Tackling financial fraud with AI: Qonto CEO Alexandre Prot

0:02Hello and welcome to the Sifted podcast, the show where we help you get to know the brightest and boldest people and companies in Europe's startup ecosystem. I'm Amy, Sifted's editor and your host. And today I'm joined by Alexander Pro, co-founder and CEO of Quanto, the French fintech unicorn. Quanto banks 500 ,000 small business customers in eight markets across Europe, helping them do everything from banking to accountancy, lending, savings, and as of recently payment terminals. The company is profitable and has 1 ,700 employees across five offices. Its goal is to hit 2 million customers by 2030 in many more countries with a bigger suite of tools and services working with bigger and bigger businesses.

0:48Today, we're going to talk about how Alex plans to make that happen. Welcome to the show. Hi, Amy. Thanks for the welcome and Thanks for having me. We saw each other briefly in Berlin in May. I know you've also been at Money 2020 in Amsterdam since then too, and obviously VivaTech in Paris. Why have you been hitting the road so much? Yeah, it's true. I've been traveling and doing quite a few events lately. I think it's very much linked to our European expansion and the fact that we are and want to be also more and more visible. We're also hiring across the different markets, having partners in those different markets.

1:26So I'm combining, let's say, these external events with also meeting, you know, our teams internally, meeting potential partners, potential new, you know, people that can join the team. So trying to do that efficiently, but indeed, pretty often on the road these days. So let's start talking about acquisitions. You're one of, I guess, a handful of European tech companies, which are still relatively young in the grand scheme of things, but in the interesting position of being able to and interested in acquiring other businesses. What are your plans around that? So maybe before we talk about the plans, just a quick reminder of what we've done in the past.

2:05We've mostly grown organically. So you were mentioning 500 ,000 SMB customers across Europe. And so most of these SMB customers we've acquired organically, one by one, basically, and opening their business account with us and doing more more stuff with us. But we also made two acquisitions over the years. The first acquisition we made was three years ago. We acquired Penta. Penta was a small, let's say, conto focused on Germany. So much smaller than conto, but much stronger in Germany. And as we wanted to be much stronger in Germany, it felt like a very good acquisition for us. And three years later, we're very happy with that first acquisition.

2:43We acquired Penta when they had about 50 ,000 customers. We migrated all these customers to our Conto platform. We integrated the whole team. And now the Penta product and platform have been discontinued, but we're much stronger in Germany. It's a very much geographical expansion acquisition. Then the second acquisition we made was a year ago, a French fintech called Regate, basically for product expansion, doing more sophisticated financial tools for SMEs. But also Regate is selling directly and serving directly accountants which is a slightly new channel for us. Accountants have always been big partners and stakeholders of the Conto ecosystem, but now we're also serving them directly, strengthening even more our finance management platform for SMEs.

3:29So one geo-expansion M &A, one more product or channel expansion M &A. And so now to answer your question about the future, these are two dimensions where we see opportunities for us in the future to be stronger on the geo-expansion, go faster in some of our existing markets or maybe launching a new market through an acquisition. And same goes for the product expansion. We have a pretty big and strong team on the product and engineering side. About 40 % of our total team is on product and tech. But still, there are some great teams out there, great products that have been built. And so if there are some opportunities for us to go faster, to bring an even better product to our customers, we'd be happy to seize them.

4:10Of course, we have to be careful because buying a company is pretty easy. The tough part is actually to integrate and to do that well. And we know that sometimes, you know, it slows down the whole company. Sometimes you break companies when you acquire them. So it's a mix of art and science. And of course, it's not easy. So we have to be pretty careful, not only about, you know, how much we pay, but more about how we process the whole thing, how we align with the existing team and so on. But we believe, and we've seen that successfully in those two first deals. So we believe that we can go even faster on our goals through acquisitions.

4:44How would you mark your homework? How well did those first acquisitions go? What could have been improved for the next time? So I think on the first one, the easier part is that it was exactly the same business model, exactly the same product as we had, but they were just stronger and bigger in Germany. and it was pretty clear that after a year or so, we would sunset the Penta product and in the meantime, we would basically improve our product for Germany. So the Conto product for Germany and migrate all customers on the Conto product. So from an operations perspective, it was not easy. From a product perspective, it was not easy, but from a strategic standpoint, it was pretty clear and the direction was pretty clear.

