All right. Good morning, everybody, and welcome here. What we're going to do today is try to give you as much transparency on what we actually do at Adyen as a company. What I actually like about, maybe anecdotal, but what you see here is a little movie. In this film, you see a robot working on a payments terminal. Why do I think that this is funny? That is, if you are putting new software on the terminal, there are about 150 tests that an engineer needs to go through. Adyen is a company where about 40% of us are engineers. If you ask an engineer to do boring, repetitive tasks, they obviously are going to think about something better than that. We had an engineer who was interested in robotics, and what he did is he built a robot to do the test.
When there's a new version, overnight, a robot is doing the test. There's a little camera filming what's in the display, and the test involves doing everything that could go wrong with the terminal. Say, halfway the transaction, suddenly it pulls out the card, stuff like that. Each time when we find in the field something that our terminal didn't react optimal to, we put it in a new version and it's tested. We started with one robot. Now we have a whole room full of robots doing that. We say you're redefining payments. Of course, we always take a different look at things. The robots do not exist, we make them ourselves. They are 3D printed. More anecdotal, but that's what you see here. This is the agenda for today.
After a quick introduction, I'm going to hand it over to Brian and Edgar here. This is the team that we have present here. We will be around also during lunch. We end the day with Q&A, there's enough room for questions and interaction with us. What I like here is that we founded the company with myself, the CTO, Arnout, and we have four engineers that we started with. Michiel was one of them, he's part of the founding team. This is something that should be recognizable for you and that we're going to talk about today. Why do we feel that working on a single platform puts us in a very different position? Total addressable market, we don't think it's a limitation to what we do. Look at who we work with.
We have a proven track record of working with the largest, most advanced customers in this market, always consistently, they give us more and more share of wallet. Ingo will talk about the numbers, also talk about how the business model works with further growth in the company. We defined for ourselves as a company the three things that we do. On the highest level, what is Adyen's objective? Number one, obviously, is to help our merchants grow. If we help our merchants grow, we grow with it, but also that defines the differentiator between us and our competitors. The better we are in helping our merchants grow, higher authentication rates, more countries, unifying channels, all the things that we do, it's a reason, it's the delta that they see if they move from competitor to us. That's number one focus.
Secondly, we redefine the payments market. What does that mean? The way how things used to operate, we don't take that as a given. We challenge it. Maybe somebody will talk today about how we integrate our payments terminals. We do that in a different way because how it's commonly done, we feel creates too much trouble. That something is standard and that that's done in a certain way, we challenge. This one, we have fun whilst doing it. We feel that if we are capable of quickly developing new software, quickly take it into production, really helping our merchants, that will create an environment which is attractive to the highest talents. We've never done an acquisition because we don't think we should veer away from doing that single platform, but we also don't like doing it.
It would be negative on being able to attract the best talents. The way we've structured the company is a company in which your day job, it's not that after work, but during work, this is something you feel the traction. You feel that you can actually have an impact. This company feels that the people who work for Adyen feel that that's the fun in their job. I'd like to hand it over to Brian and to Edgar.
Thank you very much, Pieter. Unfortunately, I'm not Ingo. I won't be talking about numbers. You'll have to listen about product for a little bit, pardon me. Pieter talked a little bit about us doing new things in the market and setting new expectations and really turning over payments on its heads and doing it differently. If you've come to one of these before, you've seen this slide. There isn't a discussion that we don't have with a prospective merchant Adyen in which we don't start the discussion with this, it's the fundamentally different approach that we've taken to payments in the industry. If you're not familiar with payments, the top is the value chain that most payment companies embrace. You have a merchant on one side. They interact with a gateway or an API layer.
Underneath that gateway or API layer is a risk management system or a totally another company managing that risk management. There's then a set of acquirers who then use underlying processing systems. That underlying processing system then manages that communication between that acquirer and the schemes or the networks, Visa, Mastercard, Amex, Discover, JCB, CUP, et cetera. The other side is the issuer, then the communication goes all the way back. Entering 16 digits, goes through all those players, comes back, and you get an approval or a decline. We believe that this model is fundamentally broken if you're trying to achieve the best amount of performance and efficiency in the market. There's a few things that go wrong there. First and foremost, most of these different entities are completely different companies and systems, all at different stages of growth and maturity.
Generally, what a merchant working with a traditional player in the space would see is degradation of data between those different points. Because of that degradation of data, degradation of performance, generally seen in authorization rates. For every 100 transactions that go through, do 100 get approved? Do 99, 98, et cetera? That is the primary metric that any merchant is looking for in terms of their payment partner and what they can achieve. What we did is we came into the market with a fundamentally different approach, which is to have one universal platform that is completely channel agnostic, that can achieve the entire value chain with one platform and one partner.
This not only achieves significant benefits from a cost-saving standpoint, you have one contract, one partner to work with, but we achieve better data, better performance, and better overall global reach because we have one platform. Plugging that platform into different regions over time becomes very easier. Edgar is going to speak to that at length in just a moment. Now, what does this result in? We like to talk about four key benefits for our merchants. The first is global reach, letting merchants find customers throughout the world and do payments with them. That, at the end of the day, is why we're here. It's why we work with merchants. The second is unified commerce, that's a new way of thinking of a channel-agnostic approach to payments that our merchants are embracing with us.
The third is data enablement, doing things with data, the fourth is ongoing innovation. We believe that at the end of the day, that's what truly sets us apart, our unparalleled speed in terms of bringing innovation into the market. Now, I'm going to come back and talk about two through four. Edgar, who is our Head of Global Acquiring, is going to speak to the theme of global reach.
Thank you. Hi, everyone. My name is Edgar. I'm looking after the global card networks and payment methods that we have available on our platform. I would like to talk you through what it means to have an overlap of global domestic licenses with the strength of our single global platform, how that helps us innovate, and how that helps us provide the best services to our merchants. To start with, I would like to show you the map. You've seen in our recent reports that around 70% of the processed volume now is on full-stack acquiring. That means that a transaction, when it leaves the merchant until it reaches the card network, does not leave our rails. We are in full control of data quality, of processing quality, and innovation.
The benefits that it has for a merchant is something Brian will talk more about, but also on the card network side, this is something more and more new. That means that if we develop a new product with the card networks or we implement a new global payment method, let's say Apple Pay or Google Pay, or we derive or we build Real Time Account Updater with Visa or Mastercard, we don't need to build it market by market. That's the common practice in the payment world. We don't need to wait until local acquirer, local process is ready until we can offer it to our merchants. We build it once, and we push it into every domestic market that you see here. That gives us our speed. That's the benefit of global. Of course, why do we localize our offering to merchants?
Why do you not offer acquiring out of one region for your merchants? Many of the merchants we serve are either having an entity in one of these markets or in multiple markets. The way card schemes are set up is that to grab the full benefit of payment processing, you need to set up a local registration into the Visa, Mastercard, Amex network to qualify for key benefits. Before we can do that, we as Adyen, we set up domestic entities. We integrate with the relevant bank systems. We make sure we meet all the regulatory requirements, then we apply for a license, from that moment onwards, we can start onboarding merchants. What do we then offer in each market?
When we offer them the localized service of Adyen, we can provide higher authorization rates because the bank of a shopper will see a transaction coming in as a local transaction, which is considered lower risk, there's a higher likelihood of approving a transaction. The second is lower processing costs. The card scheme rules and pricing tables are set up in such a way that domestic transactions are priced lower than international ones. Because of our pricing model, global Interchange++, we pass the benefit on to our merchants. Third one is faster settlements. Domestic transactions are typically in a single currency, get cleared over the local clearing solution. We get our money faster, we can pay out to our merchants faster. It's another key benefit of localizing our offering.
The fourth one, which sort of comes more into the product space, is the local shopper experience, which is maybe less quantifiable, but it's very key. Our merchants spend a lot of time to fully localize their connection with a shopper, setting the right language, opening up stores. In your payment experience, you expect as a shopper, something you know and recognize, something your bank has explained to you. Contactless works in a certain way. You don't get a weird surcharge on your card. You recognize what is written on your statement. All these small things matter, local acquiring helps to obtain that goal. If you then take the concept of local localization and offerings, helping a merchant present themselves as local as possible to a merchant to the next step, you come into the domain of local payment methods.
You know that on our platform, we're offering more than 200 different local payment methods. In every key market where we're active, we're making a mix of global payment methods like Visa, Mastercard, American Express, all the ones a shopper knows, add the relevant local payment methods to it, so that if a merchant goes live with a website or a store in a country, the shopper has the right choice, the merchant can present whatever is relevant to reach that shopper. To make that more concrete, I would like to run through four different scenarios. The first one is Germany, not that far away, I know many in the room are coming from more card-heavy markets. It's good to realize that in many markets it works slightly different.
Even in Germany, it's not that common if you pay online to use your credit card for that. It's more common to identify yourself against your own bank using your IBAN details, for instance, or your bank details, the transaction is processed over the SEPA rails. If you want to go reach your German shoppers, you need to have this as a payment method, in your checkout flow. Second one is Alipay. Recent reports say that in domestic electronic transactions in China are now over 50% done over the Alipay network. That means that if you want to offer a service in China or you want to reach Chinese shoppers outside of China, you can actually make an impact by offering this as a payment method. Even here in the Netherlands, where everyone has a debit card in their pocket and pays contactless all the time.
