March 30, 2022
Besides the impact on people's everyday life, COVID also forced a radical shift in paradigm for the payments industry. Companies had to quickly digitalize their business, and e-commerce, along with online payments, experienced a big boost. Did that boost also translate to an increase in payment fraud? In this talk, Rogier van der Voort (CEO of PCN and podcast host at In Check with FinTech) covers this and other fraud related topics with João Moura and Nathan Troussdell (Fraudio's CEO and COO, respectively). Enjoy:
00:00:01 - Rogier Rouppe van der Voort (RRV): Payments is an industry that has incredibly wide moat. Throughout my career, I've evolved with payments technology. A lot of financial services are changing quite quickly. I always knew I was going to start my own company. Welcome to In Check with Fintech. Welcome everyone, to another episode of In Check with Fintech organized by PCN. My name is Rogier Rouppe van der Voort and on this show we interview people from the fintech space about market trends and developments. Today I'm joined by João Moura and Nathan Troussdell, who met each other at Payvision a couple of years ago and co-founded a company called Fraudio, two years ago. They've built a solution that fights payment fraud, merchant fraud and money laundering, through a SAS based patented AI super brain. They're Amsterdam based, operate globally and are part of the world class payment ecosystem. They've recently raised €2.7 million in Seed Funding and are backed by Salt, as well as several VCs such as Big Start Ventures and Bynd, as well as the likes of Volt, Comply Advantage and Viva Wallet. So without further ado, welcome to the show guys.
00:01:12 - João Moura (JM): Thank you very much, Rogier.
[00:01:14 - RRV]: Great to have you.
00:01:14 - Nathan Troussdell (NT): Yeah, thanks. Great to be here.
[00:01:16 - RRV]: Yeah, likewise. Great to have you as well. It's always exciting to hear the founder story from, especially founders. So what's the founder story of Fraudio?
[00:01:29 - JM]: Yeah, good question. So myself and Nathan worked for an acquirer in the Netherlands, and I was responsible for our data science efforts. Nathan was the Director of Data services. And together we built a very strong team, data team there. And our responsibility was to help the company extract information from data and to build products that improved processing through data and AI. So I was responsible for developing this product that automatically retried failed transactions. And it did so in a smart way, tweaking certain transactions and looking for candidates for retries, which we would then retry when a transaction fails. This was a very successful product. So on average the increasing revenue for our merchants there was above 2% in terms of their volume. And this of course raised a lot of attention and we wanted to roll this out to every merchant over there. And in order to do so, we needed to be in control of fraud. So this acquirer was being served by a couple of large vendors in this space, fraud detection vendors, which we try to use, one we couldn't use, the other one, the results weren't satisfactory, so we developed our own. It turned out to be very good, and one thing led to the other. We decided to spin-off. We brought a lot of data, a lot of codes, and I managed to convince Nathan and three other co-founders to join me and to start Fraudio. This was as you said, this was a couple of years back. Since then we've been developing a product. I have been building the team, building our sales strategy, marketing, et cetera. And here we are.
[00:04:04 - RRV]: So you saw an opportunity basically when you were working at Payvision and you jumped on it.
[00:04:09 - JM]: Exactly.
[00:04:10 - RRV]: I love it. Can you In a nutshell, maybe talk about how Fraudio works. What is it? What does the solution do? How does it work?
[00:04:18 - JM]: Sure. So unlike what I think is really everyone else in this space, when we want to onboard a new customer, what we do is we receive a little bit of their data, and then we look at their data and we translate, or map, their data schema to our own internal data schema. We then put their data together with everyone else's data. We already have billions of transactions. And we then produce AI models, machine learning models, from those transactions, from that centralized data set. This then allows us to build a product that's very strong, that's built from really a lot of data. And we don't need to do this in a custom way for this customer that we're trying to onboard. We simply need to plug, to translate their data and allow the connection to be done, to allow the integration to be done. And then it's really plug-and-play. And from day zero, from the moment that they start sending transactions, we start sending them scores that are relevant, that are very good, from our centralized AI. Off the back of that, and the more data that we receive, we continuously re-train our models and improve them with all the data that we continuously receive. And so, we make sure that the models remain relevant over time. So it's a solution that is very good today, and that in a couple of weeks it's going to be better. In one month it's going to be better. And in the future, it's going to be even better.