5:30And so I think that was an execution challenge But the idea, let's say, was pretty simple. On the Regate acquisition, which is integrating two products, it's by definition slightly more complex. And also this new customer that we're now serving directly with accountants, it's also a channel that we knew, but we're not serving directly. So we're also learning as we're growing on that channel. And so it's more like it's more unknown. So it's not directly linked to the acquisition or the integration, but more that we're entering, you know, something which is less known to us. So we're learning more, we're discovering more, and we're also making some mistakes in that channel, which is new for us.

6:15But overall, very happy with the ecosystem that we're building, because in the end, as we want to make it much easier for any small business owner in Europe to get started, to grow, to manage their business. Of course, accountants are a very, very important stakeholder. And so having this directly into accountants and making it smoother from banking to accounting, you know, all the way is, of course, super important. So we have very good feedback from customers, both accountants and SMEs. But it's just the execution that is slightly different from just applying the same playbook as we have done, you know, kind of forever.

6:51and how what's been your approach in terms of obviously when you acquire those companies you often bring on board the founders and their teams and then naturally what often happens is those people sort of see out a certain term and then they decide to leave and I believe that's what's happened with Raygate's founders reasonably recently how do you approach that in terms of what is best for Quanto in terms of how long that original team sticks around for like what role do they need to play? And then when is it maybe the right time for some of those people to leave? Well, generally, in most deals, you often have this kind of two-year, I think it's kind of the market standard, you know, two-year, either it's an earn out or bonus or integration contract, whatever it's called exactly.

7:39But I think this, you know, 24 months on Penta, this is more or less what happened. So one of their co-founders who was leading Penta at the time, Lucas, stayed for about two years and left when, you know, we all expected that, you know, he would leave at that point. And I think the whole integration transition had been made, which was good. And we had, you know, someone new running Germany and so on. So that was very expected. And I think very smooth transition on the Raygate acquisition, Laura and Alexei, the two co-founders left more or less about a year after the acquisition, things were, you know, well underway.

8:15And it also makes sense for kind of, you know, to strengthen the real integration. You also have, it's hard to have, you know, two cooks, let's say in the kitchen or three cooks. And so that's pretty efficient that, you know, when there is no clear goal and need or role for the, you know, the founders that have been acquired that they leave. And so for your next acquisitions, because I believe you're on the hunt, we've obviously spoken about geography and product being some particulars. Are there particular geographies you'd be interested in acquiring businesses in or particular areas of the product that Quanto would like to strengthen or isn't doing yet that are obvious areas for you to look at?

9:00Yeah, well, on the geos, I mean, our key geographies today are, the four markets we launched initially. So France, Germany, Italy, and Spain. More recently, we launched the Netherlands, Belgium, Austria, and Portugal. So frankly, any of these eight markets could be interesting. Given our positions and our growth in France, we're rather looking at the other GOs to be growing even faster outside of France, but still very much focused on the European Union and actually just for compliance and license reasons. And so most probably in some of our eight markets. So the focus right now is really on these eight markets, not even expanding further today because we just launched four new markets.

9:42So it could be any of these. On the product piece, it's mostly around what we call financial tools. So Conto is really very much the banking at the core. That's what we have at the core of our finance management platform. And around that, we have financial tools, which are all the adjacent topics that are very important for small businesses, such as accounts payable, accounts receivable, cash flow management, expense and spend management. Invoicing is part of accounts payable and accounts receivable, but it's also very important topic with e-invoicing. So tax compliant invoicing become the norm across the European Union.

10:18And so on these topics that we started building after the core banking piece, we have built nice products, but we know we can and we want to do much more. So there might be some good additions there, either by having small teams that have been building something really nice there, or maybe larger products that could be a good complement. But we don't depend, let's say, these acquisitions. We more have like a strategy to build things mostly ourselves, but we also have, if I may say, the luxury because we've been profitable for two years and we have, you know, some cash available to accelerate if we have some good opportunities.

