Nik Storonsky: The most uncompromising banking CEO

Nik Storonsky, Founder and CEO of Revolut, joins John to break down how Revolut scales across 80 million customers worldwide, their strategy to challenge Amex and Chase in the US, and why products live or die by strict retention curves in their "new bets" process. They also discuss building a profitable fintech around subscriptions rather than lending, how they are cutting 210-day compliance approval cycles with AI, and why CEOs should skip conferences in favor of product development. Timestamps (00:00:00) Introduction (00:00:21) Revolut growth (00:04:03) New bets framework (00:12:35) Why bank lending is like selling options (00:20:12) Automated cross-sell machine (00:27:39) Stripe Radar (00:28:10) Lightweight users to primary accounts (00:33:16) Building the AI-pilled bank (00:39:04) Taking on Amex and Chase (00:42:36) Products over conferences

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00John Collison?

  Do you want to go head to head with Amex Platinum and Chase Sapphire?

• 0:03Speaker 01

  Yeah.

• 0:03John Collison?

  How will you compete? Nick Storonsky is the co-founder and CEO of Revolut, one of the world's leading neo-banks. They started with foreign exchange, but now do everything from bank accounts to business services to mobile plans. And they're expanding in the United States. Cheers.

• 0:17Speaker 01

  Cheers.

• 0:21John Collison?

  So as you look back on it, how did Revolut get to scale? Like what worked and caused growth?

• 0:28Speaker 01

  I mean, to be honest, I was never kind of optimizing to get scale quickly. I was just working on the product and then solving problems and then initially when I started, we had probably one to 10 customers per day. Then it went to 100 customers, then to 500, and then to 1,000. Now we have about 100,000 customers joining every day. So I was never kind of trying to optimize to get as many customers as possible. I was more focused on building products, gaining license and so on.

• 0:56John Collison?

  We presumably thought a lot about distribution dynamics and how to get customers to refer other customers and things like that. Like kind of the growth, consumer growth side of things.

• 1:06Speaker 01

  Yeah, we actually started doing growth, I think, 2022. First time where we started having like serious budgets on growth.

• 1:15John Collison?

  But you didn't for the first seven years of?

• 1:17Speaker 01

  No, we didn't. It was like purely organic, yes. And that's when we built, you know, all the models, you know, paybacks or TVs per customer. And then also build a referral marketing product as well.

• 1:30John Collison?

  But the first seven years was just organic, building financial products you thought were good.

• 1:34Speaker 01

  Yeah.

• 1:35John Collison?

  But how do people discover them?

• 1:36Speaker 01

  Road of mouth, plus we've done obviously PR as well.

• 1:41John Collison?

  What were the best products from an acquisition point of view?

• 1:46Speaker 01

  Well, the best product was initially first product we launched. It was a multi-currency card with zero FX fees.

• 1:53John Collison?

  It's interesting that both you and Wise were able to get to scale based on consumers frustrations with FX fees. Like the frustration with the incumbents in that product was so high that it has birthed multiple successful fintechs.

• 2:13Speaker 01

  Probably. I mean, if you look at banks, the majority of products are actually very bad, right? So it's not only kind of multi-currency transfers, but also credit, business credit, business products. Almost every single product is really bad. I mean, acquiring, used to be bank, right? So I think it wasn't about multi-currency transfers or card. It was more that we picked up certain products and we made it so much better. It could have been acquiring that you've done, it could have been maybe credit as well, wealth and trading, Robinhood done it.

• 2:46John Collison?

  So you think there's nothing special about starting with multi-currency cards? You think you could have in another universe started with a lot, you know, checking accounts?

• 2:53Speaker 01

  Yeah, I think as soon as you build much better product compared to banks, which is also much cheaper, then it's pretty easy.

• 3:00John Collison?

  What was the second product that worked that, you know? When did you prove to yourself that this wasn't a fluke and that you could repeatedly, repeatedly build financial products?

• 3:08Speaker 01

  Well, business accounts, we build business accounts. And offering wasn't about multi-currency card anymore, it was a digital bank account. And without any marketing, we first started probably with 10 business customers a day. Now we're onboarding maybe one and a half thousand business customers a day. And then only, I would say 30% of it is paid marketing, the rest is still organic.

• 3:33John Collison?

  One of the things that, looking at Silicon Valley,

• 3:38John Collison?

  successful product companies tells you is there are many, many paths to heaven. You know, there are many ways to build a successful product organization. And you have Apple that famously doesn't do any kind of focus group testing. And they just build the products that they know in their heart of hearts are right in this very functional organization. And then you have, you know, other Google or Facebook develop products in totally different ways. What is the Revolut system for, for building products?

• 4:07Speaker 01

  So we just separate a small team of 10 people, smart people, smart owner, good designer, good developers. We just build what makes sense from first principles point of view. So it can be either replicating existing product, but building it so much better. Or trying new products which we think makes sense. But to be honest, everything new is forgotten all, right? We haven't built any product which worked, which wasn't an old product, which was built badly by other organizations. We have internally a new Bets process where everyone in the company can raise an idea. And then it goes through validation. And then if idea is validated, then it goes to, well, called committee of people who've done it before. And then they decide whether to fund the idea or not.

• 4:53John Collison?

  What's validation?

• 4:53Speaker 01

  Well, first, I think it goes to team which validates, like, you know, unit economics, financials and so on, whether it makes sense. The majority of cases makes sense because cost of the team is pretty low compared to if idea is useful for customers, they pay for it. You can generate like tens of millions of dollars, right, compared to cost of the team, which is small. And then it goes to people who built new bets before, and then general managers. And then they also validate it from, again, purely common sense perspective. And then if everything is approved, then we separate a small team. We hire people there, additional people there.

• 5:36Speaker 01

  Then they build a product within nine to 18 months. And then we look at the growth of the product. So if there is a lot of growth, then we scale up the team. So one team becomes two teams, three teams, four teams, department and so on. And then for the bets which don't grow, we sunset them. And then for bets that grow but not very fast, we just leave them alone.

• 5:59John Collison?

