It’s mad to think two young Irish guys, school friends, would have the audacity to try to build a digital bank for the continent of Africa. 

It’s all the more astonishing to me, because I’ve known them since we were teenagers. Tiernan Kennedy and Barry O’Mahony are cofounders of Umba, a digital bank currently operating in Nigeria, and with multiple other markets in its sights. Kennedy serves as CEO, O’Mahony as CFO.

The scale of their project is mind-boggling. To pull it off, Umba needs an app as slick as Revolut’s. It has to build a credit scoring system from scratch using machine learning tools. It needs the discipline and processes of an established bank. It has to win over regulators in each market. It has to pick the right business model, and build it in the right order. It needs to raise millions from venture capital backers. It has to attract users. And, just as importantly, it has to convince them to deposit their money. All this it has to do with one foot in its San Francisco headquarters, and the other in its bases in Lagos and Nairobi.

Where would you even start?

The road to Lagos

How does one end up as CEO of an African FinTech? Kennedy and O’Mahony have worked together for years, on and off. After college, they started Krunksoft, a software business that built apps for A&L Goodbody and others. O’Mahony spent a few years playing professional rugby with Munster. Then came the idea for Umba.

“Doing something in Africa was a huge aim,” Kennedy told me. “I went there first in 2010, spent the summer there. And really wanted to go back and do something there. Because Ireland has a long connection to Africa. There’s a tonne of people who’ve worked out there for years. I remember, one of my neighbours used to be in Lagos every single month. 

“And then – long story short – I went to America and did a startup. I went there and joined a startup accelerator. I was representing Trinity, and our team was competing with Harvard and Cornell and all these big US universities. We came second, which was a huge result, but the visa system precluded us from actually running that business in the US. So I decided to stay on and got a job. And then, right away, it was focusing on how I’d get my green card. 

“And then the minute I got the green card, it was, ‘what do I do next?’ I had been working in machine learning. I was working in consumer home security in the US. I’d learned a lot. I had gone from individual engineer hacker person to, running a 30 person engineering team in New York. By that point, I felt had earned my stripes enough to build a company, because I’d been involved in board meetings, I’d been involved in certain parts of fundraising, I’d seen VC funding, and how it all works. So I had seen how companies are built in the US startup way. 

A lot of investors here [in San Francisco] love the big idea. Like crazy off the wall stuff – more than the New York investors.

Tiernan Kennedy

“And then I had gotten my green card, and really wanted to do something big and meaningful. As you know, I don’t have a banking background at all. But I had the engineering machine learning side of things. And then had that African experience that I’d seen [the advanced Kenyan payments system] M-Pesa in 2010.

“I saw first hand how M-Pesa was changing the payments landscape in Kenya in 2010, then Venmo taking over the US payments space a few years after, and then Revolut entering the European Market and getting massive market share with their full banking offering. 

“I was working in distributed systems and machine learning, and Barry [O’Mahony] was working in the Kenyan payments space, and seeing this growth first hand. There was clearly a gap for a digital bank in Africa, where digital payments had spread like wildfire but most people still didn’t have bank accounts. Since we launched services in January in Nigeria we’ve really seen that there is a massive appetite for this service- we’ve had over 100,000 downloads since we launched there in January.”

Pulled West

The Umba story is obviously about this big opportunity in Africa. But it’s also about the intense gravitational pull of the United States. A project this grandiose couldn’t happen from Ireland. It couldn’t even happen in New York City, where Umba was founded and got its early backing. 

Kennedy and O’Mahony decided Umba needed to be in the heart of things, in San Francisco. “A lot of investors here [in San Francisco] love the big idea,” Kennedy said. “Like crazy off the wall stuff – more than the New York investors. So we’re able to raise money. I was out here for three weeks meeting investors, and it was going really well. So we said ‘screw it, why not give it a go out here?'”

So even though Umba’s customers and regulators are to the east, in Lagos, Nairobi, Accra, Cairo and Kampala; Umba is getting pulled further west. It’s where the money is and the talent is. “There are a lot of people here who have already made their money, and now they’re now into the next thing,” said Kennedy. “Or they’ve scaled a company from two people to 100 people in two years. We are looking for those types of people.”

For example, Umba just hired a new chief technology officer. “He took a company public on the NASDAQ in 1997. So he’s been around the block. He has raised money off Sequoia, raised money off DFJ. And he’s already made his money. He’s 55 and he’s going to work for another 15 years. He says ‘I want to do something to tell my kids about that I’m proud of.’ So that we fit that thing.