10:53But we're not forcing ourselves to do M &A either, precisely because we know it's tough to do it well. It's tough to have find exactly the right fits. And so we see our growth over the next, you know, five years, I would say, I don't know, 80 % or so organic and maybe 20 % through M &A. And Quanto is actually licensed as a payment institution, isn't it? Rather than a fully fledged bank. So you have to protect customers' money by separating it from your own payment flow and you can't offer direct lending services. Is that something you plan to change at some point in the future? Do you plan to become a fully licensed bank?

11:32Yeah, maybe just before answering your question, what you said is perfectly right, except that with the license that we have, which is indeed a payment institution license, we can still do a bit of financing. And the way we do it is by actually offering financing, leveraging our own funds, our own equity. And so we launched about 18 months ago, a product called PayLature, where we can finance invoices up to 30 ,000 euro. and our customers, if they have, I don't know, to buy computers or laptops for 10 ,000 euro, we can basically finance those invoices for them and they repay us within three, six or nine installments.

12:10And so that's already a significant financing offer that a lot of our customers are already using, but it's a short-term financing. So it's less than 12 months and it's financed with our own funds. So obviously pretty different from what a fully licensed bank could be doing because they would be leveraging, you know, customer deposits to finance credit lines to other customers. And so to get back to your question, this discussion around, should we keep our payment institutional license? Should we move to a banking license? We've had that in the past. A few times we've checked that, we checked, you know, we weighed the pros and cons basically.

12:42And the cons were more important than the pros, especially because you needed to have a very important capital requirements. And at a time when we were not profitable, that was seen as, more of a stretch and less obviously a good idea for us to do that. Also, we're super busy building the core banking piece, the financial tools, and also getting started with some financing activity like the Paylature product I mentioned. Now that we're profitable, that we've built already quite a lot of things, and we could potentially become a bank as an actual next step for our customers, that's something on our minds.

13:17And that could happen in the future. We haven't made the decision yet, but that may happen in the future. And what would that unlock if you were to become a licensed bank? What kind of things would that enable you to do? Well, I think three main things from a customer perspective. The first one is indeed more financing options. So not only short term, not only up to 30 ,000, 50 ,000 euros, but it could be longer term financing, higher volumes and so on. So that's definitely one. The second one is on other products such as savings, investment. So basically the full suite of products that traditional banks can offer and being even more replacing traditional banks.

14:02A super important number on Conto is that 78 % of our customers use Conto as their primary account. And so, of course, if we want that to further increase, and also as we want to go up market, slightly up market, serving companies with not only 5, 10, 20 employees, but also 30, 50, and even more, of course, offering more products, including financing, savings, investments, so on, is important. And the third piece is that today we're still partnering with quite a few traditional banks. And as we grow, being more independent from these partners would probably make sense to be more agile, but also have even more our destiny, let's say, in our hands.

14:42Very recently, we built our own card processor. So all the card transactions that are made with our Conto cards are processed internally, which increased card acceptance rates, which improves product velocity on card products. And so getting a banking license would, you know, in the same way, also enables us to do more internally to own more of the value chain, let's say, and to be, you know, not potentially dependent on some banking partner's speed. And banking partner speed is, you know, know, a bit of an oxymoron. And so probably that would help on that side as well. So is this something we could see in the next year, an application for a bank license?

15:22Yes. And where would make sense to have it? In France? Well, we're, you know, European and European Union, that's our field. And so, yes, it would probably be in France, given the current license we have is in France, payment institution license, passport across the different countries, but, you know, it would make sense. And we have pretty good relationship history with the French regulator. They saw us, you know, from the beginning. They saw us go from, let's say, zero to 500 ,000 customers with all the different steps. They know us pretty well in terms of governance, all the things we've implemented.

15:59So it would be the natural regulator for us. To turn to money of a different kind, in November last year, the Financial Times reported that Quanto was talking to investors about a possible sale of shares at a new 5 billion euro valuation, which is obviously something we've seen quite a lot of scale ups do recently to try and give early investors and in some instances employees a chance to get liquidity. What was the motivation behind that on your part? So indeed, we're doing, you know, let's say smaller secondary round for liquidity for employees, because obviously getting stock options is great, but it's even better if you can, you know, exercise them and, you know, and get the real benefit of those stock options, you know, over time.

