn’t acquire Vella Finance; rather, the shareholders of Carbon did. That distinction matters because it means Vella operates separately from the bank itself.
That said, the Vella team has played a big role in building out our business banking product. Before that, Carbon was mainly focused on retail customers. We did have some SME lending, but we weren’t offering a full suite, no SME deposits or transactional banking.
Vella came in with both a capable team and a ready-made platform, which really helped us level up. Two years ago, all we did for SMEs was disburse loans into their existing bank accounts. But with Vella’s infrastructure, we were able to essentially replicate our retail experience for SMEs. Now we can offer lending, deposits, payments, everything. It fast-tracked our entry into full-scale SME business banking.
How exactly does AI show up in your everyday experience for your customers?
Ngozi: AI is still evolving, and it’s not perfect. That’s why we test everything internally before rolling it out to customers; we want to make sure it works.
Right now, we’re running about five AI projects. Some are straightforward, like generating ad imagery instead of hiring models. Others are more advanced, like our debt negotiation bot, which personalises offers based on customer behaviour and has helped improve recoveries.
We also have a fraud management bot that monitors real-time emails from other banks. It verifies transactions, checks customer status, and alerts our team, cutting out a lot of manual work. We’re aiming to automate even more soon.
We’ve even tested an AI voice bot for customer support. It had all the right data, but it struggled with real-world challenges, like background noise and call quality in Nigeria. It showed us just how important local context is.
So, our approach remains: test, learn, and refine before going public. That’s just good business.
Currently, do you have any AI models actively assessing borrowers’ creditworthiness or helping to determine their interest rates?
Ngozi: We’ve been using machine learning from the very beginning—it’s kind of the godfather of AI, if you will. That’s why when someone applies for a loan, we can decide in five seconds; we already have a good sense of whether they’re likely to repay or not.
Machine learning is a core part of AI, so in that sense, we’ve been working with AI-related tools since around 2016. It’s not new to us. And yes, we’ve used it not just for assessing credit risk but also for detecting fraud. We can flag suspicious behaviour based on repayment patterns, like when something just doesn’t look right. So really, machine learning has been a fundamental part of our tech stack from day one.
Do you use your own proprietary credit scoring model, or do you rely on credit bureaus or partner banks?
Ngozi: We do use credit bureaus; it’s a legal requirement, but they mainly tell us if someone has outstanding debt or is in default.
Our real decision-making, though, relies on our own proprietary credit scoring models. These look at your past repayment behaviour, and we also pull in external data, like whether you’re on LinkedIn, how long you’ve been active, and if that info matches what you shared in your application.
We’ve been refining this model since 2016, and we update it every quarter to reflect changes in the economy. Our data science team runs multiple models at the same time because new customers behave differently from those who’ve borrowed a few times or even long-time users who’ve been with us for years.
How do you balance speed and automation in your credit decisions?
Ngozi: We don’t think speed and fairness have to be trade-offs; we aim to do both. Our system is fast, but we also assess whether a customer can truly afford the loan. If we think repayment will be a burden, we’d rather reject the application than charge a high-risk interest rate.
We also offer longer-term loans to ease repayment pressure. And as customers build a good track record, we reward that with better terms, larger amounts, longer durations, and lower interest rates.
Even then, we stay within reasonable limits. Someone might be a great borrower, but if they can only handle N100,000 responsibly, we won’t push beyond that. It’s all about keeping lending sustainable and fair.
The bank of the future. How is that vision playing out so far?
Ngozi: The first is about independence. Right now, a lot of our banking infrastructure still depends on third-party providers, Amazon, Google, and cloud services abroad. So, if there’s downtime in the U.S. or Europe, it can shut things down here. We’ve felt that pain. And we’ve seen banks like GTB and Sterling also working to localise their infrastructure. So, owning more of that stack is a key part of the future.
But the more important part, for me, is that the bank of the future should be almost invisible to the customer. People don’t wake up wanting a bank account; they want to book a flight, pay for healthcare, and insure their shop. Banking is just the tool that helps them do that. So, I think the future is about deeply understanding the customer, anticipating their needs, and being quietly present in the background when needed.
That’s where AI can really play a role. With the right data, handled responsibly, AI can help us build a much more contextual, real-time picture of each individual. That allows us to offer products that fit their unique situation, because no two customers are truly the same, even if they look similar on the surface.
Chijioke: The last part of it, for me, is how we define a business. Ten or twenty years ago, when we said ‘SME,’ we pictured a small shop with a physical location. But today, if you took an Uber to get here, that driver is an SME. That’s their business. They’re servicing a car, feeding a family, maybe paying back a loan.
Or think about the person selling on Instagram, laptops, clothes, phones. They might not have a shopfront, but they have customers, cash flow, and ambitions. The traditional banking system doesn’t recognise them. You walk into a bank and say, ‘I need N50,000 to buy fuel for my Uber,’ and they’ll say, ‘We only do N10 million loans for equipment.’ That’s a disconnect.
At Carbon, we try to meet people where they are. We recognise that today’s entrepreneurs might be on Instagram or doing logistics or dropshipping from home, and they still deserve to be treated like a business. They need capital, support, and respect. That’s what the bank of the future should deliver: access, relevance, and dignity.
How many customers does Carbon serve today, and how far have you gone in terms of credit access?
Chijioke: We have over 5 million registered customers, but our active base is smaller than that. Considering we must stay profitable, we have to say no to a lot of loan requests because not everyone is creditworthy. A lot of customers come to Carbon initially for loans, and if they don’t get the amount or terms they want, some of them leave.
Beyond lending, we’ve built out a range of products and services that are compelling. We have accounts, debit cards, and Buy Now, Pay Later (BNPL).