When you want to pay online, you cannot do that with your debit card. You need to use a payment method called iDEAL, which basically is a flow where if you check out online, you get redirected to your banking app, you authenticate, you select your current account, you come back to the merchant environment. With PSD2 coming up, we'll explain more on that, we expect to see more and more diversity in payment methods in Europe coming up, with different kinds of authentication against your banking environment. The fourth example I would like to highlight is Brazil, which is a very card-heavy market. People pay with their Visa or their Mastercard, they still expect at moment of checkout to select how many installments instances they would like to pay. If it's three, six, 12 slices, they do that to manage cash flow.
As a shopper, if you don't get that option, you feel something is off. Also for us to localize our offering, we're building all the relevant local variations of card payment processing to make sure that the merchants can reach their shopper in a way that the shopper feels trusted, that they get exactly the experience they like, get the most out of that. The main thing I want to explain here is that our single global platform gives us speed in terms of innovation with the global payment methods and the global card networks. We build something once, we push it into every market, which is unique in the market and also in our partnership with Visa, Mastercard. That helps us drive speed.
We complement it with relevant local payment methods. On top of that, next to all the data that generates, we also have a global team, of course, to help merchants excel in individual markets so that we can also give relevant advice on how payments work in a market, how they need to set up their checkout flow, and which payment methods they need to offer to reach the shoppers in the best possible way. Now on to more the merchant focus and product for Brian.
Thank you very much, Edgar. Going off of the theme of, let's say, I think there was a China example there, Alipay and WeChat Pay. If anybody in the room has been to China lately and you've gone to sort of like, let's just call it a normal restaurant, so non-high-end, right? The way it works nowadays is actually you sit down at your table, there's a QR code at the table, you scan it, and the menu is actually in WeChat. You are not handed a paper menu. You go through, you select all the different things that you want. The food comes, waitress brings the food, and at the end of the meal, you're generally presented with the check, and then you can pay with WeChat Pay, Alipay, China UnionPay, et cetera.
The interesting question is, what kind of payment was that? Was that an online payment? You did it with your phone, or was it a point-of-sale transaction because it was done in a store? We like to talk about unified commerce because what's happening all around the world is that the distinction between channels are becoming gray. Just like the example that I just gave, it's not important actually where the consumer was physically located. The only distinction between those different channels is the device that was used to initiate the payment. We are one platform, and from day one, we had the benefit of having a team of engineers and experts who had done payments before. They knew that if Adyen was going to succeed, it needed to be channel agnostic. Right?
From day one, even before we had started integrating terminals and traditional point-of-sale in-store transactions, we already had a platform that could do that and could be flexible enough to handle that sort of thing. Whilst the majority of the industry uses the phrase omni-channel to describe having multiple channels within one company, we prefer the term unified commerce. Unified commerce is going beyond simply offering in-store payments, online payments, and in-app payments, and it's about connecting the dots between those. It is not uncommon that at a traditional retailer, there are two heads of payments that actually meet for the first time when we're bringing them together because there's a different head of payments for the in-store channel, a different head of payments for the web channel, and sometimes even in-app.
That sort of segmentation of their thinking also extends into everything from what partners they manage to where their data is isolated. When a merchant then embraces unified commerce, which is a complete combination of all those channels into one solution, you can get one integration, one contract, one back-end, and that has benefits from a business perspective. Most importantly, there's a holistic view of the shopper, right? If you have one platform that's handling all of those payments, there really isn't a distinction between data in terms of seeing where shoppers are going between those channels. I'll speak to some of those benefits in just a moment. At the end of the day, though, it is about these three different channels. I'm going to talk about our in-store offering, our in-app offering, and our web offering.
I think it's extremely important to note that this all sits on top of one platform. You could go to a similar presentation at another payments company, and whilst everything would fall under their name, underneath, those are all different platforms that come from different acquisitions over time. The traditional way of going about building a payments company is to just buy companies over time who specialize in particular region or channel. We are not shackled by that sort of thing. When we look at in-store, right? Let's look at your traditional sort of retailer, where maybe the vast majority of their transactions are in-store. This tends to be what they look like. They have a lot of different device types. Per region, they tend to have different PSPs, and then per country, they have different processors and acquirers.
That means it's not unusual for a large global retailer to be managing 30, 60, and sometimes 100+ contracts and partners. One of the main things that we provide on top of our superior technology and payments experience is the simplification of this. It is not unusual for a large international company to be eliminating potentially dozens of different partners and integrate one implementation across all of those regions with one set of contracts and one solution. Not only is it a superior product, but you can also look at it from a macro perspective across your business as a retailer and realize all those benefits of reducing complexity. One thing that I think is really important to note is our Terminal API and why on a thematic level that really sets us apart.
Most in-store transactions where there's an interaction with a terminal, the interaction between the terminal and the merchant's system is done with very rigid implementations that are commonly called libraries. Basically, there's a piece of software that sits in the cash register. It is designed to then interact with that terminal, and the problem you have with that is versioning, where everything needs to be in sync between the two. If you're going to make a change in the terminal, you need to make a change in the library. You want to make a change in the point-of-sale library, you need to make a change in the terminal. Things break all the time. Because things break all the time, everything slows down. You want to implement Alipay and WeChat Pay for a merchant ahead of Singles' Day coming in three months?
That's going to be difficult because you need to have both sides of the fence dealing with that implementation. The online world figured this out a long time ago, they figured it out with APIs. The entire point of an API is that the implementation layer, the way you communicate between devices and systems, that remains consistent. Behind that, you can swap things out. You can change systems all the time. As long as the language you use to communicate between those remains consistent, you're good to go. What we've done in the industry is that we've taken a big leap forward and we've embraced API interactions with terminals. This is extremely rare, in particular in the enterprise point-of-sale space. It's positioned us very well because what this allows us to do is implement new features very quickly.
There is an inevitable move over time towards cloud point-of-sale. Right now, the point-of-sale system, the brain of the store, is generally a box covered in dust that sits in the back room, and it communicates locally on that network. Over time, more and more of that is going to transition to the Internet. In fact, some regions like China, cloud point-of-sale is much more common than localized point-of-sale. There is no distinction between us interacting on a local network at an API level or interacting at a cloud level with that merchant's point-of-sale system. The main theme there is forward-thinking. As in-store evolves, because it will evolve, it's probably going to be the area that changes the most over the next 10 years. In-store transactions are going to look more and more like e-commerce-like interactions.
We're going to be very well-positioned because of our embracing of APIs early on, whereas more traditional players simply aren't going to be able to keep up. One thing I will note, you'll see a little kiosk over here on the left-hand side. During lunch, Dirk, who is a Product Manager who works in our point-of-sale team, is going to be able to do an interactive demo with you of our Terminal API, and he'll be able to answer technical questions on that. I highly encourage you to check it out. Overall, for the proposition, I won't go over all of these points, but I think at the end of the day, it's about quality. It's about having just simply superior technology, because the traditional way of going about payments in the in-store realm is that there's many different partners at play.
Someone else is writing the software on the terminal, someone else is dealing with the hardware, someone is dealing with the logistics, someone is doing the back-end. We embrace that full-stack, all the way to the software on the terminal. The example that I used before, which was implementing Alipay and WeChat Pay in the terminal with three months to go before Singles' Day, we did that. We had a high-end retail merchant who needed it for Singles' Day, it won us the business because we were able to implement it. We were then able to implement it, that immediately goes to all of our customers in all the regions that we operate because we have one platform, where as soon as we plug something in, it becomes immediately available globally for our entire customer base. I'll speak briefly about in-app.
I think there's a few themes that really come to mind. We are one of the more predominant players to work with for app-based companies if you're looking to do global payments. I think there's a few things that have really pushed trends in that area. The number one probably being the Uber-ification of payments, right? When you do an Uber, you don't really think about payments. You just get in the Uber, you get out of the Uber, and payments are this thing going on in the background. We work with a lot of merchants who want to embrace that sort of card-on-file, tokenized payment method sort of approach. They also want local payment methods, and they want flexibility, right? That's where Adyen Checkout comes into play.
I won't speak at length about this, but I want to talk a little bit about implementation and how you implement a payment offering. Generally speaking, there's two really wide ends of the spectrum in the industry. You have APIs on one end, and there's varying qualities of them to have a payment solution. You interact with that API, you build the UI yourself as the merchant. That's what most enterprise merchants look to do. You have the other far end of the spectrum, which are SDKs, software development kits. Those are generally payment providers for smaller merchants, where the whole idea is two lines of code and then you're done. You plop this in your website and your app, we do everything for you. You're good to go. That's not really our segment, right?
What we've really gone into the market with something special is an in-between offering. Now, the majority of our customers interact at an API level. They don't want that two lines of code and then you're done sort of approach. We have that, but the majority go for the API side. What is special about Adyen is that we offer something called components. Now, if you are a merchant, you are thinking all these other things about your business, how to reach customers. You're generally not thinking, how does a Dutch person want to interact with iDEAL? How does a Chinese person want to interact with a QR code? How does a German consumer want to enter their IBAN? We think about that every day.