16:47We're all talking about, you know, in three, five, 10 years, you know, potentially larger, or let's say real exits kind of, and maybe we'll have a discussion around that later, but it's of course also good to have some partial liquidity. And so that's what we're offering to our employees. But the other aspect, which I think those articles were mentioning, was around getting also new investors on board for the next stages of growth. And this is something we'll probably be doing next year. I think we're now very much focused on operations, execution, our geographical expansion that we mentioned.

17:21And so when we have better visibility also on how we want to further expand to your question around the banking license, what markets, potentially also, you know, IPO timeline. This is not a short term goal. But of course, this is also something that, you know, investors are asking. And so we want to be getting clear answers to all that and then probably doing some larger secondary. Interesting. Are you also going to be out raising, you know, a next equity round? I mean, you hinted there that applying for a bank license isn't a cheap process. I guess my question is, you're profitable at the moment.

18:00In what situation would it make sense for you to go out and raise new funding from investors? So indeed, we're profitable. We don't intend to be raising additional capital. If we were to become a bank, we would probably not need to raise additional capital. So the main or the only reason why we could rate additional capital is if we do, you know, large or very large M &A deal paid, you know, mostly in cash. And then we would, you know, we could need more cash. But let's say for the, you know, business as usual, we won't need any additional funding. So you're not currently fundraising? No. and you you strongly hinted there about a potential ipo what are your what are your thoughts and plans around that actually i was more doing the journalist with myself so i was doing kind of the question and the answer because it's a question that you know we very often get feels a bit like the buzzword and maybe especially in europe where unfortunately we see too few ipos and so no the very very simply put that's definitely something that we're getting ready for in the sense that it might happen one day and it takes a few years to get ready.

19:13Also, because given we are a financial services company, we have a license today, a payment institution license, potentially one day a full banking license. And these licenses require having your books in order, having your governance in order, having your internal controls in order. All these topics are very similar to what is required to be a public company. So it's not like it's, you know, kind of additional work for us. And so indeed, we don't know what might happen, you know, three, five years from now, it might be an IPO. And so we're just getting, let's say, slowly but surely ready for that.

19:50And that was more what I meant. But there is no short term plans on IPOing. You mentioned, I think, in the introduction that we have half a million SME customers today. We're targeting 2 million customers in 2030. So that's really the main goal today. And on that journey, we might have some opportunities to do some M &A, potentially to IPO, but we're not spending too much time on that. We're really focused on growing the business, making our customers still super happy as they are today. As we can see with this 4.8 star rating on Trustpilot and quite a lot of other metrics going in the same direction.

20:26But that's really the key focus for us today. Are the various stock markets beginning to court you? yes and no i mean obviously when people talk about ipo there is also like the big question of you know ipo venue and and you know what what index and so on we're not too interested in that today so we get we get some emails but most of the time we actually ignore them so it sounds to me like quantos in a in a pretty good position you know you're profitable which a lot of a lot of startups aren't you've got you've got money to finance acquisitions what could still trip you up what are the the challenges that remain ahead well challenges are to keep on you know innovating being fast and agile while at the same time you know being robust being serious on you know compliance on internal controls on governance and these things tend to be a bit let's say contradictory somehow.

21:29And so we're getting the team organized and the company organized so that some people, let's say, have this kind of serious aspect and robust aspect as their major, and maybe their minor is around innovation and so on, where other people are much more on innovation, cracking things, testing things, because even if they don't work or even if there is a problem, that's not a big issue. And so if you work in compliance or if you work in fraud or if you'd rather be on the, let's say, making sure it works well before you launch, if you're on the pure software side or on the, then it's good to be trying to be making some mistakes to launch in beta and to see what happens.

22:12If you're doing some fraud stuff, you probably not want to launch too much in beta. That's kind of a bit of the spirit. So of course, it's a bit of a simplistic answer, but that's really the point. And same for the markets. France is our largest markets, historical markets. So of course, what we do in France is pretty different from what we do in our most recent markets, like Netherlands, for instance, Netherlands, we just launched a few months ago, we have a small, very small team, 10, 15 people working on Netherlands, but that's way enough. That's kind of a small startup. So that's way enough for the launch, to take off there, to be agile.

