Adyen N.V. (ADYEN.AS) Earnings Call Transcript & Summary

Okay. Perfect. I'll set off this afternoon session. I'm Adam Wood, responsible for software and payments research at Morgan Stanley. It's a great pleasure to have the CFO of Adyen with us, Ethan Tandowsky. Ethan, thank you so much for joining us in Barcelona.

Yes. Thanks for having me. Good to see you all.

So I wanted to start off on the acquiring side and start off on the platform story. I think you talked a lot about how the single platform gives you the structural advantage versus legacy players, particularly now as we get into a world where there's new technology emerging that we can lever. You had your Investor Day this week, which I came along to and thought that was really powerful. Could you talk a little bit about how those foundations of your platform are changing? I think you talked about 3 layers. And you talked about behavior-based identity. Could you talk a little bit about how that now plays in the acquiring platform and what you're trying to do?

Yes, sure. So we talked about basically the 3 layers of our foundation of our platform. The first 2 are things we've been building for a long time and dynamic Identification, what you just talked about is what we talked about more recently. So the first 2 are the single global platform. That's essentially that we've built one tech stack completely organically, which allows our customers to process payments all around the world on one technology platform. That's both online transactions and in-person transactions. That's basically something we've built since day 1. That's always been there. In 2017, we started to add banking licenses on top of that. So we got banking licenses in the U.K., Europe and U.S. And not only is that just a licensing structure, which allows you to offer certain products that you wouldn't otherwise be able to offer, but it gives us full end-to-end control. So basically, we are the only party in the chain between, for instance, different payment methods and our customer. And on top of that now, we're getting to scale, right? So we are now at EUR 1.3 trillion in payments volume over the last 12 months. With that scale and with that full end-to-end global control, what we see is that we can tell a lot about individual shoppers, individual behavior based on the behavior we see on our platform. And that's really important now because technology is moving the world faster than ever before, right? One of the ways that, that can be understood is fraud, right? In e-commerce, One of the big challenges is fraud. And the game has always been you're in a race with the fraudster, right? So every time you are adjusting, the fraudster is adjusting and then you're adjusting and the fraudster is adjusting. And it's always important to stay ahead. That's how we can help protect customers. Well, now the pace at which a fraudster moves is so much faster with AI, right? I mean even simple things like if you open a new bank account, you send a picture of yourself with a passport next to your head and you say, that's what gives you the clarity that you are who you say you are. But that's -- in this world, that can be recreated in 10 seconds, right? That's something very easy to recreate with AI. So we want to leverage the behavioral patterns we see on our platform to better inform that a person is who they say they are in any interaction, and that needs to be in real time, right? That cannot just be a static check done a year ago or two years ago. That needs to be real-time behavior to inform decisions for our customers.

Could you -- I think this is really interesting about the data that sits on the platform. I think one of the things we've thought about Gen AI differentiation is the proprietary data that some companies have and obviously, you're sat on this incredible volume. Could you just talk a little bit about the data that you have access to and then how that can help you onboarding merchants, identifying fraud, just go a little bit deeper on what data you're looking at and how that helps?

Yes, sure. So let's stick with the fraud example. There's many uses of the data that we have, but let's start with the fraud example. So we have a lot of data on the platform. Again, the significance of the size of the transaction volume we have. Let's use one example, right? Maybe we gave this example earlier, but maybe to get to our -- we did an Investor Day in Amsterdam. So let's say, somebody came to our Investor Day in Amsterdam, they booked an airline ticket. And they booked a hotel and they booked a rideshare to get from the airport to the event. And maybe on the way, they stopped and got a coffee. Each of those things informs the behavior of that person, which we can use for future transactions. And think about that set of 4 transactions. What's most useful? Well, you get a certain set of data from an airline ticket because you get the actual real name of a person connected to that ticket, but you also get some really interesting data if they bought coffee in person, right, different from if they bought online transactions. If they showed up and bought coffee in person, it's also much more likely to be them than an online transaction that's happening. And so really assigning risk levels to different behavioral patterns is where we see a lot of advantage that we can drive for customers. So we'll be better positioned to help our customers know when somebody is truly who they say they are. And that matters a lot also if you take it into a world of agentic commerce or a lot of the changes that will develop, right? The challenge is going to be, is that person really who they say they are? Is this actually a valid transaction or not a valid transaction? And so much of the ability to do that will be the foundations we've built and the quality of the data set that we have today.