We have experts around the world who think about that every day, and we provide compartmentalized software libraries where the developers in the merchant side can just plug those in. They have all the flexibility in the world to work with our APIs, but if they want to use these components, they're completely open-source and on GitHub and everything. They can slice and dice them and make sense of them. Now, I'm going at length on this, so I'll move on. What this really means is that we've embraced what our merchants are asking for, which is, "I want the API. I want control." At the end of the day, they're not experts in global payments, and we are, and we help them build these experiences. We help them build these experiences in-app and online.
When you look at web payments, there's a lot of functionality that you need to consider. We have a lot of different segments that we work with. We work very actively with airlines, which if you want to get into a geeky discussion for an hour of how airlines are special snowflakes in payments, we can do that. We work with a lot of airlines globally. Traditional e-commerce merchants, sharing economy, digital goods, recurring. At the end of the day, what they're looking for are multiple things. They simply want that advanced functionality. They want all the little nitty-gritty things to offer the exact experience that they want. They want tokenization. They want high-powered tokenization technology that we offer them so that they can keep those credentials on file and offer those one-click experiences that they want to give.
We can help with that lifecycle management of that card. As you get a new card because your card is expired, we plug into technology on the scheme side that allows us to keep those up to date. We were the first acquirer to go live globally with things like Real Time Account Updater, which allows us to do that. Of course, as I spoke to at length, API and SDK-powered integrations that give them that flexibility that they're looking for. How does this all come together? I'd like to show you a video that sort of brings a lot of these concepts together, and I would ask you to think back to my opening comment in this section about what it's like to order something at a restaurant in China with WeChat. We think of a scenario in which unified.
Of all those workstreams. We're very advanced people in the UX design. Tim Hudson is actually somebody I know for a long time, and with a background as an artist, which I think is interesting, that if you then make the step over to engineering, you get a combination of unique skills, which I think you can find back in our products. Rob wasn't really at the very first team, but has been for a long time with us and has the background of building this whole platform. These are the workstreams that I talk about. Currently 22. Yes, we had less of them, and we'll get more of them over time. What are examples of workstreams? A workstream could, for example, be authorization, looking at authorization rates and doing everything with AI to improve that.
Each time when we see the possibility to move to a higher authorization rate, and especially online, this is very relevant. It still happens if you buy something online that sometimes does not work. That can be for insufficient funds, but unfortunately, it also happens for other reasons. That's a workstream to look at that. Of course, we have a workstream around data, or we have a workstream around local payment methods. All those workstreams together make sure that we can react in a very quick way to this constantly changing environment. Michiel.
Thank you. I'm going to talk a little bit about the technology and the platform, which Brian and Edgar have talked so much about the product side of what the platform can do. I want to give a little bit of background on what the platform looks like when you peel back the covers. I've been on the engineering team since day one as one of the first employees. I've seen this codebase grow. We've always strategically thought that the best way we could grow this company and the best way we could expand was by keeping a strategy of having a single codebase. A single codebase makes it possible to have a single platform. We've all seen the importance of having a single platform.
It means that all of our engineering efforts are concentrated on improving, fixing, and expanding that one platform and that one codebase. If we talk about one of the products which Brian has highlighted, for instance, 3D Secure 2.0 or something about Real Time Account Updater. Most players in the market have this problem that, yes, they can implement it, but which regions do they implement, which platforms do they implement it at? A global merchant will typically be connected with one of our competitors into multiple of their platforms. They'll be connected into a platform for U.S., they'll be connected into a platform for Europe, maybe even two platforms from Europe. One to do the local payment methods, one to do the cards. They might be connected into a platform for Asia, and maybe even different banks for Australia.
That creates a problem because once you develop a new product, as a company, if you have multiple platforms, you're going to do a business case evaluation of, does it make sense to expend all that effort in a product which might be very relevant in one market but only partially relevant in another market. That means for the merchant side, a big disadvantage because they will have the experience that something which will work in one continent or in one integration will not work in another integration. They have to customize and basically deal with one company they are connecting into, but on multiple platforms, and they won't get the same functionality and experience across the base.
This is where having a single codebase, which we can deploy all over the world, means that every time we work on a piece of functionality or we improve our platform, that improvement or that new functionality becomes available instantly across the globe. That means that for our engineering teams, we get so much bang for the buck for developing a piece of software because of the scale and the reach we get from it, that it makes a lot of sense. The business cases are easy to build because we only need to have a small upside because we can get so much leverage from our code improvements. Another aspect is that from the get go, from the beginning, we saw that it was important to base all of our software on open-source. There are a couple of good reasons for this.
One is independence. If you build your software on databases, infrastructure, application servers, et cetera, which are not open-source, you're tying into sort of technology path of your partners, and you're basically going on a ride with that technology. That means that if one of your partners decides or one of the platforms you're using decides to take a certain direction, you either have to follow that direction or you have to change the underlying technology. With open-source, you're much less dependent on that. Typically, open-source projects, you can change the direction of them, or you can extend them in such a way that they will always fit your business needs. Another advantage of working with open-source is that you're not paying any licensing costs.
Not having to pay any licensing costs means, one is you keep your cost level low, the other one is you make your engineering decisions based on technical assumptions or on technical reasons rather than licensing reasons. A good example is that for instance, when we have our payment processing front ends, which are essentially microservices, when they store the transaction, they do that in a database. And each payment front end has its own database. If you were to do that with a commercial database vendor, that would be prohibitively expensive, and you would never do that. While it makes really good sense to say that if you're handling a payment transaction, you should store it in the best possible way of storing it, which would be in a database.
That's just one example where using open-source gives you an independent path and keeps you in complete control of your decisions and the path you take. Another aspect of the platform, I think Brian has mentioned that briefly, is everything we do, every status change in the system, every transaction or anything that happens during a transaction is put into a high performance double-entry bookkeeping system. This accounting system is highly scalable as well, and we use it as an event sourcing model. I know this is unfamiliar for a lot of you who are not in technology, but it basically means that you can see every transaction and the status of a transaction, what it looks like at a certain point in time as being the culmination of a set of accounting events.
Because the accounting events and the transaction events are one and the same thing with us, we get a very high precision in our bookkeeping. It means that we can rely when we're building new products, that if we log all our event transitions in our bookkeeping system, that the outcome will be very precise. We don't have to worry when we're developing new products, that there is a data quality problem or that we need to get the financials completely right because the bookkeeping system guarantees that. That translates into a high confidence our merchants have in our abilities to process transactions and report on them accurately and pay them out accurately. That's really important. Furthermore, I think we've also talked about the rate of change. We are super Agile. We have subscribed to the Agile philosophy.
We rely heavily on automated testing, continuous integration, and we have weekly release cycles. Simply being able to listen to our customers, find out what needs they have, being able to implement that change and get it into production within the span of a week, that's always been historically a great advantage to us and will continue to be a great advantage in the future. How do we deploy this software everywhere across the world? Well, we have multiple data centers. What we've done is we've split up our processing capabilities into regional groups. We identify the U.S., Europe, Asia, and Australia or Australia and New Zealand as the main regions where we're processing, and we have data centers in those regions. We set them up into an active configuration in each region.
That means that we don't have a primary site and a disaster recovery site, but rather than that, we just route transactions to all sites simultaneously. That means that if we ever see a problem affecting one of our sites, whether that's a problem we have internally or it's a problem that's caused outside of our network, for instance, a power outage, a hurricane that might happen somewhere or some other event which threatens to compromise the stability of one of our data centers, we can remove that data center from processing without affecting the uptime of the entire platform. That provides a lot of resilience, and that provides a lot of scalability in our ability to take transactions. Why do we feel it's important to be in each region with our data centers? Well, it's a lot to do with merchants.
We want to be close to our merchants when they're connecting into us. Geographically close means you get lower latency, which is an improvement service level. Transactions go quicker, but they also have to travel less over the Internet. The longer a path is over the Internet, the more chances, the more hops it has to take, and the more chances there are for a service interruption. We don't want that. We also operate in a kind of a microservices model where transactions are entered, the merchant enters a transaction to us, but then we route that to the place where that transaction gets acquired. Where that transaction gets acquired is typically where the card holder is holding his data. That could be on the same continent, but that's not always the case.
The merchant infrastructure might be centralized, for instance, in the U.S., but they might be doing business in Australia. Ultimately, that transaction, which gets entered by the merchant in the U.S., needs to actually travel to Australia. That means that we can also have our hardware and our data centers near, for instance, in Australia to process that transaction locally. That transaction will flow from the U.S. where the merchant has a quick path to our infrastructure. We take on the message and pass it all the way down to Australia, and it's all within our visibility and within our scope of control. Then we push it into the network there. That gives us, I think, a better, allows us to achieve a better quality of service than really any of our competitors in that respect.
Because we have so many links between these data centers, dedicated lines running between those data centers, we can have an optimal path for every transaction. We monitor the health of all of our connections between the data centers, and what the performance is of all our acquiring network links. If we see that there's any degradation either in a network or in a connection to, for instance, Visa or Mastercard from a certain endpoint, then we route around those endpoints. We're always optimizing for giving the best experience to our merchants. What it kind of looks like, because I know this is a really abstract concept, is in the traditional routing configuration, you would see a merchant connecting to, for instance, a U.S. merchant would connect into a payment provider maybe in Europe, and that transaction needs to travel through a number of hops.