“There are people out here that can do this. It’s not like ‘oh yer man there, he’s pretty good’. It’s like, he’s 10 times better than me. He’s handled an R&D budget of $50 million last year. So he can just take this and run with it, because it’s not new to him. It’s small for him.”

San Francisco is coming down with product managers and full-stack developers. But that’s only half the story. Those guys will deliver a good product, but actually getting it past regulators and into customers’ hands in Nigeria is a completely different challenge, needing a different team. Kennedy said: 

“From from the get-go, it’s always been boots on the ground. And it’s been either us there or our staff are there [In African markets] full time. Because you can’t really run a business in Kenya or Nigeria, without having people there – and really good people too. So what we’ve done is hired people out of either big banks or the fintechs. And they come in knowing what they’re doing.

“That’s been really helpful because we’re really strong right out of the gate in machine learning technology, company building, working with investors, that kind of stuff. 

“Where are we obviously want to bring in expertise in regulation – a huge part of our business. And customer experience, how to work with customers. And then marketing – how to speak to customers in the right way that makes this bank attractive. And that’s not going to be the same for every country.”

Banking is one of the most heavily regulated industries in the world, and Africa is no different. It’s made harder there by the fact that each market has its own central bank with its own requirements. Kennedy said:

“An early learning for us was that I used to be over there, handling regulation issues very regularly. Now we just hire really good people in the market, hire people out of banks who’ve done it before to deal with all the stuff. Because then they can be there week-on-week, in those meetings, building relationships. And they already have the relationships as well, they know the people. That’s been a massive needle mover. 

“So now when we look at new countries. It’s always, Who’s the compliance person? We hire them first, before anyone. Because it takes it takes the longest, you know,

“The hardest thing about working in Africa is they are all separate central banks. They might follow a similar framework, but they are different. So that’s one of the things we’ve invested in.”

Keeping the plates spinning requires a lot of travel. In the four-week period I’m in contact with Kennedy for this piece, he’s in Kenya for two weeks, California for two, and shortly to head off to Lagos for two. And when teams are distributed across 10 time zones, that changes the way things are done. 

“We’ve had to be asynchronous from the get-go,” said Kennedy. “Everything has to be written down in tickets and documented. Because it’s the only way you can run it. You have to ask a question and have it picked up an hour later, or seven hours later. 

“But the benefit is you end up with this big paper trail of how everything’s done. So in some ways, it builds in a process that’s better. 

“Running the business as an async organisation through software tools becomes a massive advantage in the long-term. If the question is ‘how does onboarding work?’, you can point to a Loom video that’s been done about how our onboarding works. That scales really fast. So you can take on 10 people, train up your customer experience staff, show them ‘here’s how it works’. 

“But of course it’s been a learning curve. Time zones are tough. I’ve been here since half six in the morning in San Francisco. Because you need to be on calls with our African bank partners, our consultant who works with the Central Bank. You’ve got to obviously manage that as well. So it’s challenging, but then, being here is massively advantageous to build a strong team, and then to raising money as well.”

What’s the right fintech banking business model?

If you think of the big FinTechs in Europe — Revolut, Curve, N26, Monzo. These companies have all busted a gut to get to where they are. They’ve done everything right – built great software, won over customers, won over regulators. But yet they’re all betting on completely different business models. Monzo and Revolut are betting users will pay fees. N26 is going more slowly, with the hope of winning customers’ trust and ultimately getting into lending. Curve aims to aggregate the rest of the industry. And the banks are hoping their traditional deposit and loan model doesn’t get disrupted. Someone has to be wrong.

What I’m saying is that executing on a strategy is hard enough. But it’s even harder, as a fintech, to pick the right strategy. It’s a fast-moving and complex industry. The winning companies have to solve all these hard problems; and anticipate how the industry is going to develop; and build the right business model to fit. 

Most of the fintech companies you’ve heard about have been focused on getting their apps into people’s hands. And they’ve been very successful at that. Revolut has 1.2 million users in Ireland alone. 

But getting apps onto people’s phones isn’t the same thing as building a profitable banking business. Take Revolut, for example. Lots of people have a Revolut app on their phones. But not a lot of people actually pay their salary into Revolut. This is true of all the fintechs. And it’s a critical point.

The incumbent banks have a lot of weaknesses, but their big strength is that everyone puts their salaries with them. This gives the banks very cheap funding, which they can then lend out at a profit. Because in banking, the real money is in lending. 

Revolut, Monzo, Curve can have all the bells and whistles they like. If people continue to pay their salaries directly into their AIB accounts, the fintechs won’t be able to lend and they won’t be able to make much money. I’ve written about the challenger banks’ challenges in the past.