And this is -- it's about going from static to dynamic in real time. And I guess for you, it's the scale that you now have on the platform, but it has to be one platform because if that was fragmented over 10, it wouldn't -- you would...

Exactly. Exactly. Exactly. That's what allows us to connect those transactions to one another.

So we carry on with the example of the fraudsters using AI accelerating their pace of innovation, if we phrase it that way. We see a lot of legacy players now starting to have more and more problems in their tech stack, raising money to solve this. I mean I think we've thought for a while, like at what point is this tipping point that market share starts to accumulate even more quickly to modern players and modern platforms. Is this now a catalyst that you see they start to struggle? Is there any evidence of that in the market yet?

I would say that complexity is rising, right? And in payments, there's always been a discussion on is it a commodity or is it a functionality game, right? We've been talking about that as long as I've been at -- and it's a commodity if complexity is decreasing. And it's a functionality game if complexity is increasing. And I think what you see right now with the way that technology is evolving is that the complexity is increasing, right? Both we talked about fraud, but also the sales channels, like the means of distributing your products. We think that over the next few years, agentic commerce will play a bigger role. That adds a lot of complexity because it's not just going to be one solution. It's going to be implementations of that agentic commerce flow in multiple different setups and multiple different protocols. And so what you see is this kind of fragmentation while technology is moving faster than ever. And that should be a reason we continue to gain a significant -- a larger share of this market in our view.

That makes perfect sense. I think as well, it feels like -- and maybe this is just an external perception. It feels like the innovation has really stepped up a level and that level of differentiation of you from other players has gone again to a different level. One of the things you've talked about is the Adyen uplift product. There was some pretty amazing stats at the Capital Markets Day around how that's helping merchants. But that seems a good example of the innovation. Could you talk a little bit about what that is and then the benefits that's delivering for your merchants?

Yes. So Adyen uplift for us is a suite of products, which ultimately help our customers optimize the full payments flow. We call it the payments conversion funnel. And whereas previously, there was a lot of discussion, especially in e-commerce around authorization rates, and there still is a lot, right? The percentage of transactions which ultimately get approved. That's a big discussion. If you compare it to in-person transactions, 10% more are declined on e-commerce than in-person. So there's still a major problem to be solved there and auth rates are still important. But what we've seen over the last few years is that other metrics have also become important, things like payments costs, right? Each payment method has a different payment cost structure. Things like when to authenticate or tokenize the transaction has become more relevant. How do you manage fraud? And where we're uniquely positioned to drive value is that we can make a decision across each of those dimensions together. So not in isolation, not optimizing for one specific component, but truly driving optimizations across the full funnel. And I can give one example maybe to exemplify it. So we've put a lot of effort into building out our U.S. debit offering. In the U.S., if you use a debit card, there's regulation, which requires that there's the option to send that transaction to multiple networks. So that means if it's a Visa card or a Mastercard, you can send it to that network or you should be able to send it to an alternative network. There's a few of them like STAR, Accel, NYCE, Pulse, for example. Now that's always been available basically since the financial crisis around 2008, that's been the regulation. But what we've really worked on is bringing the performance up to a level where that routing decision actually becomes possible, right? Because even if there were previously cost savings available by doing that, if the performance wasn't at the right level, then it didn't make sense to send those transactions down those alternative rails. So we've put a lot of effort into making sure that the integration, the implementation of those alternative rails is at a quality where you don't need to sacrifice performance or cost. And because of that full funnel kind of view, that then becomes very relevant as a recommendation to our customers to say, "Hey, if you use our U.S. debit offering, we will not just optimize cost for you, we will optimize cost and higher performance at the same time." And that's really difficult to compete with because most of our competitors, almost all don't have a single platform structure. They have either parts of that product offering operating on different platforms or they have different markets operating on different platforms or they have different sales channels like in-person and online operating on different platforms. And so bringing that cohesive view is very, very difficult compared to our setup.