Then it goes to the data center in Europe, and if it actually needs to end up in Australia, then again, there is a number of hops. If anything goes wrong around here, the merchant can't connect into us. We don't know about it. We only know about it when the merchant calls us, and that would be a bad thing. Similarly, when we connect into an acquirer network, which we're not regionally close to, and we get something like a timeout or some problem happens, then we don't know if that's a problem in the infrastructure somewhere or it's a problem at the acquirer end. In order to eliminate that, we will route a transaction. The merchant will connect into one of our data centers, which is geographically close.
Depending on the optimum path, we can route that transaction either directly or through another data center link to the place where it needs to go. That just improves service levels and also means that if something does go wrong somewhere in our network, that we are able to route around it without having the merchant suffer the consequences. Some key things around monitoring and risk management. If you look at our resilience, I think the rerouting is a good example of where we're resilient, so talking about that. If you look at resilience from another perspective, we've gone to great lengths to make sure that the online transaction part, the part where the consumer is waiting for a transaction to complete, is actually completely decoupled from all of our back-end systems.
That means that if anything would slow down in our accounting systems, if one of our major databases needs some sort of maintenance, even if one of our data centers in Europe goes offline, where all this data is aggregated, where the reports are run, then it doesn't affect transaction processing. In that respect, we have a high level of resilience. In order to monitor that resilience, we employ cloud providers. We have extremely good internal monitoring. The internal monitoring is actually so good that most merchants, or merchants quite commonly these days know about problems they have in their infrastructure by us seeing drops in traffic and our support teams actually reaching out to these merchants telling them, "Hey, we see a sudden drop in traffic. Is something going on?"
Quite often, the merchants, that will be the first the merchant even knows about problems in their own systems. It's no good us noticing that things might go wrong in our system or that we can monitor our own platform or even that our monitoring is saying everything is okay and everything is up and running when merchants might have other experiences. This is something common to the Internet, is that we don't control every part of the Internet. If a merchant on their data center is trying to connect into our APIs and is having trouble doing that might be completely outside of our network. It might be on some hop, on some router, somewhere on the route between the merchant infrastructure and our infrastructure.
In order to get a visibility of what that looks like, we use cloud providers all over the world, and we have monitoring endpoints in every geographic region, and we cross-monitor our own APIs. That means that we're seeing the same picture in our monitoring, as merchants are, have experience in connecting into us. Rather than saying, "Oh, nothing is wrong with our infrastructure because everything looks good from our side of the world or our side of the pond," we can actually go and look at what does our infrastructure look like in terms of availability from where our merchants are sitting. Then we can make decisions about that as well. A good example is, there was a power outage in San Francisco last year, and we have a data center in Sunnyvale, which is pretty close.
The data center itself, obviously, because data centers have backup power, et cetera, the data center itself wasn't affected at all. In fact, all of our monitoring of that data center, even from Europe, didn't indicate there was any problem. We started seeing that monitoring from an AWS location in U.S. East, which is probably where a large part of our tech merchants are connecting in from. Basically, if AWS U.S. East goes down, for most Americans, the Internet is down because Netflix doesn't work and Twitter doesn't work and a lot of stuff doesn't work.
Once we started seeing that AWS U.S. East on the East Coast of the U.S. had trouble connecting into Sunnyvale, which is probably caused by something to do with the power outage, maybe an overload somewhere in a power station on some network router, somewhere along the lines, connecting into that infrastructure. At that point, we take that data center offline, and we don't route any new transactions to it, and we rely on our data center in Miami. This way, the merchants aren't affected by something which probably most other providers would say, "Well, this is outside of our scope of control. We can't do anything about it." We can take control a lot further. Security. There's just so much I could say about security, but I just want to highlight one interesting program we have.
I'm not sure if you're all familiar with the concept of penetration testing. Penetration testing is essentially when you as an organization get a couple of security researchers in, they are specialized in trying to probe your network and trying to probe your infrastructure and see if they can get in. A pen test is done regularly, of course, for the PCI certification, security certifications we do. There are also large merchants, very large merchants, who would like to do a penetration test to evaluate our infrastructure. That's fine, and we like doing that. To a certain level, and with a large group of merchants we have, if we were to accommodate all requests, we would basically be running two simultaneous penetration tests every week of the year. That would consume so much resources, we can't do it.
What we've come up with is a client pen testing opportunity program, which is a group of merchants who together would like to do the pen testing. They, as a group, submit proposals for interesting pen tests they can do with us, interesting ideas for doing these penetration tests. We select a number of these proposals and then execute them. What's different than just running the penetration test for that merchant is we then share the results of those penetration tests among the whole group of merchants so that they also get the insights achieved there. They can all get at least the data from the penetration tests. I think a question which comes up regularly is, yes, we have a very quickly growing business. How do you accommodate that exponential growth in volume?
I think it's funny to see that in the first approximately 10 years, it took about 10 years to get to 1 billion, the first billion transactions on the platform. The next billion of transactions took just about a year to do on the platform. Where you're used to seeing transactions come in and you reach the 1 billion mark, the next year you've already achieved the 2 billion. Within six months we were at the next billion. This is a huge amount of growth you need to accommodate. There's two aspects to scaling out this infrastructure to be able to make sure that you can actually handle all these peak volumes and all this growth. One is the parts where you're connecting into or the parts which are involved with taking that online transaction.
The moment you enter your credit card, you punch the submit button. At that point, you want that transaction to be processed. As I said before, that essentially runs in a kind of a microservices architecture. The microservices architecture means that you have individual servers which are in charge of, have a certain responsibility. You can add more servers to that to add capacity. One thing is important, though. They need to be what's called stateless. Stateless means that there should be no dependencies between the servers. The moment you create dependencies between them, that doesn't allow you to scale. Having a stateless architecture where each server can operate individually without needing to consult its peers, not needing to communicate about the things it's doing, that allows you to add additional servers and actually increase your capacity.
We get a lot of that for free because we've been expanding our data centers. Expanding your data centers naturally just gives you more servers. Having more servers is more capacity. Because those data centers, as you go along, also include newer versions, newer generation hardware, that hardware is faster. Not only are you gaining more servers, but you're also gaining faster servers. That covers the additional capacity we need quite nicely to be able to process peak volume of transactions. The accounting system is quite a different matter. You can't really scale an accounting system on a single database beyond a certain point, because at a certain point you will just exceed the capacity of a database system. What we've done there is we've clustered and sharded those databases.
That means that suddenly you're dealing with multiple accounting databases, which is fine in terms of raw transaction volume. You can scale that out by just adding more accounting databases. You suddenly lose the single point of truth because you're always able to say what's the balance of something or what does the financial picture look like on your accounting database. Now you have multiple accounting databases. A lot of processes rely on the fact that there's a single source of truth database. Now you have multiple of these.
What we've re-engineered a few years ago, we've re-engineered all of our core processes to work in a streaming manner, which means that they are getting these events which are hitting the database, not from querying that one database, but actually getting these events in from all of the databases together then acting upon that. That's a model that scales out really well, and it's easy to make parallel so that you can scale out horizontally. That means together, on the one hand, the distributed system we have with the microservices which we're expanding, the other hand, the scalable accounting system, that means we have no scalability concerns for the future. That's it from me. Thank you. Pieter?
Okay, back to sales and marketing. A little bit on the background. When we started Adyen, we were seven people, six engineers, and me doing sales and marketing. It's back to my roots. What do we do with our merchants? It's funny if you see the relationship which we build up with our merchants over time. It is a true partnership. That means that through account management, we know what's going on, we know which problems they have. Through the discussions with them, we are able to expand our business to their new regions. Of course, we also use them or help them by getting input on what to develop next. These are quite intense relationships. Sometimes we see merchants using our product in a way which we hadn't expected ourselves.
Think about a merchant. Because we track the transactions, we can see because we can actually link everything together because we only have one platform, we can see online users, we can see store users. We can see what happens to the online volume. If you, in a certain region, start with a physical location, what does that mean? Does that mean more transactions? Does that mean attrition in your online sales? What we can also see is if you then add another store, where does the volume of that store come from? Those are the things that we look at together with our merchants. Things like, how often is a shopper back in your store?
I've been in meetings where the merchant, for the first time, got insights into what their consumers were doing, which they never had before, because for the first time they could see that in their physical stores, shoppers were coming back three times a week, whereas they expected maybe once a month. You can see how often it happens that a shopper cannot fulfill the transaction because they have insufficient funds. You get all sorts of data, which is very useful for your marketing. That's the type of relationship that we have with our merchants. If you look at who we target, of course, we target the large international merchant. When is a merchant truly interesting in that segment? That is if you're in multiple countries or if you are on multi-channels. If you sell both online, have store, mobile, and all that together.
That's also where we started. We started off building Adyen with online merchants in multi-regions. Then later we added the point-of-sale to that. What's also an interesting segment for us is mid-market. Why is that interesting? Because what we have seen is by rolling out our services, in some markets, we just happened to become an important domestic player. How should you think about that? You start in Brazil. Which merchants do you get? In our case, often American companies starting to target Brazil and rolling out in Brazil. European companies rolling out in Brazil. That's your first merchant group. As you continue, at a certain point, the local merchant sees that the quality which they get with us is higher than what they can get in the local market. You become a domestic player.