[00:06:21 - RRV]: Right. And that's all done in real-time.
[00:06:24 - JM]: That is all done in real-time.
[00:06:27 - RRV]: Cool. So basically, by looking at the data of merchants, I think also PSPs, or at least you're looking at data sets, you can basically identify patterns. And on the back of that, see them, which transactions might be fraudulent or where AML might be going on.
[00:06:42 - JM]: That is correct. So in real time, what we do is the transaction fraud detection. So fraud detection in payments, in real-time payments, but then in batch, offline, twice a day, we produce reports that contain alerts about merchants that are initiating fraud. So think about merchant bust-out fraud. When a merchant processes a list of stolen credit cards or cards that are obtained illegally somehow, and they process those cards, they process payments and receive payments in their own merchant accounts. And then eventually, once the funds clear, they get the money and run away. So this is done offline. And we also, in the same way, using pretty much the same pipelines, internal pipelines, we also look for certain money laundering patterns.
[00:07:55 - NT]: I think it's really interesting how we're getting this perceived in the market by the customers we have and the people we talk to. And one of the biggest frustrations that we faced when we're at Payvision is just how long it took to see value, from the existing providers in the market. And this was really the genesis of the company as well as João's kind of already said. And it's been kind of identified that a lot of companies struggle with making a commitment which Fraud providers to work with because it can be upwards of six months and cost six figures to test something out. And so the way that João was looking at this with this really kind of elegant solution is really revolutionizing how quickly companies can see value from these kind of products. And we see our AI super brain as a platform, really, that companies can plug into to then get access to these three products on the transaction fraud, the merchant initiated fraud, which is fast becoming our star product for payment service providers and acquirers, offers them immense value to kind of grow faster with more confidence and controlling big fraud losses from happening. And then also the money laundering all just from this simple API accessible platform.
[00:09:18 - RRV]: I guess any merchants or payment companies, it's immensely important to have an insight into fraud. Right? And the patterns that come with that. So I can imagine how valuable solutions such as yours is to this guys, or companies.
Sponsored Ad [00:09:32]: We started PCN twelve years ago with a view to serving the fintech community from a growth perspective. Since 2008, PCN has helped household names in fintech as well as the largest global merchants, grow with the best talent who have specific financial technology experience. If you are a VC with a portfolio of fintech businesses, a scale-up looking to hire the best talent, or a merchant looking to hire ahead of payments, or an entire payments team, get in touch today for a no-obligation consultation on how PCN can help you accomplish your hiring goals.
[00:10:12 - RRV]: Great. Let's talk a bit more about those patterns. Obviously, it's hard to miss what has happened over the last 18 months or so. You guys obviously started just before that. Interested to maybe discuss this further on from here during this podcast is kind of how the pandemic has kind of changed the fraud patterns, if you will. Maybe can we start talking about what did fraud patterns look like before the actual pandemic, and how have you seen that progress, or change over those last 18 months?