And is that -- particularly the direct debit routing, is that where you actually have to go and work directly with those schemes. It's not just your technology that needs to be right, it's the ability to work with those players and get them to a place where the partnership which I think has happened with some of the European schemes as well.

It's quite complex. We have to work with them. And we also had to build a lot on our end to kind of -- to make sure that the quality was at the right level, right? So it wasn't only like explaining to them and having them build, although they did build as well, but also on our side, we built specifically to make sure that the performance on those networks caught up. And that is quite a lot of effort. I think what I'm most confident about is that because we have this foundational -- the single platform component, the pace at which we can drive and deliver innovation just looks very different to what the legacy providers can provide because if they want to bring innovation to -- if they want to bring global innovation, that often means they need to build it, I don't know, 5x or 12x or 20x to bring that around the world and bring that across channels. And we need to build it once and move on to the next thing, and that allows us to keep driving that innovation for our customers.

Now that makes perfect sense. Is there any way you can give us some ideas of scale in terms of how -- where merchants are with adopting Uplift, how material that direct debit routing in the U.S. is across the business?

Yes. So maybe a couple of stats. So Uplift is a suite of modules essentially. We call it -- we have 5 modules or so. One of them is called Optimize and optimize is the auth rates piece I was talking about, so how many transactions do you get approved. That has always been a part of the Adyen offering and continues to be. So almost all of our customers use Optimize. I think where we've seen the biggest traction is Protect. Protect is our fraud tooling. We see 2/3 of new customers signing up for Protect from day 1. That's a much higher number than we've seen historically with our existing customer base. We've made a lot of investments in that product over the past years. Now all of the rule -- there's no static rules anymore. It's all machine learning based, right? Again, coming back to this point of fraudsters are moving quickly, if you have static rules in place like if X happens, then do Y, you quickly need to change that. And you're much better off changing that in real time based on the information you have elsewhere on the platform than having to manually go in and change those rules per customer, right? And so we see a lot of time savings for our customers, but also just way better performance as we've built that out, leveraging machine learning and AI. So there, we're seeing strong traction as well. That is a separately monetized product, albeit the best benefit we get from uplift and better performance is that we get more share of wallet over time, right? And share of wallet, we shared building blocks Tuesday as well. Share of wallet continues to be the biggest opportunity to drive growth in any given year. And share of wallet is basically the proportion of payments volumes we do with any customer, right? So we work with large enterprises. It might be that today, we do 20% of their volume. And the biggest area for us to grow is that we get to 30% and then 40% and then 50%, right? That's the opportunity. And the way that you drive that is better performance. And uplift is how we deliver that better performance on the product side. And connected to that, we usually differentiate through service level as well, through our account management team working day-to-day, understanding deeply our customers.

And I guess it creates a flywheel that the more data the better the performance, the more merchants, more data in that creates...

Exactly.

By the way, it sounds a little bit like cybersecurity software going from rules-based to dynamic and how you deal with it. It feels like there's a similar change in this industry that has happened in...

Yes, that's true.

On The other thing that you called out this year, and I guess this makes sense and something we've been excited about is you've obviously had a phenomenal online digital business, but then you've moved into in-store. And we're starting to see more digital customers move into unified commerce where they're doing both. Is that just the natural progression? You've called it out. Is there something that's kind of tipping this over? And again, maybe just a quick recap on the benefits of what you can deliver in unified versus what someone would get from a different provider.