After that would first be the largest domestic player. You see that the mid-market segment is also boarding on our platform. What we have done is we made it into a segment and have a specific strategy for those mid-market merchants. The third part is, of course, by working with all those different merchants, we develop additional products. Think about, we have a large airline portfolio, specific airline fraud. If you think about the large international merchants, there are marketplaces in there. We have specific marketplace products. Products around boarding small sellers. Payout products to fund those sellers. The different merchants that we have give us future new growth areas. Potential products which we can sell there because we work so closely with them. How do we find merchants?
For the large enterprises, for us, the most important thing is to get into a discussion. That can often be conferences and not necessarily payment conferences, but we like to be at specific industry conferences. Think about airline conferences, retail conferences. Those are the type of conference where we are present. The intention there is to build up the relationship. What we also do is knowledge-sharing events. That can be we take speaking slots at general events or we invite our customers at our events. I find the second part actually quite amazing. To be able to do very little and hear our merchants speaking to other companies which aren't merchants yet, how it works to roll out with us.
Because we have such an enthused merchant base, I find that the most amazing way of finding new merchants and starting to build a relationship with them. Mid-market is a little bit different because although direct marketing and directly signing up is an element, it's also important to be integrated with the platforms that they use, but also with building partnerships with the system integrators. That's what we're doing in that segment. The start of Adyen from day one has always been transparent pricing. Why transparent pricing? In different parts, in the U.S., the interchange is built up out of the fees of Visa, Mastercard, a markup, and then the markup that the acquirer gives to the merchant. Those three components we always have from day one made transparent to the merchant. Our large merchants know their markup, so they know what they pay to us.
Then there's the assessment fees, and then there's interchange. Because we have taken that approach from day one, means that where in different parts of the world blends were very common. We want to work with a merchant base where we say there's transparency and also in our contracts, it gives a discount table. More volume means that, of course, the average price per transaction goes down. It also means that the contracts automatically stay in balance with the volume that we get. That you avoid, which has been very common in the payments industry, that you avoid, that you actually, if you would take a bandwidth between where all the merchants should sit, that you actually know that some merchants are overpaying and that they're a struggle had. We tend to look a lot further, and we want to build up long-term relationships.
That's not a situation which we find very admirable, where traditionally companies actually thought overpaying merchants are fantastic. Pricing structure, we never changed. We ask a fee per transaction, and then we have the transparent markup. It's the model with which we started, and we've never felt the need to change that over time. What do we sell? In our discussion with the merchants, it's always about what is currently obstructing further growth in your business and how can we help with that. We don't do cost-based pricing, what we sell is the value of our product. Cost-based pricing doesn't really work for us because all the cost that we have is running Adyen, the cost in what we do in developing the product. A transaction in itself, of course, doesn't carry cost for us. It's value-based pricing.
We look at the value that we have with the merchant. If you see how a sales process runs, it's a lot of discussion. It's about what could we do for you and how much more business would you do if we indeed deliver on that? That's how we close the deal. What you see is that others cannot follow. Sometimes I hear back in the market that we push pricing down. Sometimes competitors do not see much else to do than to really drop pricing and hope in that way to keep the merchant for a bit longer. Ultimately, if you really do the math on it, almost always we can prove that we earn back more than the whole invoice. It doesn't really matter. What is important is how can we help the merchant grow the business and how much additional sales do you do.
There are different functions which we have working together. If you see how we structured sales is, what we don't want is that our best salespeople are stuck with doing account management. That's what you sometimes see. The ones who brought in the deals are nurturing their deals. They probably get commission on that, so you don't see them in the market again. We have made the separation there, where we say sales does sales, then it get handed over to account management, and account management grows the business. For us, a large part, above 80% of our growth comes from merchants which we land. If we land the merchant, it's always with a subset of the business. It's land and expand. The account manager is actually the one who builds that out from the initial transactions, in a certain region.
Say, a merchant just goes live with us online in the U.S., but we know what will follow is online in Europe, stores in the U.S., stores in Europe, and that whole rollout process is done by the account manager. Marketing. If we look at marketing, what has been specifically very effective for us is marketing around events. Think about takeovers of subway stations that we do. What we do in marketing around events can also be that we throw dinners, that we host, that we take over the coffee area, and there we have 10 specialists there who start the conversations. The way how we have set this up is to be able to sell in a country does not mean that we need to have an office there. It can be, of course, helpful.
If you see how we built the company, what we do to go to a country is first we do sales in that country out of another part of the world. When we see that we have a certain base, we consider if it's logical to have a domestic office. Our offices are sales offices, account management offices, customer support offices, and then dependent on the size, it can have local marketing, it can have local compliance, and it can have local engineering. The main engineering hub is in Amsterdam, but dependent on the size. If you look, say, San Francisco, which is about 100 people, then all of that will be represented in that office. I think in total, these are 18, I guess. We're now at 20. This is, of course, something. We're always looking at new markets.
Sometimes we have a person on the ground there testing things and seeing if we move on. If you're smart to go through LinkedIn and you think you figured something out, unfortunately, that might not give you so much data because we are constantly looking what's the next step. How do we reward our salespeople? Salespeople are on commission. And the structure which we have there is a plan which has evolved over time, which I've been running for a very long time, which is based on the actual margin that the company gets in, and is over a long period of time to avoid opportunistic behavior there. What we do there is we ask people to work together in different regions. We do have what we call open skies. Salespeople can sell in all over the world. We have one rule.
If you are, say, in the U.K., and you sell in the U.S., then you need to involve a U.S. office to avoid that you do things which are not very cultural sensitive. We, of course, run the central CRM where you can look that you're not starting to get involved in a merchant which we already have discussions with. In our sales process, we involve others. We want all engineers to be at least once a year with a merchant. We have account management involved, so when we go to a merchant, it's not a single salesperson. It is, of course, the high end of the market. It's not a single salesperson, but it's a whole group that help the merchant, and that looks through how we can improve life for them. The autonomy we spoke about.
The autonomy is you can sell in the market that you want, the product is very defined. What we sell is a product that we have, and if we feel there should be a change to the product, we always look at, is this a change for a specific merchant, or is this a change which all merchants or at least a larger group of merchants would like? Specific implementations for a single merchant, we don't do because we don't feel that's scalable and that would create our own legacy. If you look at how that work in practice, in account management, I would think it's good to think about an example. Say you are a large merchant and you run into a specific fraud pattern. What you now suddenly see is, hey, I'm under fraud attack, what to do?
When you call with Adyen, or maybe we see it even before you see it, because that also happens as we monitor, we are able to look at the pattern, design something for it, and take it live within hours. We can monitor, did that indeed solve the problem or do we indeed make another iteration? That's a collaboration which we then see with account management, engineering in a specific workstream. In this case, it would be the workstream responsible for the fraud system. That's why we have such a specific relationship with our merchants. A little bit more about marketing. What is our ultimate aim in marketing? In marketing, it's of course name recognition, but it's also we have thought leadership. We have thought leadership, obviously, in unified commerce.
There is, for retailers, so much challenge if you look at how quickly the environment is changing, that they are looking for us to tell them what is now, what's possible. Who is pushing the borders of what can be done? Through white papers, through meetings, we tell them and we take them through what's currently possible, and we lead them in the next phase. They are actually really exciting meetings sometimes, to hear about the challenges which they have, the challenges which you have in luxury retail, for example. Where do we run our campaigns? Outdoor advertising has proven to be more successful for us than we thought. Of course, we do everything, well you see here the list. What I like, specifically, are the Adyen events where we have specific themes where merchants talk to each other. Let's have lunch. Any questions so far?
You can. Oh, sorry. I'm Hemmo, for those of you who I haven't met yet. Hi. We've got lunch at the end of the hotel. As you walked in, just go back down exactly the same way. We've got a nice Indonesian style lunch, and we'll start back here at 1:00 P.M. When Ingo and Pieter conclude with the next section of the presentation, we'll have about an hour reserved for Q&A. We'll have plenty of time then. Thank you.
[Break]
Hi everyone. Hello. If everyone could take their seats, please, we can get started again. Thank you guys. All right, thanks all. We've got three more subjects. We've got team and culture coming up, financials, our growth strategy. After which we'll have about an hour for Q&A. We'll have a mic in the room. People will go around and you'll have every opportunity to ask the management team all the questions. For now, here's Pieter van der Does with team and culture. Thank you.
Thanks, Hemmo. There's a clicker. What we did in Adyen is we started talking about the Adyen Formula. What is the Adyen Formula? The founders of this company, all seven people of the original team, we had worked for a company before. What we noticed is that when we started building this company, more and more, we talked about stories in the past, stories about early Adyen, how we run our business. It's nice to do repetitive tasks, but it might also be good to write down a few thought lines behind it and tell that directly to people. Whereas a lot of companies have something aspiring, we have the Adyen Formula, which is very practical.
If you run through the Formula , this is something that always comes back in every yesterday, we had all engineers together, for sure, you know there's going to be talk about the Formula . Every meeting that we have in discussions, we talk about the Formula . What's the thought line behind the Formula ? Eight sentences are not really going to do it. It's our answer to the question: how do you run a global company and run it with speed? Actually, the Adyen Formula is about speed. Let me give you an example. We pick up the phone and don't hide behind email. It's very easy to send an email off to the U.S. asking for something, ending with please advise.