[00:10:45 - JM]: Sure. So it is known that about 70% of fraud, or at least it is believed that 70% of fraud is initiated by organized groups. And these are groups that control what are called botnets, so networks that operate machines, so computers, cell phones, et cetera, that have been infected with malware and viruses. And then they extract cards from those machines. So when you introduce the cards to pay for something online, those numbers get pushed through these botnets to these organized crime groups. There's also other ways in which these organized crime groups operate, so in a more physical way, via extortion, for instance, via ways that are a little bit even less savory. And that was largely what was going on pre-COVID. So the economy was booming, at least in Europe and in a lot of places in the Americas. And so businesses were goods and fraud was kept at a reasonable level, I would say pre-COVID. Nowadays, and after COVID because a lot of businesses were hit, and quite severely hit by the pandemic, of course - retail was hit with all the lockdowns, et cetera - what we've seen has been a shift towards what is the merchant initiated fraud. So this is when businesses themselves initiate payments with stolen credit cards. And in this case, we've seen both a lot of new merchants being registered with acquirers passing the onboarding system, so being able to pass the KYC processes, KYB processes, start processing transactions payments, and then only after a couple of weeks, one month, six months, chargebacks start coming through, but the money is already not there and there's really no one to go after. We've seen this a lot. And also what is surprising, and I would say a consequence of the pandemic is that a lot of good businesses, so businesses that have been around for years before the pandemic started, had, let's say, the need and the opportunity to start themselves doing this. So processing stolen credit cards in their own accounts. And this has been a trend that we've seen emerging a lot, especially in Latin America but also in Europe. This, of course, is nothing but a mirror of society and the impact that this pandemic is having in the economy and in society.
[00:14:33 - RRV]: That's crazy. Yeah. What does that mean in terms of what these merchants do? All they do is fraud? Some of them at least not good merchants, but the ones who indeed, who pass the KYC past the onboarding with the acquirers?
[00:14:50 - JM]: A lot of them interleave good transactions with bad transactions, which makes it, of course, harder to detect. If you are an organized crime group, you have the capabilities to build online stores, e-stores, and to make them look very real and even make them look very real to customers. So they effectively build businesses that are half-viable. But then, of course, what their intention is, is to actually process these lists of stolen credit cards. It's a lot easier, if you think about it. If you go on the dark Web and you buy a list of 10,000 stolen credit cards or if you operate one of those botnet networks and you get those credit cards yourself, it's a lot easier to have your own merchant account and to process it and to receive the money directly than it is to buy digital goods from a website, a voucher, or even physical goods from a website and then try to move them in a market, try to sell them. And because of this, how easy it is nowadays to open a merchant account, acquirers are being heavily exposed. Also, what before was considered to be high-risk and low-risk is, in my opinion, almost getting reversed by these new trends. And the reason for that is simple. If you are in a traditionally considered to be a high-risk segments, namely gambling, adults, digital goods, etc, you are very heavily scrutinized these days when opening a new merchant account. So you need to go through a pretty grueling KYB process, KYC process. There's a lot of checks against sanctions lists, et cetera. And so it is difficult if you are a fraudster, it is difficult for you to get one of those accounts up and running. And then of course there is fraud there because it is easy to get, if you have those goods, it is easy then to sell them in the market. And so it's a valuable way to extract money from credit cards. However, it is a lot easier to pass the KYB and KYC processes if you are in what's considered to be the low-risk segments. So mainly retail, etc, etc. And because of this, a lot of acquirers that have been focusing on low-risk are now struggling with fraud and have massive losses when it comes to fraud. And that's really where we can help, because we can help them continue to onboard easily, new merchants, and we help them then control fraud while it's happening and shut down these fraudulent merchants often many weeks before they would otherwise be closed and hence save these acquirers a lot of money.
[00:18:35 - RRV]: I can imagine. It's very interesting.
[00:18:37 - NT]: Yeah, exactly. It's amazing because often these PSPs and acquirers, once the company is onboardded, they'll have pretty static rules. Right? And they're really focusing on the front door. But as João's says, there's a lot that they may be processing for a year, two, three years, and then due to some circumstance, they then move towards being at least partially fraudulent. Maybe their profit margins have been completely eroded. So they just go, "all right, I'll just do this for a while". So, if you've got 100,000, 200,000 merchants like some of our customers do, you just can't do this even with a really big fraud team of, you know, tens and tens of analysts. So allowing our AI super brain to kind of monitor and prioritize investigations, just allows them to be so much more efficient in getting cases early as João's says. And that can be the difference of hundreds of thousands of euros saved and also just being far more efficient. You can't scale a team of people easily and fraud often happens in bursts. Someone will find a vulnerability and they'll really go for it if they feel like they're not being detected. And so your fraud cases might double or triple, but you've only got 30 people who are all working really hard doing their best, and they just can't handle that kind of caseload. So we feel like it's a really strong value proposition that we have to really help these companies control fraud but also just have healthy environment for their team and be able to kind of grow in an organized and controlled way.