Yes. So our general model is land and expand -- land and expand. You start small with the customer, you prove out the value. And then from there, you gain more proportion of their volumes. One of the ways that you expand with customers is into different sales channels, right? So it could be that a customer wants to start with us in person in a specific market and then they want to add e-com later or they want to start with e-com and then they want to add in-person later. I think the general trend is that you see more and more businesses having both channels, right? That was really accelerated through COVID, and that continues to be a big focus. And the benefit you have when you have a unified commerce need when you work with Adyen is that you can process payments both across online and in-person together through one integration, and it provides you one data set, right? So you can understand your shopper behavior very differently if you have that data unified in one place, you understand how specific segments of customers are moving more online or maybe online and returning in store or a lot of these transaction trends, you can see if you have a holistic and unified view. So employing that land-and-expand model also means that some of our just digital customers, so customers who just work online with us, move into unified commerce as they add an in-person channel. We have seen that accelerate in Q3. So we called it out. It's a general trend that we've seen over time that we get a bigger share of wallet and that share of wallet comes from more geographies or more sales channels, comes from more products over time. And so it's just a continuation of that strategy that we've been employing over years. But I think we're quite uniquely differentiated in unified commerce. There are very few companies who process both online and in-store across the world. There's almost none that do that across the world and that do that on one platform. It's very difficult to find that solution.

That makes sense. When we continue on unified, it feels like there's been somewhat of a vertical strategy. I guess, looking into vertical finding the pain points, how to help merchants with their specific problems. Luxury has been a phenomenal success for you. Could you talk about the next steps where we're seeing the next opportunities in verticals where you're moving?

Yes. Sure. So we started in luxury retail. Think of LVMH, for example, or Burberry, right? These types of businesses who first focus on customer experience because a big benefit you get when you work with Adyen is you can improve the customer experience. So they were the first ones to really focus on payments as a part of that strategically in person. Since then, we've expanded a lot to broader retail. So think of like an H&M or Zara or those types of businesses. And then more recently, we've gone into hospitality, food and beverage have been big areas where you've seen like the convergence of various sales channels, right? So then maybe in-app becomes important or ordering ahead and picking up in the restaurant or in hospitality, of course, a huge set of international customers, loyalty plays a big role in both industries. So those are our fastest-growing verticals today, food and beverage and hospitality. And I think overall, what we've seen is that more and more verticals see payments in their tech stack as strategically important to driving unique customer experiences, right? And so usually, there's first movers in any given vertical. And as those first movers gain the advantage of better technology and therefore, the better customer experience, then the other players come along. And so our expectation is that more and more verticals will move in this direction of prioritizing technology, prioritizing payments as part of that, and that will be an opportunity for us over time as well.

I wanted to talk about -- because there's always a lot of debate with investors about how much of the market is open to you. And you shared quite a lot of statistics around addressable market at the Capital Markets Day. Maybe just to kind of level set to start off with, could you just talk about how big you see the addressable market for you in payments today, what your share is, just to remind us of those stats?

Yes. Sure. So we've essentially sized the total payments market today at EUR 34 trillion in payments volume, of which we do EUR 1.3 trillion of that. We took out China domestic because we don't have an offering, and we took out sanctioned countries, and we essentially think the vast majority is, therefore, available to us, which is EUR 26 trillion of that EUR 34 trillion. So if you then compare the EUR 1.3 trillion to the EUR 26 trillion we think is addressable, we have 5% share today. So there's a significant opportunity to still gain share. And one of the ways that you could look at it is like who are the biggest players in this market. It's very fragmented, but the biggest players have 3x the share we have, right? So you could debate me if it should be EUR 26 trillion or it should be EUR 24 trillion or it should be EUR 23 trillion. But the reality is that there are competitors which with worst technology with 3x the market share. So we should at least be supported to grow to that type of level if we continue to execute and differentiate ourselves. And that by itself represents a major growth opportunity. Then next to that, the payments market is expected to double over the next decade, just like it did over the last decade. And that's on top of the gains in market share that we should see. And then lastly, we sized the opportunity outside of payments because we've broadened our product offering over the years. We're broadening it into financial products. And the opportunity I just mentioned was a volumes opportunity. In financial products, it makes much more sense to size these based on the revenue opportunity and the revenue opportunity we see just in embedding it. So leveraging our customers to go to market with these products is EUR 127 billion and expect it to grow 20% on a compounding basis over time. So that's also a very sizable opportunity that we will go after together with our customers. It's also a revenue opportunity for them. So we're mutually incentivized. If you compare that to, we did last year, EUR 2 billion or so in total revenues. it's a very sizable opportunity as well. And that's the long-term opportunity that we're going after that we're very excited about.