The problem is that that email has a high chance of getting an aggressive email back, somebody has to get involved, you spend a lot of time on actually finding out what was the intention of the sender, what was the intention of the receiver. If you would have picked up the phone and have an adult discussion, suddenly you see that you get much quicker to a resolution. We find out that if you work with 20 officers around the world, you should discuss. We say here through email, of course, we also invested in video conferencing stuff, have dedicated rooms. It's really easy to walk into a room and to see your colleague who's at the other side of the world and have a normal discussion.
Each time that an incident happens because of email, we think it's unfortunate and that we should avoid doing that. I also do it in my daily life, by the way. The balance between being, the Dutch are known for talking straight to the point of almost being rude. The challenge there is how do you do that in an international company? We say to everyone, it's of course within your culture, but say how it is. It is something which our merchants appreciate. If something goes wrong, you will hear from us what happened. Where often those discussions are more about proving that it's not your mistake, we are much more neutral. We want to be transparent internally to our colleagues, to each other, transparent to our investors. That's why we have this day.
I hope that you also feel that we give you all the data to make up your mind, also we want to be transparent to our merchants. It has helped us a great deal. Decision making. If you make decisions only at the central level, at a centralized point in the company, you are working in a quickly changing environment, that is not the way to structure your business and be ready for a quickly changing environment. You have to find something else. What do we say to people who are working, say, in a workstream, encounter a problem, and feel that they should do something? We don't want individuals to take a choice, to make up their own mind, because that's error-prone.
What we say according to the Adyen Formula is discuss this with a few people within your team, but also in other teams to find out, is my idea actually as good as it is? If you have discussed that and have collected the different views from different people, and they're all positive about it, please execute it. If you don't do that, please don't. That is related to speed. That's a way how you structure the company and can capture that. Winning is more important than ego. I think all of us are proud if we have a certain accomplishment, but it doesn't work really well with getting input from others and claiming things. We know that if a salesperson stands up and say, "Look, the big contract I brought in," that the engineers will think, "How difficult was it? I built something so brilliant.
Don't be so proud of yourself." We are more focusing on celebrating releases, celebrating the progress of the company, and not celebrating individuals. It's a choice which we made together, and it's something which we consistently, as a board, need to push. Because before you know it, you have national organizations who suddenly start building their own P&L and who are saying, "Maybe you have that customer, but actually we landed it." To be able to run this, we run it as one company, as a global company. Each Adyen office, if you walk in there, it's very similar to another Adyen office in another country. To run this in one culture, we do exchanges that you move from office to office. That's the way how we do it, that is winning is more important than ego. We pride ourselves in our wins.
Everybody is very proud of this company rather than just proud of his or her own contribution. Oh, it does not move. Oh, now suddenly it does. If you look at the management team, it is quite specific how you do this if you do not want to have many layers. We have six people who are board members, and around us, we have collected a group of people who are specialists in their area and who work very closely together with us to make sure that we have a wide team where you can go to sharpen your ideas. Some of them have been with us for a very long time, but what I think is very powerful is that you also see there are new joiners, which very quickly run through the organization, have worked in different departments, and are being added now to this slide.
I think that Bert Wolters in the right here. He started with us after university, did an engineering degree, started working on data, and then very quickly moved up to be responsible for a much larger part of our technology. We also have people that are almost there from the beginning. It is a very good mix. In country, we have people who are very If you are responsible for the U.S. region, like Kamran is, you are of course instrumental in our growth, and we are working very closely together with those regional managers. Kamran, for example, has been a payment manager for Netflix. His knowledge about the merchant side of payments and in-depth knowledge about payments is extreme. Everybody within that group has in-depth knowledge. Is this where I hand over? Well, there you go. Martine is Head of our HR.
Thank you. Hello, everyone. My name is Martine Buis. I joined Adyen early 2015 as Head of HR. When I talked to Pieter and Ingo and Arnout by the end of 2014, I was very impressed by these guys. Super smart, inspiring, and so dedicated to build this successful tech company that I simply wanted to join this team. That is what I will be talking about. How have we been able to build this team, and are we still able to attract and retain the best talent while we scale? I go one slide back and talking about this management team. What makes this team unique, I think, it is the real team. There is a lot of payment knowledge, and they really live the Adyen Formula, which makes my job as an HR director pretty easy.
What I think is also very important, and that is what I will be talking about, we as a team are extremely capable of building a strong team around us. That is what I will be talking about. How do we grow the team? By the end of 2015, we had about a little bit over 300 people. Last year, we had almost 900 people. And you see that is extreme growth rate year-on-year. This growth rate is far behind our revenue growth rate. We deliberately take this approach of measured growth on our headcount. We do that for two reasons. First of all, we just do not need so many people to scale our business. Secondly, because we want to keep that unique Adyen culture or Adyen Formula alive. That is so critical to the success of our business.
We are very focused on who we hire, and when we hire. That's also highlighted by the fact that a lot of the people from the first days, like Michiel, especially the developers, are still with Adyen. Also it is highlighted by the fact that the number of regretted losses is very low. It's below 2%. I think that's due to a very rigorous hiring process, but also because of our great working environment that we offer. Let's look at our talent pool. I think it has been mentioned before, in essence, we are a tech company. We have a focus on sales on the commercial side. Still 80% of our people is in tech and sales or commercial roles.
I think the cool thing about this is when you compare these numbers to last year, the number of tech people and the number of commercial people, that percentage has gone up, while the percentage of other function is going down. This shows me that we're still very focused on continuous growth of this business. When you look at our talent pool, it's average age of 32, so pretty young. That's even excluding our interns. What is also very unique about Adyen is an extreme international organization. We have more than 70 nationalities working in our 20 offices in our 17 countries. How cool that is, I realized yesterday, Arnout was organizing his fourth tech event here in Amsterdam.
We had about 400 tech people together, I was talking to developers from Zimbabwe to Norway, from Colombia to Shanghai, and they were all talking about the cool development stuff they were doing. That, I think, is what makes Adyen unique. We embrace those different people with those different backgrounds because we believe the best ideas come from diverse teams. That's also why we have the Adyen Formula line stating, "We include different people to sharpen our ideas." You constantly check with each other. Is this the best thing to do? That's also what we do as a management team. We constantly check in with each other to make sure we do the best. How do we find people? I already stressed that we deliberately take an approach of measured growth.
We don't want to grow that quickly in headcount because we don't need to scale this business, but also because we want to keep that culture alive. We are very selective. We can afford ourselves to only hire the best out there. We want to keep the bar high at the door. We don't want to hire the five, because then tomorrow we cannot hire the nine anymore. How do we do that? We have those three selection criteria we use. First, most important, it has to be a good formula fit. People have to fit in our culture. They need to be team players. They need to be humble. They need to talk straight. All these elements need to be there, because if that's not the case, you won't be successful at Adyen.
Secondly, we want to hire the smart people, the eights, the nines, and the 10s, because we have that environment where the really smart people can flourish, and they will be successful. Thirdly, what we're looking for is people that can make an impact. We want people that build the business with us, people who enjoy building and not people that come here to do a job. We want people that get things done, people that hit the ground and run. That's what we're looking for. If we hire all these people, where do we find them? There are a couple of sources, but there are two main sources that I want to highlight, and that are very important for us. That are the sources that really provide us the opportunity to really get to know the candidates.
One of the most important is our referral program. We introduced the program in 2015, Everybody can refer someone. You get a nice referral bonus. The only prerequisite is that you only refer someone that you for sure know it's a good formula fit. Almost 40% of the people we currently hire is from that referral program, which is a really high number. People enjoy referring because it's great working here, and they want their friends and the people they know working here as well. Another interesting source that we use is our intern program. We have set up a pretty unique intern program starting by the end of 2016. We provide bachelor and master's students the opportunity to work on a part-time basis at Adyen, and mainly in support roles.
We have those smart kids working for us in support roles, and what we ask them to do is to improve and automate our processes, to really keep on innovating from day one. 50% of those interns we hire on a permanent contract later on. I, in my team, have a good example. One of the first interns that we hired was in the recruitment team. That was two years ago, a little bit, two and a half years ago. She is now leading our operational support team. From two years from university, now leading operational support at Adyen and being responsible of 25 interns herself, which I think is a great example of how you can grow your career within Adyen, but also how we can really find good candidates who are successful later on within Adyen.
What tools do we use to assess our candidates? Of course, we do all the normal things. We do the normal CV screening. We have a coding test for our developers. We also do things a little bit different because we're Adyen. We do it the Adyen way. One of the most important tests is the airline test. What does that mean? That means that we ask everyone who's interviewing, always test, would you like to spend a flight with this person, with this candidate, to one of our biggest hubs, to San Francisco or Singapore? That means, would you enjoy taking a flight from 10 to 12 hours with this, sitting next to this person? If you enjoy, probably your colleagues will as well.
If you don't enjoy sitting 12 hours next to this person, how would you ask your colleagues to spend 40 hours a week with this person? It's a no-go. This test has to be passed by all interviewers, and it has to be positive. Another specific Adyen thing is that every single candidate in the final stage sees a board member, which I think is pretty unique. The reason for that is because we want to make sure that it's a good formula fit. Also give the candidate the opportunity to meet with a board member, but also to give the board member the opportunity to really get to know the person and later on have that relationship. With these tests, we want to make sure we keep the bar high, and I think we're doing a pretty good job.