  But I'm surprised that ideas would need to be validated economically because you're distributing products across millions of customers. Won't the cost of, you know, can't kind of any idea support a five to 10 person engineering team to go develop a V1 of it? Like, it's very unlikely that a product won't work economically when you have such a large customer base and you can kind of fix it with pricing.

• 6:21Speaker 01

  Like examples, we launched Revolut Lite across many countries, 100 plus countries, without the banking license and without our e-money license. Purely like, you know, money transfer business. And then the idea was to build Venma internationally.

• 6:36Speaker 01

  It didn't really work out because we thought distribution costs would be much lower. So it didn't really grow organically fast enough. So we had to do performance marketing and then unit economics or the performance marketing didn't really work in terms of payback. So in majority of cases, 90% of cases, yes, you're right. If it works out, it works all very well financially, but there are some cases, exceptions, which you still need to validate.

• 7:04John Collison?

  Brian Armstrong from Coinbase has talked about how when he was first pitched USDC, he thought it was kind of a silly idea, he wasn't convinced, but sure, okay, if you guys want to go work on that, have at it. Are there ideas that came to you that you were initially not fully convinced, but ended up being successful?

• 7:23Speaker 01

  To be honest, I'm always, in 99% of cases, I'm always up for new ideas and just try them.

• 7:29John Collison?

  You're just open-minded, yeah. Maybe the idea that surprised you, you didn't expect it to work. You were open-minded, but you were pessimistic, but then it had performed your expectations.

• 7:38Speaker 01

  I just had too many failures in the past when my judgment was wrong. So now I'm like fully optimistic. I think it's, it's not about the idea, even though idea is important, I think it's about execution, to be honest. You can make almost any idea work.

• 7:56John Collison?

  Can you? Even the bad ones?

• 8:01Speaker 01

  If it's not that bad, then you can make it work, provided you spend enough effort and enough time on it.

• 8:05John Collison?

  And people are bringing you the plausible ideas.

• 8:08Speaker 01

  I mean, if the team is very strong, then even if idea is average, you can still make it work. There are cases when idea is great, business model is great.

• 8:16John Collison?

  Yes.

• 8:17Speaker 01

  Team can be bad, it can still fly.

• 8:19John Collison?

  Well, what's your favorite example of the great idea, great team intersection?

• 8:26Speaker 01

  Well, I think our subscriptions team or department is very good. And then when they came and then they said, okay, we want to launch like additional ultra subscription. I thought it was, idea was okay, right? It ended up, you know, being amazing bet because it's a very expensive subscription. So ultra is effectively ultra bank account where you pay about, I think it's 50 euros per month. But then it gives you additional, a lot of additional benefits. And I thought, I mean, idea is okay, probably this business won't be very large. It ended up being very, very large business.

• 9:03John Collison?

  The closest comparison, it's kind of like an Amex premium card or something? Yes, yes.

• 9:10John Collison?

  I'm so interested to hear you say that growth dynamics were not critical in the early days of Revolut because firstly, so many financial products spread. Because, you know, a payment transaction has two legs. And so it's just very well suited to viral product growth, whether it be, you know, bill.com or, you know, PayPal, where you can send money to an email address, even if they are not signed up for PayPal. Like, it's a very classic acquisition dynamic. And so it's interesting to me that that wasn't the case for, for you. And also, I mean, the peer-to-peer product is now so successful.

• 9:46Speaker 01

  Peer-to-peer is definitely an acquisition driver, especially in countries where they don't have a very good payment system.

• 9:52John Collison?

  Yes.

• 9:52Speaker 01

  So for example, in certain markets in Europe, we

• 9:55John Collison?

  Ireland.

• 9:56Speaker 01

  Ireland is one of them, Romania is another. So yes, it worked in many markets where a bank payment system is pretty old.

• 10:07John Collison?

  Okay.

• 10:07Speaker 01

  But I would almost say it was primary kind of position.

• 10:10John Collison?

  Yeah. Because I would draw a distinction between word of mouth, where someone has to like go to a bar and be telling you, hey, I discovered this great new product last night, and then viral product adoption mechanics, where if I want to pay you money, you become a Revolut user and then you can tune it. It's like, have you thought of, you know, sending money and you can promote that? You know, Meta does a very good job with this, where when they want to grow new parts of the product, they promote them in the app. And again, you have knobs and dials similarly that you can use. And so the in-product growth stuff, to me.

• 10:40Speaker 01

  Yeah, of course. We tried a lot of things, P2P included. Like problem with banking product KYC onboarding, right? There's a lot of friction.

• 10:48John Collison?

  Uh huh.

• 10:49Speaker 01

  So even if you have this viral things, then.

• 10:52John Collison?

  This is a steep drop off.

• 10:54Speaker 01

  Yeah, it's a lot of drop off, yeah.

• 10:55John Collison?

  Yeah.

• 10:56Speaker 01

  So it's not really viral.

• 10:57John Collison?

  Yes.

• 10:58Speaker 01

  Plus, you cannot really sell it worldwide, right? You always need to sell into countries where you have licenses.

• 11:04John Collison?

  Yes, but it can be viral in the same way that, you know, Facebook was viral in the early days, where they would go college campus by college campus. Like, you can't do the cross-border transfers, but there's so much activity within a country that you can get viral growth within a country.

• 11:22Speaker 01

  If you can onboard subject to one click.

• 11:24John Collison?

  Yeah.

• 11:24Speaker 01

  As a Facebook, yes.

• 11:25John Collison?

  Yeah.

• 11:26Speaker 01

  Yeah.

• 11:26John Collison?

  Which, and again, you think the KYC is a bigger impediment.

• 11:29Speaker 01

  So when we first launched, there was a lot of virality, because we're doing onboarding. I think it was up to 1000 or 100 euros. You can do it without KYC. So back then, this regulation allowed to do it. So there was a lot of virality, but then I think it was stopped 2017 or something.

• 11:49John Collison?

  So the KYC changes actually made the viral growth loop a little harder.

• 11:54Speaker 01

  Oh, yeah, 100%, yeah.

• 11:55John Collison?

  Though you can presumably still do a lot of nice viral growth, where, you know, if there's money waiting for you and you just need to go through KYC to claim it, then like you can do one leg of the transaction and not complete it, right? And that will still be plenty enticing for a customer.