So if I just come back on the payments market, first of all, the pushback I'd get is that's true. The legacy players are at that level, but they're charging single-digit basis points for commodity big box retail, grocery. And so should we be excluding a bigger part of the market? Do you think -- is that fair? Or actually does more of that come into your view over time?

I think the general trend is that more of those volumes have come into our view, right? So if I take this business 10 years ago, we've been in the U.S. for a long time. We didn't go into the U.S. thinking that we would ever win U.S. volumes. We thought we'd go into the U.S. to sell to U.S. headquartered companies and help them international where the complexity was, but it turns out that the U.S. became really complex over the last decade, more payment methods, unified commerce brought complexity, platforms now, which is a big part of our offering, added a lot of complexity. That's how we go after small, medium-sized businesses. So the complexity has been rising. And now think about a world with agentic commerce, right? That's another lever of increasing complexity. So in general, I would argue that the complexity has been increasing over time as consumer expectations have also been increasing. And that combination gives us the perspective that if you play it out over time, more and more of this business or of the available business will be looking for a differentiated solution than a commoditized one.

If I kind of go the other way around and come to Europe, again, the pushback I get there is in digital, your shares, we estimate it might be getting up to about 20% and so people like, well, how long can they continue to grow from that level? I mean you also shared some data around penetration of cohorts. Is that a good way maybe to push back on that? Are you getting to a penetrated sales level that can't keep going up?

Yes. So the combination of e-commerce and Europe is the longest combination we have, right? That's where we started Adyen. So that's where we have, let's say, the most maturity in terms of the market. Having said that, our view is we have around 15% share. And that 15% share can easily go much higher, especially as we've been growing outside of digital core, right, which is unified commerce. And we never had real SMB penetration in Europe, where SMB is a really big part of the European market. So there's a very significant opportunity for us to continue to grab share even in what's the market where we've been the longest. Europe continues to be one of our fastest-growing regions even at this scale. There's so much complexity across all the fragmentation you have by country. You have different payment methods in every market. This is a market that's also often, especially on the in-person side, been serviced through banks, bank -- country by country, bank by bank. And we can help simplify that complexity. So we still see a lot of potential in Europe.

So you mentioned platforms a few times, which I guess is another big runway opportunity. Could you maybe just explain what are you doing on platforms? How does SMB come into Adyen's opportunity because of what you're doing in that space?

Yes. So what you've started to see, and it started in the U.S. and has come to Europe over the last years is that small businesses are starting to get financial services from what were historically software providers. So let's take an example. Let's say you are running a restaurant. You very likely now use one of the software providers, the SaaS solutions to operate your restaurants, right? There's a number of them that you could use to run your restaurant. What you've seen is that they've all added on financial services to that. They start with payments. So now you can also buy payments from those providers. And then you can also provide other financial products, by other financial products, like you could get a bank account from them or you could get a corporate card to manage your expenses or if your oven breaks, you can get capital from them, right? You can get a short-term loan. And all of those products are connected to one another. So we've built out those products as well. And they all interact with each other. So if you -- if we issue some -- we call it Adyen Capital. If we issue capital, a short-term loan to one of -- to a restaurant, we collect it through payments every day. A percentage of it, we just collect daily, and we make the risk decision based on their sales history because we see their payments activity. And we're seeing a lot of success with this model. So it started mostly in U.S., moving to Europe. But for us now, it's a part of our business, which is growing over 50%. And that's basically entirely payments today. But over time, that's going to become broadly -- more broad these financial products as well. And those financial products are already a really important part of our pitch today because so many of these platforms, they have the ambition to roll out these products over time, and they start with payments to do so. We did some research. We see that the majority of small businesses are actually expecting to get payments in financial products from a platform rather than a bank, for instance. And that's a big shift. And the reason that they're interested in it is, again, the embedded nature of those products working together also with the product suite. And because each vertical kind of has their own specific needs or requirements, we're seeing that mostly this part of the market is being picked up by the vertical SaaS providers in each vertical. So we talked about food and beverage here with a restaurant, but you could also imagine if you run a hair salon, there's a vertically oriented platform to manage your business. And I keep repeating this example because I love it, but there's one to run -- if you run a pizzeria just in the U.S. you should use Slice, a platform there that's just for pizzeria in the U.S., right? So it can be very vertically oriented down to the specific needs of that customer group and their own demands, but that's absolutely how we think the SMB segment is going to be serviced. And because it's represented by a larger enterprise platform behind it, it's a good fit to our technology. So we basically enable them to go sell into those SMBs.