We only hire one out of each 171 people applying. We're very selective. To give you a little bit of a background, when I was running these numbers with the team, and I was looking at the numbers before and after IPO, it was really cool to see what that did for our employer brand. Before IPO, we had about 18,000 people applying every year. After IPO, it's 50,000, so it more than doubled. From an employer branding perspective, we are no longer that unseen Dutch tech company. Now we are out there, and people know to find us. I've been talking about our team. I have been talking about how we find talent and how do we keep talent. How do we retain people? I think that's evenly important.
If you ask people within Adyen, "Why do you enjoy working here?" I think it's these three things. It's the opportunity to create your own path. We don't have standard career paths. Instead, we provide a lot of freedom, and we provide a lot of autonomy. We have a lot of growth opportunities, like the workstreams that Pieter was mentioning. We have the Adyen Academy. You have incredible merchants to work with. We have an exchange program, international transfers. A lot is possible, but you have to go and get it yourself. It's your responsibility. Secondly, why we enjoy working for Adyen, it's just that feeling that we're in there and building the business together. It feels like a once in a lifetime opportunity to change this industry together, which makes it just fun to work here.
Lastly, and I think we stress that quite a lot, it's our company culture. It's the Adyen Formula. No matter if you enter the office in Shanghai or in Mexico or in New York or in Berlin, if you enter, it feels like Adyen. You feel the energy. You feel the speed. You feel the merchant focus. It's just fun working together and make this happen. That's what Adyen is about. Thank you. That was my story about how we built Adyen and how we're still able to attract and retain talent well before the scale. I now hand the mic to Ingo, we're all waiting for
Thanks, Martine. What I'd like to do is give an update on the financials. Financials haven't changed, so it's going to be a bit boring story. What I'd like to do today is give a bit more insight how we book revenues, how it exactly works, our P&L. I think moreover, explain that we still have a one P&L philosophy. We built this company as one global company, and the only thing that really matters is how much margin we make per customer. We don't make P&Ls per region or per office. We think it's not relevant, and that's also what we really don't want to do. I think this is the perfect summary of our business model. If you look at processed volume, you see a large increase because we focus on helping our existing merchants to grow with us.
Most of our revenue growth comes from existing customers that were already live on the platform. Of course, like Pieter explained about our sales strategy, it is finding that right balance between, at one end, building with our existing merchants, at the other hand, finding the right new customers on our platform that are your security or your insurance for the next year's growth. The increase in processed volume is translated to net revenues. Net revenues is the key KPI that we focus on. How do we make sure that the absolute number of net revenues increases over time? How do we make sure that if a customer brings additional volume, that the total amount of absolute margin grows? This is not a relative margin business. This industry is often looked at take rates. The net revenue is divided by processed volume.
We're not managing on take rate for that reason, because we only believe in absolute margin. I will come back on that in a second. Thirdly, EBITDA. I think it reflects the increase in EBITDA, how we built this company. Revenue and cost are unrelated. Revenue is the result of the growing volumes, and cost is mostly cost of the team. As volume grows quicker than our team, you get economies of scale. We are at EBITDA of EUR 182 million by end of 2018, and this has grown significantly. If you look at how we book revenues, above the net revenues, there is a gross revenue line. Internally, we sort of ignore gross revenues. The reason why we ignore gross revenues is because it includes the interchange and the scheme fees that we pay.
The interchange is the fee that we pay to the issuer, and the scheme fee is the fee that we pay to the networks. The reason why we ignore that number is because interchange and scheme fee is passed on to the merchant, and it really differs per region. The interchange in the U.S. is almost 10x higher than in Europe. European Commission has capped the interchange. In the U.S., people or the issuers like to give big rebates to credit card holders. That's why you see that the interchange is so much higher. If you look at our gross revenues, it includes processing fees. That's the fee to use our gateway. It's a settlement fee, which is basically the fee for using a payment method like Visa. That also includes the interchange and the scheme fee, and its other services.
Other services includes terminal service fee for using our terminals, also the fixed income. If we deduct the interchange and the scheme fees, you get to net revenue. From net revenue, we go to operating income if we deduct mostly cost of the team. Cost of the team is still about 70% of our total cost. If we add the depreciation amortization, we are at the EBITDA level. If you look at a single payment, on the top, you see a shopper making a EUR 100 payment on our platform. What does the merchant in the end get? Well, in the end, it gets, in this example, EUR 98.24. The way how we get there is that we need to pay EUR 1 to the issuer, which is the interchange fee, EUR 0.50 to the networks, the scheme fee. We receive from the schemes EUR 98.50.
We deduct our fees, in this example, EUR 0.06 as a processing fee and EUR 0.20 as a settlement fee, so EUR 0.26 in total, you get to the EUR 98.24. The Adyen fee, the green box, that's what is net revenue. That's also where we focus on, to increase that number. That's the key focus of how we manage our P&L. A very important measure of our company, of the health of our customer base is the concentration. Concentration has gone down over time. If you look at current numbers, for processed volume, the top 10 is about 36% of our base, which is slightly lower than the year before. The same for net revenues. The top 10 is about 31%, which is almost slightly lower than last year. We think that's normal development. We are onboarding more and more bigger customers on our platform.
Our expectation for the future is that this concentration will further go down. On processed volumes. The key thing of our account management group is to make sure that we work on a new project with our customers. We really want to focus on, if we work with our largest merchants, on their plans for next year and see what kind of projects we need to implement. That's what the account manager does. The account managers works with the merchants, also tries to understand what it means for a development perspective, if we need to develop certain functionality. That's why we can have this growth. That's also why most of the growth in processed volume is coming from these existing merchants.
The net revenue growth, what we have seen over the past year, although we do not manage on take rate, we see an increase in take rate because the full-stack volume on our platform is increasing. The fact that we are in multiple regions now, both the gateway and the acquirer, is helping us in our income. The strategy of building that global platform that also Edgar was talking about is helping us because you see that the share of full-stack on our platform is increasing. Because that share is increasing, we earn more per transaction. That's a very positive development on our take rate. Take rate is up to 22 basis points.
Of course, for the future, it really depends on how the base further develops, because we think that if we board big merchants on our platform, that it's a positive if volume increases, that the fee per transaction declines. That has a negative impact on take rate. That's also exactly why we focus on the absolute margin. We really want to make sure that if a merchant brings more volume to the platform, that the absolute invoice with that merchant is increasing. If we work with an account manager, their main assignment is to have an increase in the invoice over time. That's how we look at it, and that's also how we continue to look at it, because you want to be in a position that if a merchant is very successful, that you can have a negotiation about the relative price per transaction.
From a volume perspective, because this graph reflects the volumes, you see the development in 4.5 years. 2.5 years in 2017, 2.5 years in 2018, how it increased over time. The 70 and the 89 on the right-hand side, this is the EUR 259 billion processed volume. Still most of our volume comes from enterprise. That's how we started off the business. The majority of volume is initiated in enterprise merchants. What we do see, though, is that if you look at the mid-market, which is a new initiative, we see an important increase. It is now at EUR 2.1 billion on our platform, coming from EUR 1.9 billion over the first half year, and it's getting to 2.5% on our platform. We see that the mid-market strategy, you see the first signals that it's taking off.
At the same time, it's still very early stage. More development we see, of course, in unified commerce. Unified commerce, we only started this initiative about three, four years ago. It's now 11% in the second half of this year of our volume. I think that's pretty unique because we basically built in two, three years' time a 11% volume on our platform, a completely new business, because these are transactions that are initiated on the point-of-sale. There's no e-commerce transactions, are really on the terminal-initiated transactions. This gives, I think, a very good sense for how we can further grow this number because we have very high expectations of our point-of-sale proposition. I think the numbers reflect that this increase is there. On the regional side. If you look at net revenues, how is it distributed over the different geographical regions?
We look at billing address. Let's take a big customer, Facebook. If we work for Facebook in Europe and U.S., we send invoices to their European entity, which then recognizes European revenue, or we send a invoice to their U.S. entity, then it's recognized as U.S. revenue. If you look at the graphs, Europe is still most important. That's of course how we started off this company. We started off this company here in Europe also with a European acquiring license, That's why we have the main base there. We started to add, though, all the other regions, specifically, if you look at the dark green boxes, that's the growth that we had in 2018. You see that North America and APAC are very important for that growth.
The highest relative growth is in those regions, we of course continue to invest in those regions. Here you see the benefit of having one platform, because the international merchants that work with us, that started off in Europe with us, it's for them very easy to switch on U.S. or to try us out in the APAC region because there's no different technical integration needed. It's the same platform with the same procedures. You see more and more merchants testing us in the other regions. On operating expenses. Cost of the team is of course the most important cost driver. The measured growth strategy is not related to the fact that we can't afford it financially, but it's more culturally determined, that we want to take a responsible way of growing the company. Of course, it helps in economies of scale.
The other costs that are important to our business, of course, the sales and marketing cost. We started to advertise the last two years, specifically around retail events, because we see that retail events, where a lot of retailers come together, that it has a real impact if we have brand advertisements out there. Of course, platform costs are important. We host the co-location in third-party data centers. We don't have any type of cloud infrastructure with commercial clouds except for the monitoring that Michiel was referring to. Also, the platform cost is relatively low and very good manageable, because it's our own infrastructure, and we can build it and manage it ourselves. Of course, the key question that we always get is, okay, how far can you bring the EBITDA margin?