• 12:09Speaker 01

  That's true. Yeah, we also have it.

• 12:13Speaker 01

  But again, it's like, you know, this K fact is never larger than one.

• 12:16John Collison?

  Yeah.

• 12:17Speaker 01

  So you cannot really experience financial growth. It's always, you know, less than one.

• 12:21John Collison?

  But you're not, like, there could be a K greater than one product.

• 12:26John Collison?

  Not too pessimistic.

• 12:28Speaker 01

  Everything we've tried so far, last 12 years, always less than one.

• 12:33John Collison?

  Okay, so you're in the K is less than one business. The thing I find interesting about Revolut's business is you're a bank whose primary revenue source is not net interest margin, which is obviously unlike other banks. Did you arrive at this accidentally? Was it part of your thesis for Revolut? Like, you're very different in that way, and it's quite asymmetric versus other banks.

• 12:52Speaker 01

  I think it's pretty, the balance is pretty simple, because we started in UK and Europe. And then in UK and Europe, conditions are very tough, right, for startups. So interchange is very low, plus credit is extremely regulated. You cannot really charge customers very high interest rate, same as you can do in Mexico, Brazil, for example. So you end up, for example, you'll have like 10, 20, 30 startups starting in financial services, and then the majority would die, right? And then we were forced to optimize our transactional business and to effectively make it profitable. If we would start in Brazil, Mexico would obviously be focused on credit business, because it's less regulated compared to Europe.

• 13:37Speaker 01

  And we would make a lot of money on a credit business, but in Europe, it's not possible.

• 13:41John Collison?

  But aren't you developing a balance sheet? Because you have one side of it, which is the deposits, and you could build the other side of it with loans.

• 13:48Speaker 01

  Yeah, but before we got deposits, I mean, we couldn't build a credit business. Now, now.

• 13:52John Collison?

  But now you could.

• 13:53Speaker 01

  Yes, now, yes, we are.

• 13:55John Collison?

  And will you build a traditional bank credit business, where you have the deposit side and you have the loan side? Or will it look atypical in some way?

• 14:05Speaker 01

  Well, I think our business model, and then my personal opinion is we want to have the least amount of risk. So even if we lend, it won't be a large portion of balance sheet. So at the moment, target state is less than 10%, and then the rest we would secure it as.

• 14:22John Collison?

  Why?

• 14:23Speaker 01

  Well, first, the simpler the model, the better, from a greater point of view. Plus, again, if we want to be a tech business, right? We would rather be in subscription business. As soon as you go into lending business, your multiples become similar to banking multiples. So if you look at the tech business multiple.

• 14:43John Collison?

  Yes.

• 14:43Speaker 01

  You can usually get like 40, 50 PE. In banking business, you'll be like if you get 10 PE.

• 14:50John Collison?

  But is that because, because like I'm always skeptical of multiples-based arguments, because ultimately, the business is good or it isn't. And, you know, you can educate investors if the business is in fact good. Like, when we started Stripe, people thought payments businesses were horrendous and deserving of low multiples and everything like that. And like, over time, they came to appreciate payments businesses more. But if we had been fixated on kind of the multiple, you know, that people ascribed to the business, we might have, you know, been scared off this path. And so I think you just need to decide, is this lower quality or higher quality as a profit pool? And so do you think it's like actually a lower quality profit pool? Or investors just have a weird hangup?

• 15:31Speaker 01

  I think it's more the business itself. So if you look at the credit business and banking, general credit business, several times when you do a loan or credit card, you need to put provisions against it. And then reality amount of these provisions highly depend on the models that you use, on economic conditions. Because reality, you don't know, like, you know, in three or four or five years, will you receive the loan back, right? So you kind of model it. On average, you hope to receive money back. But reality, things can change, right? And this uncertainty is being pressed in, I think, by investors, generally by low multiples, because, okay, bank can produce a lot of money now from credit portfolio, but look, in two, three years' time, unemployment goes from 3% to 6%.

• 16:15John Collison?

  Yep.

• 16:15Speaker 01

  And all profit is gone, right? Plus, they're losing money. So I think this.

• 16:20John Collison?

  Okay. So you think fundamentally, making your money via credit is this very cyclical profit stream that can be entirely wiped out in a downturn? It's very sensitive to you doing the math right. Whereas the nice thing about subscriptions is ultimately, people are paying you this money on a recurring basis. And, yeah, in a downturn, maybe some people cancel their subscriptions, but like, you can't make negative money on subscriptions.

• 16:47Speaker 01

  Yeah, exactly.

• 16:47John Collison?

  Yeah, shrink the book slightly, and so it's much higher quality.

• 16:50Speaker 01

  Yeah, exactly. It's similar to selling options, right?

• 16:53John Collison?

  Yeah.

• 16:53Speaker 01

  You kind of collect premium for 10 years, and then year number 11, you lose it all.

• 16:58John Collison?

  Yes, yes.

• 16:59Speaker 01

  So that's a problem.

• 17:00John Collison?

  Okay, I want to go back to your product development methodology, though. So, okay, you're describing the early, the early stages of there's some tops-down product ideas, there's some bottoms-up product ideas. They have to be validated that if it works, it can actually be interesting economically. You have a talented product leader and a few engineers working on it. Then what? Like, how do you decide which products actually get the in-product distribution? How do they earn that? How much are you willing to give them credit for? Oh, I think this will become a good product over time. Product developers always tell you, oh, I would be doing very well if you just gave me more distribution.

• 17:42John Collison?

  But like, the product people like to blame distribution, the distribution people like to blame the product. Maybe you can talk a little bit about how the scale up works.

• 17:49Speaker 01

  So basically, it's pretty simple, right? So we've got the team that work on the product for up to 18 months. Then hopefully, they launch the product.

• 17:57John Collison?

  They'll work up to 18 months before launching anything?

• 18:00Speaker 01

  So from the moment we decide to launch, to launch is usually 12 to 18 months. Because it takes time to get the team, onboard the team, build the product, go through all this process on the banking side. So 18 months, probably a good number.

• 18:16John Collison?

  Because you're a regulated bank, you have to go to your bank regulator to get sign off on the product.