Umba got a toehold in the market as a lender. “When we started out, because regulations are tricky around taking deposits, it was just lending, and a small amount of bill pay,” said Kennedy. “That’s how we got up and running and raised our seed round. But once we did the seed round, the perspective was really offering a full-service digital bank.” 

There are people out here that can do this. It’s not like ‘oh yer man there, he’s pretty good’. It’s like, he’s 10 times better than me. He’s handled an R&D budget of $50 million last year. So he can just take this and run with it, because it’s not new to him. It’s small for him.

Tiernan Kennedy

Unlike other Fintechs, which typically don’t offer lending, Umba decided to go the whole hog. But in order to lend money, a bank needs deposits:

“For our ambitions, we need deposits, a hundred per cent. That’s because deposits are so much cheaper than debt. Because fintechs are paying double-digit percentages on debt, and fintechs in emerging markets are paying more. So deposits are a much cheaper source of funding.

“So deposits are incredibly fundamental here. Some fintechs want to live in a world where they’re running on foreign exchange, they’re running on selling crypto, and they’re getting fees here and there. And that’s their business model.

“Ours is a very traditional banking business model. We take deposits. And we’re going to offer banking on top to bring in customers. You might make loans to 10 of them, you provide banking to the other 90, and you’re building up this huge bank of customers, and you’re using all the money they’re giving you. It’s traditional banking, and it’s the most powerful way to go. Everything else you do is just really hard to scale.”

So by necessity, while it was waiting for the regulator to let it take deposits, Umba spent years working on lending. This was hard work, but like the Kenyan marathon runners who train at altitude, it’s paying off for Umba now that it’s fully regulated. 

“The reason we did lending first is because, on the regulatory side, it’s much easier to get approved. But on the product side, lending is much harder and it takes a longer time to do. So I’ve got to build credit scores, and machine learning infrastructure. I’ve got to build all these training sets. There’s a tonne of stuff that happens in the background in our infrastructure just for us to give a yes or no decision on someone’s credit score. Because there are no agencies to credit score against. 

“So we’re taking all your bank transactions, the type of phone you’re using, how often you use it, who you’re phoning, every transaction you spend, how you’re spending it, what are you spending it on. So if you want to really analyze all that stuff, and get really good at that, that’s hard, it takes a long time. You can’t, you can’t really speed up that process. No one can build a credit score in two months, it’s just not possible, You need a good bit of time. So we focused on that first. 

“And then later as we got back to work and built up the banking side of our team, we found it wasn’t easy – but it is easier. We’re launching pretty major products every month at the moment, Because we can launch a savings account quite easily. From a technical perspective, it’s much easier for us to do that later in the game than it is to train a credit model on brand new data, that hasn’t been done before. 

Here in Europe, the FinTechs and the established banks are deadlocked. As we’ve seen, the Fintechs have delightful technology, but their customers don’t trust them with deposits. Established banks have the opposite problem. And while it’s clearly a lot of hard work to build a bank from scratch in an emerging market 10 time zones away, Umba’s big advantage is that the incumbent banks in Africa don’t enjoy as much trust as the ones in Europe. 

“One good thing is the incumbents we’re up against are extremely unpopular,” said Kennedy. “Like for example, you, Sean Keyes, might keep your salary account in Bank of Ireland or AIB because they do mortgages, they do car loans and whatever else. Whereas in our market, that’s not happening for most of the population. Those type of loans are out of reach. So we’re very lucky in the fact that there’s massive unmet credit demand.

“In Nigeria, you’re talking about $150 billion unmet credit demand, just for small businesses and SMEs. And that’s companies crying out for credit. Their economy is still growing extremely fast, in lieu of any really good credit services for smaller businesses and personal. So we’re in this world where access to credit is extremely minimal, there’s no shared credit reporting infrastructure, there’s no credit score. So the industry is really in its infancy. So for us, trust is, ‘hey, we’re offering better services than other banks.’ Whereas in your developed market, you say, well, ‘Revolut’s my walking around money, it’s a cool app. But I’m keeping my salary in AIB. Because they AIB is who I’m gonna bank with long term.’ We don’t have as much that problem.”

This is the exciting thing about Umba. Yes, obviously it’s hard to build a bank from scratch. Building a product and getting it launched in African markets, and getting it in people’s hands, and getting their money. These are tough problems. But none of the other big fintechs – the Revoluts or Monzos, are putting in those hard yards. They don’t have local credit scores and they’re not in with the regulators. Which means Umba has as good a chance at this as anyone.