[00:20:15 - JM]: So maybe to quantify this, so that people have a feeling for how much this is happening. So customers of ours who are onboarding thousands of merchants per month, actually, about 3% of the accounts that are being opened are problematic.
[00:20:41 - RRV]: That's quite a lot.
[00:20:42 - JM]: That is quite a lot.
[00:20:44 - RRV]: Yeah, exactly. That's very interesting. Now I was going to say, Nathan, on your point, what Fraudio does, it sounds like you guys are automating a lot of manual work, right. So you're way better at scaling, better able at getting insights into transaction fraud patterns, that will make better decisions on the back of that and be able to go to the team.
[00:21:07 - NT]: Yeah, exactly. It's not to replace humans, right? It's to allow a lot of simple processes or simple decision-making that is being done in a pretty erranous way, being done in a far more sophisticated way that humans really struggle with. And then that being a toolkit that is given to fraud teams and companies, to allow them to make better decisions faster and really prioritize them. So just being way more efficient. So it's really a hugely additive value proposition to kind of an existing big and well-funded industry with a lot of experts who've been doing this for a long, long time, with pretty feeble toolkits.
[00:21:49 - RRV]: Exactly. Yeah, I can imagine. So is this merchant initiated fraud? Is that kind of the latest trend and something you see the most now when it comes to fraud, AML, or is there other stuff you see even more at the moment?
[00:22:03 - JM]: So the typical transaction fraud or payment fraud remains in place. So this is another tool that we have to help stop transaction fraud. So at the end of the day, what we're stopping is transaction fraud, only over a longer period. So offline. This is something that I would say is new or at least it wasn't as prominent as it is nowadays. From what we can see in the market, that's really how the market changed the most. All other forms of fraud remain in place. For instance, Booking has those customer schemes, what's it called?
[00:22:57 - NT]: You mean the loyalty fraud?
[00:22:58 - JM]: Loyalty, yeah.
[00:22:59 - RRV]: What is that?
[00:23:01 - JM]: So loyalty fraud is when you have you're a customer of a company and by doing something, by buying something or by enrolling in some program, you receive credits and then those credits are actually valuable one way or the other. For instance, in Booking.com, you can go online, you can register a property and then you can register also accounts, travelers' accounts. And then when you spend money in a property, you get money back. If you then refer someone else, that person also receives money. So if you register property and then you register a lot of accounts and buy or pay to stay in that property, that's actually yours, you are paying the platform fee, a share, their hosting fee, but then the platform is giving you credits. And because you are receiving the money that you are paying for, the money that you receive is actually higher than the fees that you have to pay to the platform. And so you're making money. So this is just one example. There's loads of these scams almost, but we really focus on payment fraud.
Sponsored Ad [00:24:37]: Berlin, we're here and ready for your hiring needs. After some short time considering it, we decided to set up business in Germany, meaning we can be closer to clients and allow room for new business. We are set up and ready to help find your ideal candidates, help build teams and offer up media services. People create networks.
[00:24:58 - RRV]: Yeah, exactly. That makes sense. All right, so back to that merchant initiated fraud then. Do you see you just talk about Europe and LATAM, but maybe also other parts of the world. Do you see many differences? Is it in some parts of the world or regions where it's more than in others?