• 18:21Speaker 01

  It's not the bank regulator, but more like, you know, internal governance.

• 18:24John Collison?

  Oh, I see, your internal function, yeah.

• 18:27Speaker 01

  So we launch the product, and then we release it. We first release it in a test country, for example, Ireland or UK or whatever. And then if it works, if everything is fine, we release it to more countries and more countries and more countries. And then we add it to our cross-sell engine as well, which is automated. So there is no kind of, you know, distribution problem, right? It's automatically being added to, hopefully, the whole user base, if everything works fine. And because we've done it for a long time now, we have already historical data on performance of, say, about 100 beds. And then we just look at month one after launch, what is gross profit compared to previous beds?

• 19:11Speaker 01

  Month two, month three, month four, month 12. And then how is it stuck with all the other beds? If it is top 25% quantile of all the beds, we just scale it up. So we hire more people, we have two teams, three teams, four teams, and so on. Bottom 25%, we just don't do much, so we redistribute the team to do something else. And then in the middle, sometimes we keep the team, sometimes we don't keep the team.

• 19:36John Collison?

  Is a new product like that actually an option? Or are there costs to shutting it down? Like, are consumers annoyed if like they were relying on a product and then it goes away? Or are there some products you kind of, I don't know, if you had a pension product, you can't, you actually just kind of can't turn that off, because it's a pension or something? And so do you think about the cost of shutting off a product when you decide to fund it? Does that differ by product?

• 19:58Speaker 01

  Yeah, of course. We look at custom detriment, right? If there are already customers using the product. If there is detriment, we just keep it as it is. So we stop investing in it, so we just maintain it.

• 20:11John Collison?

  Can you tell me about the automated cross-sell machine?

• 20:14Speaker 01

  Automated cross-sell machine, well, basically, we have many products on the platform. And then we effectively have a machine learning model that decides based on certain events, based on customer profile, which is the most suitable product at this point of time. For example, it can be, I don't know, you land in Ireland, you have our UK account with Revolut, and then we offer to you ECIM, right? Because, for example, you land into Ireland and then you spend money on Air Allo, right? Because we see all the transactions, we can offer you similar products from ours.

• 20:55Speaker 01

  And on top of it, we also have a LLM model, which effectively behaves like as account manager, right? It, through, you know, human reasoning, it looks all the products, at all events, and all suggestions generated by machine learning model, and then decides which product to push to your customer in this particular point of time. This is kind of all automated.

• 21:23John Collison?

  I hadn't thought that you are in a much better position to cross-sell products than anyone else, because you see what people are spending money on.

• 21:32Speaker 01

  Yeah, as well, yeah.

• 21:33John Collison?

  You can just, if they're overpaying for something, you kind of know.

• 21:37Speaker 01

  I mean, we don't see the full details of the bill, right? We just see the transaction.

• 21:43John Collison?

  And so the cross-sell engine has some propensity modeled for the customer, and you're basically trying to find the conversion maximizing sale, like the thing that they are most likely to buy next?

• 21:57Speaker 01

  It's a bit more complex there. It's not only conversion, it also takes into account the GP, the product. And we also try not to spam customers, so there is certain negative value of our message to your customer. So altogether, model makes the decision.

• 22:15John Collison?

  And it's interesting that you describe products needing to perform well on a revenue cohort curve, where it's a little like a race, where like your job is to stay in the pack or stay at the front of the pack, but you're judged relative to other products. Have you ever had slow-burn products that, for whatever reason, had a gradual start? Maybe the team was just like learning exactly how this works, but like that wasn't actually any negative signal about the product? It just took a while for them to get going and then really took off?

• 22:44Speaker 01

  To be honest, it's almost never happened. I don't even remember the example when it was kind of slowly growing and then started growing fast. Usually, trend is your friend. The thing is growing in very quickly, first four months, it will continue to grow very quickly.

• 23:00John Collison?

  Because, you know, a classic bigger company behavior might be to say, "Oh, this isn't working yet, but there's a good reason for it and blah, blah, blah." You think that's all excuses and the trend is your friend.

• 23:11Speaker 01

  Generally, the statistical trend is your friend, yeah.

• 23:13John Collison?

  Yeah. Why don't other companies operate in this Revolut model of developing many new additional products to cross-sell to customers? It feels applicable to many other consumer companies. Grading them against the curve of how other products have done at a similar age, I don't think most companies, even that sell multiple products, are that strict. Instead, they make excuses for why the product isn't performing, and, "Oh yeah, we just need more staffing," and, you know.

• 23:45John Collison?

  They're not as strict.

• 23:46Speaker 01

  Yeah, I think, I think because majority of companies, they don't do many bets at all, right? So, because we've done a lot of bets and we had feedback loop a lot as well compared to other companies, and because of that, we are very strict.

• 24:02John Collison?

  Yes.

• 24:03Speaker 01

  Other companies, okay, they do probably seven, eight bets, and feedback loop is delayed, plus management changes. So as a result, they never end up implementing any kind of strict process on your bets.

• 24:14John Collison?

  Yeah. Are there products you've done that have not been well-suited to this model? I can imagine, say, a product that involves changing many different parts of the Revolut experience would be hard to fit into this new bet model, because it's fundamentally much more infrastructure rewiring or something like that, I don't know. But yeah, do you sometimes have to do products that don't fit into this standalone bet model?

• 24:43Speaker 01

  I mean, to be honest, you can fit any product into this model, even if it is a product which is very new to us, for example, not financial product, and it can be B2B product. For example, we've done a bet, Revolut people. So we built internally our HR system for ourselves, and we decided to make it a new bet. And then we separate into standalone team, plus we sell it externally. It's kind of a new bet, but we cannot really cross-sell it to our retail customers.

• 25:14John Collison?

  What about products that just don't generate revenue at all by design? And so imagine if I'm a product leader and I'm building a financial insights product, and it generates a lot of engagement with the app, because it sends people a summary of their transaction activity and ways to spend money and blah, blah, blah. But like, the point is not revenue, the point is app engagement.

• 25:38John Collison?

  How would you grade my success?

• 25:40Speaker 01

  We still have P and L attribution to it by indirect attribution of P and L, because you can also calculate it.

• 25:48John Collison?