So this is about digitization of SMBs as they go to software, then that software vendor has the opportunity to sell all of your payments and embedded financial services products in -- great for the software vendor, great for you. I guess, it's stickier, bigger revenue opportunity for a better experience to the customer.

Yes, for sure. For them, it's much stickier and it's a revenue opportunity. So they're monetizing each of those products, too. And you have some of these players who used to be software -- they used to be software companies, and they very much look like financial services companies at this point, given where their revenues come from. So it's mutually incentivized.

Perfect. What does the competition look like in that space? Is it different at all from rest of the payments markets?

There's much less competitors in that space because the complexity is much higher. So you need to offer a wider set of services. You need to do that globally. You see -- you need to often have a unified commerce setup, right? So in those cases, take a restaurant or a hair salon, both cases have an online and in-store component, something that you also need a strong offering in. So you find very few competitors in this space. There are a few also newer tech stack, more innovators in the space as well who are also offering a good offering in this space. So there's certainly still competition. But in terms of the number of providers who can provide this offering, it's much less. And we feel we have quite a lot of differentiation, which is also reflected in the pace at which this pillar is growing.

And that's been a kind of longer-term monetization opportunity. I think again at the Capital Markets Day, you gave some -- you gave a little bit of a feel for what you thought you could get from just moving on from payments to the embedded finance side. Can you share with how that changed your midterm outlook on that side?

Yes. So again, I shared the market sizing of it. If we then reflect it into how we look at growth over the coming next few years, we think that ultimately, it basically adds 1% and then rising over time in percentage to our growth, right? And that starts small, right? Because we see, for instance, in issuing, which is the financial product, which is furthest along that we've really started to get to an inflection point. So we showed this chart, the real J-curve that you could imagine. But -- still this year, we expect to do EUR 6 billion in issuing compared to the EUR 1.3 trillion that we're doing on the acquiring side. And that business is also fast growing, right? So we're really starting to see acceleration in our financial products. At the same time, it's starting off a very small base and will take time to build. And we're okay with that, right? We took the strategy of create the solution, which adds the most long-term value to our customers. That was the longer path. That was building it ourselves, all of the technology we built in-house that was getting our own licenses to do so. And so that was a longer path, but we're confident that, that's going to bring more value to our customer base over time. Capital and accounts they're earlier in their maturity phase than issuing is, but I would also expect them to get to their inflection points in the next years as well. And I'm really excited about the potential of those products to drive growth over the coming years.

And on the issuing side, I mean, you're absolutely right. It was a pretty incredible J-curve that you showed at the CMD. Is that just the technology needs time to mature to build pipeline, salespeople? Or is there anything else driving that inflection point in the market?

I think with any product, any new vertical, you also start to need to get references. You need to build up the product gets better with more volume, right? So you're constantly iterating on the product. And you need to get some scale on to the product to make those iteration cycles faster, the learning cycles faster and improve the product. And I think we're just getting to the maturity level where we both have the references and the quality of the product, which will allow us to scale. It happened the same with in-person, right? In-person also was a very small part of our business. We needed to build scale reference customers. We needed to improve the product because we built it from scratch. And all of that takes time. In-person is now 20% of our volumes. So I expect that issuing also continues. Now it hits its inflection point. It will have momentum. There's reference customers. There's the product improving itself and iterating on itself over time. It will continue to get better, and I would continue to expect that kind of quick growth on the issuing side.