22:49They can have, you know, a daily kickoff meeting and, you know, be agile, see some good progress, you know, day after day and so on. Of course, it's a bit different in a market where you're pretty mature and where you already have, you know, hundreds of thousands of customers. And it feels, you mentioned fraud there, it feels to me that fintechs in particular are especially susceptible to sort of frauds and fakes. And we're always seeing headlines that like, you know, So X digital bank has higher rates of complaint or fraud or whatever than high street banks. What's Qwanto doing to stay kind of one step ahead of the criminals?

23:26That's a very good question. So first, I think seeing, you know, the industry numbers, I think we're not having higher fraud rate with digital banks or neobanks or challenger banks, whatever the word versus high street banks. I think it's more than for high street banks, a good chunk of their customers do not even do anything online. So of course, there is less fraud, but it's more like, are customers doing things online or not? And if they are, I think the risk might be even lower with more digital products and companies that have the right tools to basically stop fraudsters. Just to give one very concrete example, I think very early on, we used IP addresses and MAC addresses, like smartphone, basically serial numbers to prevent fraud.

24:11and we discussed those with the French authorities and they were surprised that indeed not that many traditional banks were using those. So I think we can be at the forefront of innovation for UX, I mean, user experience, but also for fraud topics. And indeed, I mean, banking and payments is a lot about making sure the 99.9 % of cases that are clean go well and smoothly and fast and so on. and the 0.1 % that are not clean are blocked instantly. And so that's the challenge. And so whenever we launch a new feature, meaning a payment feature, whether it's international transfer, local transfer, the direct debits, whatever, of course, there is this question of, okay, how do we make sure that these 99 % or 99.9 % of customers or of transactions go through very smoothly, very fast, et cetera, et cetera, but we also block the other ones.

25:07And so sometimes you have some false positives, some false negatives, and the whole game is to lower the ratio of false positives and false negatives. And so technology is very much leveraged through that. We're actually using a lot of gen AI and NLMs to basically improve our fraud algorithm. The other day I was actually spending, I was very curious to understand how it was working. So I spent one hour with the team working on that. And basically it's, they train the computer by explaining, you know, what a fraud could look like. Like, I don't know, if you have, you know, two payments with the same card, like physical payments at two, you know, distant locations within a short timeframe, of course, there is a problem, you know, and training the algorithm with those things.

25:51And then the basically the model detects, you know, frauds, and then suggest them to analysts that will check them and, you know, will clear them if no problem. So we use a lot of technology there. And that's also why I was saying that, you know, new offers like Quanto and, you know, basically new fintech can do very often a much better job than high street banks. How else are you using AI at Quanto? What experiments have worked well and which ones have not worked so well so far? Well, we started using, let's say, AI a bit before it was called AI. So we've used, you know, in our operations on the, you know, customer facing topics like the chatbot that we've been using for a couple of years now.

26:34And so, of course, now we implemented some AI logic and Gen AI logic into that. So today, 54 % of our chat interactions with our customers are handled by our Gen AI chatbots. What's very interesting is that the customer satisfaction, maybe it's a bit sad, but the customer satisfaction rates with the Gen AI chatbot is as high as with the part that is handled by human agents. There's a little bit of a selection bias, because usually the human agent gets to slightly more complex cases. But still, more than half of the chat messages are handled by AI. So that's very important. And that enables us to be really good at customer care.

27:13We've always been, but even better today. So 24-7 customer care in all the languages that we support across Europe. And so AI is helping us do that. It's, of course, also helping the agents to be faster and better when they reply. So on the human parts, they also use AI, let's say, as a co-pilot. So that's one very direct use of our application of Gen AI. Then, so I mentioned the fraud one, which is an important one. But also, of course, on all our, let's say, internal daily operations, the way our engineers code, I mean, they're using AI now to go much faster. The way our designers are designing their new website pages or whatever, they're using Gen AI there.