  So you have some churn model?

• 25:49Speaker 01

  So we'll have two numbers. One is direct GP and another one is indirect GP.

• 25:54John Collison?

  And so I have some synthetic profit that I can attribute to myself, because I saved this many customers from churning by having them come back to the app or had them spend incrementally.

• 26:07Speaker 01

  I'll give you examples. So we got RAP points, it was also a new bet.

• 26:11John Collison?

  What is RAP points?

• 26:12Speaker 01

  RAP points is more like a reward system, yes. So you've got points and you can effectively using these points to buy additional products that we offer.

• 26:18John Collison?

  Yes.

• 26:19Speaker 01

  As accounting GP, it's very small. But then if you could create indirect GP, then it's very high.

• 26:26John Collison?

  That's very interesting, because I feel like a lot of the most successful companies get to an understanding of their economic equation and then take that very seriously, where, you know, Amazon says, "Oh, an incremental Prime subscriber is extremely valuable, and so we're going to set up a movie studio so that we drive more Prime subscriptions." And it sounds crazy, but like, if you really believe the math on how much a Prime subscriber is worth, then it makes sense. Or Google was always very fixated on just how valuable the incremental search customer was because of the lifetime value of the advertising revenue. And so they do all sorts of crazy things, you know, pay a lot of money to become the default Safari search box.

• 27:08John Collison?

  But again, it was all that attributed P and L. And so you think the attributed P and L revenue is very important so that you can actually then create accountability with product teams to know if something's working or not?

• 27:24Speaker 01

  Ultimately, you need to have one number which defines whether a product is successful or not.

• 27:28John Collison?

  Yeah. I think a lot of companies actually don't do that, to a surprising extent, chase through the full accountability with a synthetic revenue number.

• 27:39John Collison?

  Revolut's global scale presents a hard but familiar problem: how to stop fraudulent actors without turning away legitimate ones? So Revolut uses Stripe Radar. More than nine in 10 payment cards that show up on your checkout page have been seen before by Stripe, and so that gives you a serious head start separating good activity from bad. Add to that fraud detection models trained on more than 70 trillion data points across the internet, and you're off to the races. So if you're trying to reduce fraud without losing legitimate sales, visit stripe.com/radar.

• 28:10John Collison?

  How important is the way people think about neo-banks, as I understand it? Whether there's like an account or an account is whether people deposit their paychecks. And so you can use Revolut, like, a majority of Ireland, like, uses Revolut in some degree, but they might just use it for peer-to-peer or something like that. But they're like, they have an account. And then there's like using Revolut as your primary bank account, which in the banking industry, people basically think of as, "Do you get your paycheck deposited there?" How important is it that you move people from being a lightweight user to a heavyweight user, from like occasionally using certain features to actually being the place that the paycheck is deposited?

• 28:50Speaker 01

  I mean, it's super important, obviously, right? Because as soon as you get salary into the account, then people use a lot of other products that we have, that we have. If they're just using PTP, they, I don't know, load $20, send money to someone, it's not very profitable for us. So we spend a lot of effort to move salaries of these people to Revolut. And overall, I mean, it depends on the country. Like, it's from 10% to 20% of monthly active users, we have our salary accounts.

• 29:23John Collison?

  What moves people over?

• 29:24Speaker 01

  It's a very slow burner, right? So we actually spend a lot of time and effort building products for kids, because kids, as soon as we capture kids, then we can easily move them to 18-plus account, because it's very hard to move adults.

• 29:43John Collison?

  Yes. It's a tale as old as time. And I remember when I was in secondary school, maybe a similar experience, the local bank came and they gave us a soda and a bag of chips to sit through a presentation on like how to save money, and then they gave us all a bank account. But, you know, they were in the school, age 13 or 14, doing customer acquisition on the theory that bank accounts are very sticky. But, okay, but you see very different behaviors from the people. Presumably, like, if you have 10% to 20% primary banking share amongst kind of the regular population, presumably that's way higher amongst the younger population, because they're Revolut native.

• 30:26Speaker 01

  If you look at the different cohorts of users, then we probably capture more than 50% of users up to 35 years old in almost every single country where we're in, in Europe. And then where we have difficulties capturing 50, 60, 70-plus people who still hold, you know, majority of liquid assets, but then it's very hard for them to move, you know, from legacy bank to digital bank.

• 30:55John Collison?

  And so, what's that 50% of under 35 statistic?

• 31:00Speaker 01

  Well, basically, if you take all the countries in Europe,

• 31:02John Collison?

  Yeah.

• 31:03Speaker 01

  And then if you take, I don't know, kids, then people from 18 to 25, 25 to 30, 30 to 35, then older, in majority of countries, or...

• 31:14Speaker 01

  We already captured more than fifty percent of people up to thirty five years old.

• 31:19John Collison?

  Having any kind of Revolut account?

• 31:20Speaker 01

  Yeah.

• 31:21John Collison?

  How is this product development engine different as a regulated business versus if you were just a normal tech company?

• 31:29Speaker 01

  Well, I think in a normal tech company, it's much easier to do because in banking, you need to go through many approval process. Another way we structured, so basically because we have many banks. So then the way the approval process works, we called NEAP, New Product Approval Process.

• 31:48John Collison?

  How do you minimize churn?

• 31:50Speaker 01

  Well, trying to build the best product.

• 31:52John Collison?

  But more than that?

• 31:55Speaker 01

  More than that.

• 31:59John Collison?

  Because again, you're very quantitative, you're very product oriented. And this is the thing that quantitative, product oriented people obsess over.

• 32:07Speaker 01

  Yeah, we used to have this framework of State Machine. You're probably familiar with it. How many people went over, how many people went through, I don't know, from being PTP customers to seller customers, how many of them churned. But then it just became a little too complex.

• 32:25Speaker 01

  So we end up just having co-host of customers who joined the country. And then how many of them are still with us within twelve months, eighteen months, twenty four months. And then what we observed is that we have a smile, right? So out of hundred customers, monthly active can be, say, sixty people can be monthly active, but then after two years, it will be sixty five, seventy customers, so people return to us.

• 32:51John Collison?

  Yes.

• 32:52Speaker 01

  So we actually have, I would say, negative churn.