You've alluded to agentic commerce a few times. I guess in tech, there's always a bit of a -- there's marketing and then there's actual technology that's real and underneath it. I think there's a bit of a concern that from the press releases and marketing, Adyen wasn't as involved in the agentic side as you could have been. Again, you talked a lot about that at the Capital Markets Day. Could you talk a little bit about what you see happening here and how you're involved?

Yes. So the reality is that very few transactions are actually happening on agentic commerce, right? What -- what LLMs are being used for right now is mostly discovery. So we gave an example Trevor did, our SVP of Digital...

Nice cashmere jumper. It was a nice cashmere jumper...

Yes, exactly, exactly. So he said, find me a gray cashmere sweater for Investor Day under EUR 100, right? That's what people are using these LLMs for, but like -- they're not actually paying through that process, right? So they're getting a few options surfaced to them, and they may be deciding one, but they're still closing out that transaction at the website of whatever brands they see. We think that over time, that will change, that the buying will actually happen where that original interaction is happening. And that could either be by making a click by the individual themselves clicking, I want to buy that one or the next phase would be that they are actually instructing the agent to just make that buying decision themselves based on information they've provided. That is still a ways out. And everybody is trying to figure out what is the right protocol, what is the right environment to support those types of transactions. You have the card networks building up their own protocols, right, Visa, Mastercard. You have the wallets, right, Google or Apple looking at how to set up their wallets for that environment. You have the actual agents themselves. So think about OpenAI or Google again on the Gemini side who are looking at how to set up the right protocols. And we've done over 100 customer interviews on this to understand the needs of our customers. They're very excited about another sales channel, but they're also very worried about disintermediation, about separating their brand and their experience from their shopper. So what we've taken a very vested interest in is setting up these protocols in a way that's still merchant first, that still allows our merchants to interact with their end users in a way that's beneficial to them and which can help them drive loyalty and the same types of behaviors and patterns that they've seen over the years. We're in discussions with all of those players, right? So we're in discussions with Visa, with Mastercard, with Google, with OpenAI, with Cloudflare, right? Because there's also this interesting change that comes from some bots being good now, right, whereas historically, bots have always been bad. They were always fraud. Now there's going to be good bots as well. So how do you differentiate? So we're in discussions and building out these protocols with each of these players to ensure that the feedback from our customers is being incorporated into the setup. Now what does it ultimately mean for Adyen? If you think about agentic commerce, in some ways, it's just the next step on the challenges which already existed in e-commerce, right? So what matters in an agentic world, authenticating some -- an agent, -- are they who they say they are? And are they acting on the instruction that they were instructed on? Payment method mix is important. Are people still able to use the same payment methods for the same type of transactions? What happens when there's fraud, right? Even in the most benign case like a kid accidentally got on the computer and just typed into ChatGPT, get me candy, right? All of those things become much easier in this environment or even, hey, I want to buy concert tickets in 3 weeks. Please make sure you buy those when they open up on Friday in 3 weeks. And then in 2 weeks, you decide to book a vacation for that same week. So you want to cancel it. How do you make sure that, that intent is still there, right? So there's a lot of these challenges, which they still need to be solved. But what we take a lot of excitement from is that they'll build on the same things that we've already built, right? Again, that single platform, that end-to-end control, we've always been really good at authenticating a user. We've always been really good at making sure payment method mix is available to them. We've always been good in understanding fraud patterns. So all the building blocks, which are going to be very relevant in this agentic commerce world are something that we have real strength in, and we think that we can deliver in this framework as well.

I mean as well as you talk, it just sounds like there's a lot of complexity, which should be perfect for what you're building to.

That's the summary. That's the right summary. The complexity is increasing, and that's where we thrive.

Exactly. You also gave a framework for 20% -- approximately 20% growth over the next few years. Could you just talk a little bit about the decision to move to that kind of framework and guidance and give us a little bit of an idea of how we build that 20%...