[00:25:14 - JM]: Not really. And that's also somewhat surprising. So in LATAM, definitely the fraud ratios are higher than in Europe, but the merchant initiated fraud is similar. So the ratios appear to be similar. And we process for a lot of merchants in LATAM and in Europe, more in Europe than in LATAM, but we do see the ratios as being similar. And that is probably not only a consequence of the Pandemic itself, but also a consequence of the automated KYB and KYC platforms being the same platforms being used here and there. And so the Scammers and the fraudsters already know how to bypass those. And one thing sort of leads to another.
[00:26:16 - RRV]: Yeah. They might potentially maybe even come from one region, but they set up businesses across the world. Or would you say that people would be local as well? It's just something, it's a leap hole that is found, basically, and is shared, and then people just start to put it into practice.
[00:26:36 - JM]: So we don't know, okay? To be honest, we don't know exactly who's behind it. We know that 70% of payment fraud comes from organized crime. When it comes to merchants initiated fraud, figures are still not out there. It's a new trend, so it's difficult to know. But I would say that if it is, if that ratio remains the same, then there will certainly be international organized crime networks that operate in multiple countries in the same region or even across different regions. But there will also be a lot of small, local, more opportunistic type of fraudsters.
[00:27:24 - NT]: Yeah, and to add to that, fraud-as-a-service is a relatively new thing, with the Dark Web and also with cryptos being used for payments, for a lot of nefarious activity, I think they've become a lot more brazen. So again, who's behind these groups? We don't really know, but they are literally kind of fraud-as-a-service operations that say, "look, you pay us X amount of money and we'll get you X amount of vulnerabilities, or we'll get you X amount of kind of credit cards". And so people can kind of buy modular parts of a fraud strategy. And look, these are sophisticated groups, right? And unless you're fighting and using data science and artificial intelligence, it's pretty impossible to win. Because they are using those techniques, they can afford to pay data scientists. They can afford to try and backsolve all the issuer rules and take the time to do it because the sums that they can walk away with are huge. And they can have a certain amount of anonymity, within the dark web and within crypto payments, even though in the long run, the crypto chain can be unwound to really find out who's doing this, you've kind of seen that recently with some of these ransom attacks on big institutions. A lot of the funds have been recovered later on. But these are highly sophisticated networks that are pretty global because the payment company is very global. And when they enter a new market they're trying to take in the same easy onboarding process that they've replicated somewhere else, it allows for these vulnerabilities as well.
[00:29:07 - RRV]: Yeah, exactly. They're continuously developing, right, these fraudsters and service as well. So that makes sense.
[00:29:14 - NT]: Yeah, absolutely. So you've really got to try and stay one step ahead. So you've got to be innovating in this space, otherwise they'll be innovating ahead of you.
[00:29:21 - RRV]: Yeah, exactly. So how do you guys fight this? I understand that you see the transaction, you flag it the merchant or the payment processor in this case takes action. Do you go beyond that? Is there more that you do or is there stuff in between as you do as well?
[00:29:41 - JM]: Yeah, there is something else that we do. So we do block the transactions in real time, but we also send offline reports or alerts to start investigations. And we send those alerts with explanations for why we're raising those alerts and also with links to sets of transactions that triggered those alerts. So this effectively allows for a hybrid strategy, automated and manual, to be put in place. So when we send a black alert, our highest level of alert, we effectively are suggesting, or are telling the acquirer or the payment service provider, to shut down that merchant. Effectively, we believe that that merchant is a fraudster for sure. Shut it down. And if it's red, we suggest to initiate an investigation. And this is started by the acquirers fraud investigations' team. But of course, they started an investigation in a very targeted way, so we tell them why, and what they should be looking for. Effectively this cuts down the amount of time that the investigation takes, by a lot. And let's say, the precision, how accurate these reports and alerts are, is actually quite high. So the accuracy is good. The fraud investigations' teams won't waste too much time looking at merchants that are actually, either being a victim of fraud themselves and not initiating fraud, or just false positives.
[00:31:47 - RRV]: Yeah, exactly. That makes sense. So you basically are an advisor to these merchants, acquirers, PSPs, correct?