• 32:56John Collison?

  Why do people come back? Like, why do people leave but then come back?

• 32:59Speaker 01

  I think it's more because a large and larger portion of population started using Revolut. So people who got churned because of different reasons, they still hear about it from friends, family, and then they come back.

• 33:10John Collison?

  I see. So you have multiple word of mouth kind of shots and go. Okay, so yeah, I was curious to ask you about AI.

• 33:18John Collison?

  How AI-pilled are you in how you run the company?

• 33:22Speaker 01

  Well, I obviously started with AI about two years ago. And then goal was just to optimize customer support, translation. Now it's almost in every single area of the company, obviously engineering as well, marketing. Now we're even trying to create an AI employee end to end. So yes, we're very involved. And last ten years, I optimized a lot of things in the company just based on people, based on advantages and disadvantages of people. We build all the systems of KPIs, roadmap, process, hiring, performance management. And now it might be all redundant, right? Because of AI, in two, three years type.

• 34:05Speaker 01

  So it's kind of, you know, we're all reinventing how the company should work.

• 34:10John Collison?

  How's it changed how you hire either at an overall company level or at an individual team level? Like, are the eight-person teams, should they now be three-person teams? Or can just, you have higher aspirations for what an eight-person team can accomplish?

• 34:25Speaker 01

  I mean, technically, one person can accomplish much more now. I would even say you don't really need to have separate entry developers, iOS developers. One front-end developer can do things. You can even have, you know, one developer covering front and back end.

• 34:40John Collison?

  Yes.

• 34:40Speaker 01

  Possibly data science as well. So you can have, you know, one person doing all aspects of the product. Should probably still have designer, although it's so much easier to design. So one developer can do a design as well. So a team can be of two people, probably.

• 34:56John Collison?

  But like, you work in this relatively standardized way of product bets where you put a small team on something to initially test it. Has the size of that product bet initial team changed? And shouldn't it change if not?

• 35:10Speaker 01

  I mean, not yet, to be honest here. So we're still operating in the old framework.

• 35:14John Collison?

  Isn't that very old fashioned of you?

• 35:16Speaker 01

  Yes, pretty old fashioned now, yeah. But I just prefer to continue doing things that work. And then in the meantime, experimenting with new things. And then when we have a proven track record that new things really work for us, we will start changing.

• 35:29John Collison?

  But that's interesting because you, you think that a four-person team is too, like, kind of cannot bring a new product to market even with all the new productivity enabled by AI?

• 35:41Speaker 01

  No, four, four, four people team definitely can. It's just because all the process within Revolut are still, we're still set up, like, you know, by specializations. Okay, you have iOS developer, and they have the system. They know how system operates. Android developer, then backend developer, then data science, then ops manager. So I think it's, it will be very hard to just hire one person who will get access to all the systems and then produce with AI, all required code.

• 36:10John Collison?

  I see. So partly it's just you need to reorient Revolut a bit.

• 36:15Speaker 01

  So we need to change process here.

• 36:16John Collison?

  So how does this change the shape of the Revolut organization?

• 36:20Speaker 01

  Well, first of all, we're just trying to automate a lot of things. And then hopefully, when automation is done, we can start hiring less people. Because at the moment, I don't think we're at the stage when automation is done.

• 36:34John Collison?

  You traditionally competed with legacy banks. And so people might go to their bank for FX, or they might use a Revolut multi-currency card. And you had a lot of advantages where you had no tech debt and you had no retail branch presence to have to afford and kind of many other advantages versus traditional banks. And so you're very successful. As you go to Brazil, you compete with Nubank. And as you go to, you know, Kazakhstan, I guess, you compete with Caspi, is that their name?

• 37:06Speaker 01

  Yeah, the Kazpi.

• 37:07John Collison?

  Caspi, yeah?

• 37:08Speaker 01

  We're not in Kazakhstan yet.

• 37:09John Collison?

  Okay, but when you, when you do. How is the world going to look different when you're competing with new banks rather than traditional incumbents?

• 37:19Speaker 01

  Well, obviously, it's more difficult, right? Ultimately, when we go into the country, the decision to go into the country is a combination of how bad banks are, right? Because there's less competition. And then what is kind of, you know, our unit economics, the customer, and how difficult is to get a license?

• 37:39John Collison?

  Yeah.

• 37:40Speaker 01

  Obviously, if a country has a lot of digital banking competition, it's difficult to get a license. Unit economics is not great, we will not go there, or we try to kind of de-pretize it. And then we try to pretize markets where there are a lot of legacy banks. Unit economics is great. And then hopefully, it's easy to get a license.

• 38:00John Collison?

  Okay, so you look for markets with sleepy incumbents.

• 38:03Speaker 01

  Exactly, yeah.

• 38:03John Collison?

  That's a good market for Revolut. What predicts the unit economics of a market other than how rich people in that market are?

• 38:12Speaker 01

  Well, it's interchange, legislation, laws. And then legislation on credit rates as well. And then generally, if inflation is higher, then credit rates are even higher.

• 38:27John Collison?

  Yes.

• 38:27Speaker 01

  And then the margins are even higher.

• 38:29John Collison?

  But isn't it surprising then that you grew up in Europe where both interchange rates and lending rates are regulated? Like, wouldn't that be a bad market in this rubric?

• 38:38Speaker 01

  Yeah, it's a very bad market, yeah.

• 38:40John Collison?

  So does the rubric miss something?

• 38:42Speaker 01

  I think it's just because we started in Europe where it is very hard.

• 38:49Speaker 01

  I think out of, you know, fifty digital banks, probably like, you know, three, four survived. So we really survived in a tough competition. But then, you know, that means we are more kind of, you know, fit to compete in other markets.

• 39:02John Collison?

  Yeah, so let's talk about the US market. The US has historically been very resistant to new banks.

• 39:12John Collison?

  Remember Simple?

• 39:13Speaker 01

  Yeah.

• 39:13John Collison?

  Like, that was what, twenty twelve, twenty thirteen?

• 39:17Speaker 01

  I think that was the earliest one, yeah.

• 39:19John Collison?

  Yeah, exactly.

• 39:19Speaker 01

  Even before Europe, yes.