Yes. So the idea is relatively simple. We've been very long-term oriented, and we want to tell the long-term story again. That's why we did the whole TAM analysis more than to defend that it's not a limiter. We wanted to show the potential for Adyen, right? And the potential is that our market should support a significantly bigger version of Adyen than we are today. And that's not even to build something that's never been built before, like there's currently competitors of ours who have 3x the scale, right? And they don't have a better product offering. There's no reason we shouldn't be one of the biggest players in this space. So we wanted to, again, connect to this long-term view. We know we're on that path. That means we'll gain share. The market itself is growing. All of that should drive Adyen's growth potential over the long term. We've also opened up new product offerings, creates additional addressable market, drives other avenues for growth. So the long term for us, that's the opportunity we're going after. And at the same time, we said the growth path to get there won't be linear. There will be years we grow faster, years we grow slower. That's mostly in any given year, driven off of the growth of our existing customers, both the macroeconomic environment, how fast those customers grow themselves, but also the pace at which we gain share of wallet, and that comes down to their road maps and their priorities, right? So if a customer would say, next year, our big focus is going to be Singapore. We would say, great, we're going to help you with Singapore. If the year after they say our big focus is going to be U.S., we would say, great, we're going to help you with U.S. Of course, U.S. is a more sizable market, but we follow their lead. We follow their road maps, and we help them where they see the biggest importance. So you see that share of wallet while consistently gains has a bit of up and down based on that specific year's priorities and what that looks like. So we get that visibility about 6 to 12 months out, basically around now, right? All of our customers are going through their planning for next year. They're all thinking what are they going to prioritize, what's available to us. And we're building out specific account-by-account plans, which say these are the opportunities we're going to go after. We want to give that visibility to you all to the market. So our idea is that each February, basically, when we close the year, we'll also give our view into what the next year's growth will look like. So if you think, okay, you have your long-term view, we know we're on that path to getting that bigger share of market and an expanding market as we also grow into new products. We'll give that shorter-term view of the information we have on a net revenue basis. And so we're moving away from like that medium-term guide that we had 2 years ago. We didn't -- we still wanted to give some visibility into what to expect over the coming years. So we created this framework, the building blocks we call them. It's basically the individual levers which drive our growth in any given year and the relative size of them, but it's not something that we're going to revisit on a continuous basis. We're going to continue to talk about the long-term opportunity and connect it to what the growth opportunity then looks like each given year. That's the plan going forward.

That makes a lot of sense. And I think people will find that helpful to have the long-term framework and then the annual setup. On the profitability, you also guided to an over 55% EBITDA margin in '28. Could you just talk a little bit about how you think about balancing investing for growth, driving the operating leverage in the business?

Yes. So first and foremost, we want to drive growth in the business. I mentioned the long-term opportunity we have. We need to invest in that opportunity to capture it at the fastest rate we can. So we want to invest in the teams and grow the team. 75% of our costs are related to the team. So that's the place where we can make these investments, and we'll make those. The biggest areas we invest are in our tech teams, in our engineering and product teams and in our commercial teams, adding account managers as we add more customers, right, they focus on the existing customer base and adding new salespeople to go win new customers. We'll continue to make those investments. At the same time, we're seeing efficiencies from automation. We're making investments in AI-enabled tooling as well. And we've always been focused on automation, but this is another set of technologies, which will help us drive that even further. And that balance means we can continue to invest in the team while still growing the team slower than the revenues of the business. And that's also connected to that single platform. The fact that there is a lot of scale to the way that we've built out our functionality and our solution, which means that as we add more volumes from our customers, we don't add significant incremental costs. And that just brings a lot of scale to the organization. So we think we can both invest, first and foremost, in the teams while still seeing our operating leverage visible and growing EBITDA margins above 55% in '28.

Perfect. We're bumping up against time, unfortunately, so I'll close it there. Ethan, thank you so much again for joining us and best of luck with that journey out to '28.

Thanks. Appreciate it. Thanks, everyone.