[00:31:54 - JM]: An automated advisor.
[00:31:56 - RRV]: Indeed, yeah,. Great. Indeed. Like Nathan just said, you have to be one step ahead. You guys just did your first Seed Round. So the bridge I'm trying to make is, what is next for you guys? What is next for Fraudio? What are you working on?
[00:32:12 - JM]: So our goal is to now build or increase the capacity of our product team, to expand our sales team as well and our marketing teams, and to really start expanding to the Americas first. So we're making a big push in Latin America, starting to look at the US. This is, let's say, building the company up until the beginning of next year by when we plan to do a series A.
[00:32:57 - RRV]: That makes sense, all right, okay. So from now on, it's in the expansion of the team, expand internationally. And then in the next year, you go into the next phase.
[00:33:06 - NT]: Yeah, exactly. And really looking to with our next series A round, partner with people in the geographies that we're going to have a physical presence, really use that as a launching pad, getting into those deep local networks. There are tens and tens of thousands of potential customers for us, in very big markets. And in general, the Fraud Detection space is growing at a very rapid rate, with the tailwind of COVID making digital payments just exploding in so many different markets. And so we feel like we've got good winds behind us and just really executing at the moment, as João says on hiring a product and some sales and marketing and set ourselves up for an even bigger and better 2022.
[00:33:53 - RRV]: Very exciting. If people want to find out more about you guys, if people want to keep track of your journey. Where should they go? Where can they look?
[00:34:03 - JM]: They can go to our website, www.fraudio.com, or also just go on LinkedIn, look us up and ask us for an introduction.
[00:34:15 - RRV]: Great. All right. Okay. That sounds good.
[00:34:19 - NT]: Yeah. And if you're a PSP, or acquirer, or an issuer, you can do a free Proof-of-Results with us. Instead of doing these long six month POCs that cost six figures. What we offer is send us a snippet of data, and in a week or less, we'll turn around and show them in a real world scenario, actual data, how our models would perform, and we do that free of charge. So the proof is in the pudding, as they say. They can see our results straight away and then make a decision to move forward or not with us.
[00:34:50 - RRV]: Great. That's a rather unique proposal. It just shows how confident you are about your solution, I would say.
[00:34:56 - NT]: Yeah, we really are. We really believe in that meritocracy of the product. We wouldn't be doing this if we didn't believe in it and spun-off when we were up against some really big companies and we believe we have a better proposition for the customer, much faster to get and definitely at the cutting-edge.
[00:35:15 - RRV]: Great, guys. All right. Okay. Well, thanks for being on the show, both to share your insights obviously into Fraudio, but also about what's going on within the payments or within the fraud market. It's been great to have you and thanks everyone for listening and tuning into another episode of In Check with Fintech. If you haven't done so already, don't forget to subscribe to us on Spotify, Apple, Itunes or any of your other favorite platforms that you use, and don't forget to tune in next time. Thanks very much.
[00:35:46 - NT]: Thanks for having us.
[00:35:49 - RRV]: Thanks for listening.
Hope you've enjoyed our talk between João Moura, Nathan Troussdell and Rogier Rouppe van der Voort, about the change in fraud patterns during, and after COVID. See you on our next chat!
Fraudio is an Amsterdam based scale-up helping companies in the payment ecosystem fight payment fraud and financial crime with its unique ability to build high performing AI and ML models without costly customisation. It is trusted by some of the fastest-growing companies in the world, protecting them from payment fraud, merchant-initiated fraud and money laundering.
It's easy to integrate with products that deliver best-in-class fraud detection from day 1, allowing clients to scale their customer bases safely, reducing operational costs and fraud losses while maximising revenue. Fraudio's founders are from the payments industry and don't believe in black-box solutions. They ensure that end-users are provided with insightful and timely information to control payment fraud and merchant portfolio risk while ensuring the highest level of security and auditability.
How about trying our solution and experiencing the next generation for yourself?