• 39:20John Collison?

  Yeah, around then. And various foreign new banks have tried to enter the US market. Why, like, the US is such a quick adopter of so many technologies and invents so many of the technologies. Why is, like, leaving aside Revolut for a second, we'll come back to that. Why is the US so slow to get new banks?

• 39:40Speaker 01

  Well, I think historically, it was a very tough regulator. US regulator was one of the toughest ones. That's number one. Number two, all these startups, they couldn't really get a banking license, so they chose not to get a banking license because it was difficult. The traditional tech companies, they don't want to be regulated because it kills their business. And as a result, they jumped into partnerships with legacy banks. And it doesn't really work because, okay, we've got one tech, and then we've got compliance people in banks. And it's a different company, they don't really, you know, match together.

• 40:13John Collison?

  Yes.

• 40:13Speaker 01

  So the model was wrong.

• 40:14John Collison?

  So you think you have to be willing to own the stack end to end, including all the licensing?

• 40:20Speaker 01

  Including the license.

• 40:20John Collison?

  Which was tricky in the US because it was such a stern licensing regime.

• 40:24Speaker 01

  Yeah, plus, culture was, okay, you need to have tech separate to compliance.

• 40:30John Collison?

  You need to have tech.

• 40:30Speaker 01

  Reality is one business, right? So you also need to automate tech, automate compliance with tech and so on.

• 40:36John Collison?

  Yes. And so why did we not see, you know, a version of Simple that worked where, again, Simple tried to do the tech via partners? But why didn't someone try, why didn't someone try to start a new bank that owned the licensing end to end in the US?

• 40:53Speaker 01

  I think it was very difficult to get a license. I know Square, Cash App, I think they have a banking license now. They now activated, but it took them four, five years.

• 41:03John Collison?

  Yeah.

• 41:04Speaker 01

  Now time change, depending with the new administration.

• 41:07John Collison?

  And so is the current regulatory environment better in the US?

• 41:10Speaker 01

  Process quicker, yes. Now you can get a license within a year.

• 41:13John Collison?

  Okay.

• 41:13Speaker 01

  Before it took you, it could take, you know, four, five years.

• 41:16John Collison?

  And you guys have gotten that, you will get that?

• 41:18Speaker 01

  So we received the conditional approval. So there are like, you know, two more approvals that we need to receive.

• 41:23John Collison?

  And this from the OCC?

• 41:25Speaker 01

  OCC, yes.

• 41:25John Collison?

  Okay, very good. And so you guys will be going big in the US soon? Like, when you get the?

• 41:31Speaker 01

  Well, next year, yeah, hopefully.

• 41:32John Collison?

  The rest of that. What do need to change about the products to be successful in the US?

• 41:37Speaker 01

  Well, definitely needs to be credit card led, because you've got a lot of benefits in the US.

• 41:43John Collison?

  Yeah.

• 41:44Speaker 01

  From Amex, from Chase.

• 41:46John Collison?

  Yeah.

• 41:46Speaker 01

  So we'll do the same, it will be credit card led proposition.

• 41:50John Collison?

  Yeah.

• 41:50Speaker 01

  A lot of benefits as well.

• 41:51John Collison?

  Okay, so you want to go head to head with Amex Platinum and Chase Sapphire?

• 41:54Speaker 01

  Yeah.

• 41:55John Collison?

  How will you compete?

• 41:56Speaker 01

  Well, offering the same benefits are better, plus the whole proposition that we have, because they're offering just, you know, payments products, we're offering wealth, crypto trading, and so on.

• 42:08John Collison?

  Why launch your own Euro stablecoin?

• 42:11Speaker 01

  It's good business, obviously, as number one.

• 42:16Speaker 01

  Number two is basically there is USD, stablecoin, and there is no equivalent or credible equivalent for euros. I'll give them the other way in Europe. We should do it.

• 42:28John Collison?

  Yes, you said you've been doing this for a very long time. It's been multiple years.

• 42:32Speaker 01

  Yeah, tech was built a long time ago.

• 42:35John Collison?

  How do you organize your day? Are you waking up and going wherever your curiosity takes you? Or is this hyper-scheduled?

• 42:41Speaker 01

  Monday to Friday, it's hyper-scheduled, 15 or 30 minutes slot. So usually it's like business reviews, product reviews, one-on-ones. And then Saturday, Sunday is more like freelancing and free-thinking.

• 42:55John Collison?

  Revolut is the kind of company that has a work-from-home policy of, you know, you can work from home two days a week, Saturday and Sunday?

• 43:01Speaker 01

  Yes, true. I try to limit our external meetings as much as possible and spend 95% of my time internally.

• 43:09John Collison?

  Do you think most CEOs are too externally facing in how they spend their time? Going to conferences and all this kind of stuff?

• 43:15Speaker 01

  A hundred percent. I think I do have a location for conferences, maybe, you know, two, three times a year about the session.

• 43:23John Collison?

  Yeah. And you just want to be within the building reviewing the actual product?

• 43:28Speaker 01

  Exactly. Products and the business process.

• 43:30John Collison?

  What are you actually reviewing? Like, are you looking at metrics, dashboards with people? And kind of what are you, Nick, adding versus you'd also have an automated system that, you know, shuts down any product that's failing to meet its revenue goals?

• 43:43Speaker 01

  Yeah, basically, we have all these KPIs, roadmaps and processes, actually targeting execution. So before, I was doing it manually and now it's all semi-automated with our performance team also reviewing everything, reviewing the KPIs are correct, that roadmap is executed for each team department on the company level. So what I do is obviously product, right? Product direction, product features. And then detecting bullshit as well, right? There can be a lot of cases when everything is rosy, KPIs are good, everything is mesh. But then things can be even better, right? And then I just know it.

• 44:28John Collison?

  What is the bullshit that you're detecting? If the KPIs look good, in what way are they trying to bamboozle you?

• 44:35Speaker 01

  I mean, you need to dive deeper than KPIs, right? Because KPI can be mad because of external factors and because of luck. And then you dive deep and then you look at the product, product features and the speed of development as well. I'll give you an example. For example, wealth and trading product, I think two, three years ago, it grew up 60, 80% year-on-year. And then I thought it's not enough, right? It should have been like 300%. And then I dug deep and what I realized, the quality of product wasn't great. Features were not good. And yeah, then we changed the team and now it's growing 300%. So like you didn't have intuition for these things.

• 45:18John Collison?

  What's the easiest way to annoy you in a product review?

• 45:22Speaker 01

  Incompetency.

• 45:25John Collison?

  What kind of incompetence? There's many forms.

• 45:27Speaker 01

  Well, I mean, we have certain standards for quality, right? And everyone knows the standards. And then if somehow I see product below the standard of quality, and we've done many products so far, it really annoys me.

• 45:43John Collison?

  So shoddy product quality is the...

• 45:45Speaker 01

  Yeah, that we need to kind of, that we're kind of going into the same mistakes.

• 45:48John Collison?

  Yes. How's your AI hedge fund going?

• 45:51Speaker 01

  It's actually going well. It's now, in terms of returns, top 3% to 4% of funds in the world. So model really works.

• 46:00John Collison?

  And what is the model?

• 46:03Speaker 01

  I mean, our model is a machine learning model, so which effectively is trained on a lot of data, on startups. And based on, so it's feature-based model. And then based on our number of features, it effectively tells you which startups to invest.

• 46:22John Collison?

  Oh, it's not a hedge fund, it's a venture firm?

• 46:24Speaker 01

  It's a venture.

• 46:24John Collison?

  It's an AI venture firm?

• 46:25Speaker 01

  Yeah. So they've already invested in 40, 50 deals last two, three years, and the returns are good.

• 46:33John Collison?

  Is the classic problem with this kind of approach, that with private company investing, it's about picking the right companies and then being able to invest in them?

• 46:42Speaker 01

  Exactly.

• 46:43John Collison?

  And sometimes it's like very obviously good companies, but you just can't invest in them. How does, how does that work here?

• 46:52Speaker 01

  So basically, as you correctly noticed, first part is to create a good model, which will tell you with a high precision which startups to invest, which not. And then second, to enter the deal. And then our sales proposition to founders is that, A, we will help them with Revolut operational playbooks. So we tell them how exactly Revolut hires people, how does performance management and so on, so they don't invent things from scratch. That's number one. And then number two, I help with my personal advice. And then number three is we help them to sell either into Revolut or into customers of Revolut as well.

• 47:32John Collison?

  I see. So you're a desirable investor and...

• 47:35Speaker 01

  So added value investor.

• 47:36John Collison?

  Exactly, you can trade the added value for that. When is Revolut going public?

• 47:41Speaker 01

  Well, I hope 20, what's now, 2026, 2028, in two years' time. Subject to market being good.

• 47:48John Collison?

  What needs to be true for everything else to be ready? Or is it just already ready?

• 47:53Speaker 01

  I mean, it's still work in progress, right? We need to launch US, we need to launch a lot of our new expansion markets. We need to still prove that we can be as successful as in Europe, in the markets outside of Europe, especially in credit-led markets.

• 48:06John Collison?

  Do you actually need to launch the US before going public? Like, isn't Revolut a huge bank already with billions of dollars in profits? Like, you're plenty scaled already?

• 48:17Speaker 01

  Of course, I mean, it's not necessary, but it's good to have.

• 48:20John Collison?

  Because investors are based in the United States? And so again, this home country bias means they'll be more interested in Revolut if you guys, if they're familiar with the product?

• 48:30Speaker 01

  I mean, I'm not really optimizing on valuation or readiness for API. It's more like, okay, this is the roadmap that we want to complete, and then let's focus on the roadmap.

• 48:39John Collison?

  You mentioned helping startups understand the Revolut playbooks. Where do you think the Revolut playbook is most differentiated? Like, you talked about performance management, you talked about a few things.

• 48:50Speaker 01

  Yeah, obviously hiring, performance management, like product building as well. I think with AI, it might be irrelevant in the last two, three years.

• 48:59John Collison?

  How do you do performance management differently?

• 49:01Speaker 01

  I think it's very important to do it quarterly, because the majority of companies that do it once a year, I think it's too long. Problem is, a lot of managers, they're inherently biased against, kind of, you know, exiting underperformance. So by having a process once a year, you're not really helping the company, right? Because, okay, then you're forcing managers to kind of, you know, exit underperformance once a year.

• 49:29John Collison?

  Yes.

• 49:29Speaker 01

  Really, you need to do it as kind of, you know, fast as possible, right? You cannot really do performance management once every month, because it's kind of too time-consuming. But once every quarter, you can.

• 49:41John Collison?

  And do you try and make performance more quantitative than other companies? Like, the mapping products to where they should be on the cohort curve is maybe one example of that. Do you try to do that in other places?

• 49:54Speaker 01

  Yeah, so we, we are trying to be as quantitative as possible. So every person is assessed on their skills, deliverables. Deliverables are measured, roadmap plus KPIs. And then culture. So all these three aspects are assessed by manager plus functional manager. Plus, there is a performance review on top of it, checking that everything is correct, and then a person is assessed.

• 50:18John Collison?

  But some domains are just fundamentally more quantifiable than others, right? Like, a salesperson, you can measure how much they sold. Whereas a lawyer, like, you can only make that so quantifiable, right?

• 50:31Speaker 01

  Well, in our case, we have metrics for almost each lawyer, I think. I mean, definitely for the team of lawyers, because what lawyers do, for example, if you look at commercial lawyers and their process contracts, right? And then are they keeping what their SLAs? That's number one metric. How many deficiencies are within the contracts? Again, you can measure it and so on. So you, I mean, on the team level, we definitely have KPIs. On individual level, I'm not sure, but it's possible to do.

• 51:02John Collison?

  So you think most companies are... Because I don't think most companies really believe that you can measure lawyers on a grid of KPIs. And so do you think that, again, you can be more quantitative than you think?

• 51:17Speaker 01

  It can definitely be more quantitative than the majority of companies think. Can I quantify everything? No. So you need to have, like, you know, some other leading indicators rather than some KPIs.

• 51:28John Collison?

  Okay, we'll leave it there. Thank you, Nick.

• 51:32Speaker 01

  Thank you.