Of course, the second half of 2018, you saw an enormous impact of EBITDA margin growth. We think that it's better to look at the full year, to get a good sense for where we are from an operating perspective or operating leverage perspective. The 52% EBITDA margin in 2018 is what we believe a good reflection of the economies of scale. Of course, if we further grow the volumes, and we take a measured growth perspective to the teams, we can grow further towards the guidance that we have given. At the same time, we keep on investing. We see a lot of opportunities out there. Well, with opening of new offices, of course, we continue to expand the sales team, but also what you saw in 2018, we added a lot of new engineers. Below EBITDA, we have relatively limited cost.
Of course, the main expenses is related to tax. We have a strategy that we always pay taxes locally. We try to limit the tax optimization to have a very clear visibility in each country and have limited risk from that perspective. That's why also net income has further increased in line with EBITDA. EBITDA and net income, also from a growth perspective, are very much in line. On the CapEx side, we guided on 5% of net revenues. The reason why we can be so efficient is because we have one platform. We strongly believe that within that 5% of net revenues, we can further upgrade our platform. We don't expect any peaks in investments, what you sometimes see with competitors, because they need to replace a certain back-end.
We believe that with these investment levels, we can scale the company to the next levels and I see no problems there at all. Of course, the key question is, okay, what does it mean for free cash flow? We've defined free cash flow as the EBITDA minus the CapEx. Of course, you get then to a very high free cash flow conversion. The reason why we ignore working capital is because the most important part of our working capital is the receivables that we have on financial institutions. On the other side, it's the payable to the merchants, so you can net that out. The impact of working capital is relatively limited. That's why we keep this simple definition of free cash flow, and we get to these free cash flow conversion numbers. This is very important for us.
When we IPO-ed the business, we said let's take a medium- to long-term view on this business. We, as a management team, are long-term committed to this company, and we want to grow the business like that. That's why we also haven't changed the guidance, although 2018 was a very successful year. We believe that fundamentally nothing has really changed to the company, and we can continue the business in the next years, in this pace. We want to avoid a short-term focus. Of course, we get the question, why don't you guide on 2019? Well, in payments it's relatively easy to make your guidance for a certain year, but we really want to make sure as a company that we do the right things for the long-term, because that's how we like to build the company.
We have a unique position in the market, it takes time to get there. We have the time. That's also why we stick with current guidance. I hand over to Pieter to say some final words on our growth strategy in the different market segments.
Right. Thanks, Ingo. I wanted to elaborate actually on something that you said, which is it's easy for us to make our numbers. How we build Adyen, how we grow this company is with a long-term view. There are multiple ways to short-term look better, things are very obvious. View away from Adyen Formula. We could develop something for that one merchant and then board that volume, show it to you. Everybody will say, "Good idea." We know that long-term, we are getting things in our technology stack that we don't want to have in there, that we need to maintain, which might become a security risk in the future. We know we don't want to do that, you could do it if you have a short-term focus.
We could take on volume of merchants which we think are more in the high-risk area, like start doing acquiring of airlines, and then of airlines which are a little bit more risky. We have everyday opportunity to do something which we find long-term off strategy. Short-term, it could be a good idea. We don't run this company short-term, hence we don't guide short-term. Let me quickly wrap up. I think we covered almost everything. Why do we win? We win on both sides. We have the better team because it's just more fun what we do, because we have the single platform. Actually if you develop something, it can be taken into production. You get to work with the nicest merchants. It works on both ways.
You can work with such a good team on such a simplified problem, you're actually getting traction, which is very difficult to compete with. We have made a few good strategic choices. One of them was a few years ago, let's get into point-of-sale terminals. Let's make sure that we don't work just online, but with the vision as what's happening in the physical world with the point-of-sale is actually a repetition of what we have seen a decade ago online. That is the terminal is actually becoming an internet device. The terminal should have the payment methods of the world, and it doesn't make sense anymore to have something which is locally solved. This is something which as a company, you should solve at the full-stack.
Long-term, you can wonder if the terminal looks like what it does look like today, we want to be in that business. We want to do everything. A few good strategic choices with a simplified problem with the most talented people of the industry. That is why we win. Who do we win from? Of course, we win from the online companies. Think about the Worldpays, the Paymentech, who sometimes also have, or they both have also parts of their business catering for the physical world, but not integrated. Those are large suppliers for us. Remember that we were late in the game. We started processing transactions summer of 2007. That means that all our merchants came from somewhere else. All our merchants are actually dislodged from competitors, landed with us, like the service, expanded with us.
Also, the ones who are missing are the banks. Banks, often in the point-of-sale environment, if you are a large retailer, say for the U.S. market, you typically have terminals from three different banks. Suddenly you're going for one global solution and those three banks see that volume is gone and will not return. Local banks, often people think it's just about the online players. No, it's the more traditional bank that loses out because of us. The third one is what we call SME enablers. What are those companies? They build a nice layer over the existing rail. Easy to board, looks really good, doesn't really outperform, but it's nice and easy. It's convenient if you're small. You're usually local, so in a single market. How effective it is, how do you measure? If you do three transactions per week, what's a 2% uplift?
You cannot see that back. They're more focused on marketing, on the quick boarding, and they're less focused on the true quality that we build by owning full-stack. What happens there, some of those companies grow and then they outgrow that supplier. Suddenly those things being more regional, so limited in payment methods, limited in authorization uplifts that we can do, limited in unifying the channels, looking for specific services which we can offer. Suddenly that starts hurting with their growth and then they become interesting for us. We also see that as a third stream coming to Adyen. If you look at the merchant wins, you see that we have each year landed merchants which then expand with us.
Nike was already four years ago, if you look at what we announced last year, H&M live with us in the U.S. with their online volume with a few stores in Europe. They're at the beginning of a journey to give us more and more volume. The only thing that stands between that and having it is time and of course, for you to judge execution risk. If you look at this is a pattern which we constantly see. We lend a merchant, over time it gives us more volume. If you look at the current pipeline, you see that this process can run for a long time. Ingo already explained this, I'll do it very quickly.
The three areas where we see growth, it is the point-of-sale part, which we call unified commerce, because all those merchants do it with the intention to do everything with us. Sometimes they start with terminals, sometimes they start with online, we don't want to calculate them into that group if they do a single terminal. We calculate in that the terminal volume, the unified commerce. We have, of course, the traditional enterprises. Enterprises is everything. That's also for us, the marketplaces, we see them in that segment. Mid-market, we spoke a lot about that. It's the natural next segment. They were already boarding. We have very high Net Promoter Scores, but the Net Promoter Scores in this segment were even higher than in the mid-market.
What we do is specific approach, specific approach in marketing, specific approach in making sure the onboarding and everything works really smoothly. Specific approach in working together with the suppliers for that market segment, be it integrators, be it platform. I think it's a good idea to do Q&A. I want to invite you all on stage, at least from Adyen, the employees.
We've got two mics in the room. One I have, and the other one's held by Anna Marije Sandman in the back. Please wait until you have the mic to ask questions. Anna Marije, if you could just reach the first person. We've also got our Chief Strategy and Risk Officer, Joop Wijn here, who did not present today, but he'd be happy to answer any questions you may have. Lastly, I'd like to introduce Ethan Tandowsky, who will be joining us in the investor-facing team. Anna Marije, Hemmo, Ethan, those are the names you'll mostly interact with when you're not talking to Pieter, Ingo. This is us. It's good to put a face to a name, I think. Yeah, first question from Anna Marije, and I'll hand this back to Pieter. Thank you.
Great. Thank you. I guess as a question for Pieter, you describe yourself obviously as more of a technology company than sort of a payments business. If you think of the sort of technologies that you have, obviously the single platform. When you go to pitch to the customers, what's the sort of the distinct advantage in terms of cost and efficiency that you can get using the Adyen platform, perhaps in some numbers versus some of your competitors? If we take that sort of forward, as they build on your platform, beyond just payment capabilities, what are the other, and you alluded to some of the data and the analytics. What are the additional products you can build which are not necessarily directly linked to payments, but in an ancillary kind of software area?
It is unfortunately for our sales, not as easy as that you just go to a merchant and say, "Look, you have now an authorization rate where you have 85% of your online transactions approved. Move to us, it will be 87% for you to calculate your additional profit. There you go." In reality, what is going on there are always different elements to that. Authorization rate can be something, specific fraud patterns can be something, but what is more important is that merchants choose us to be future proof. What is nice is if you integrate it with us and you move to an additional region, or if you even do an additional channel, you never have to reintegrate. This means that this is a choice which is future proof with a track record of constant innovation.
You can, for example, say, I want to have Alipay and WeChat Pay on my terminal. Actually, you also want to have on your terminal what will be released next year and the years after. Whereas authorization rate improvements are important in our sales process, the decrease that we can make in overhead in your financial department, taking frustration out of the shopper environment, which directly relates to customer support, cost for the merchant. Making the process smoother, making it future proof, giving what we call subscription to permanent innovation. That is what makes the merchant choose for us. The second part of your question, which is related around how do you look at additional things that Adyen is developing? We are productizing it. That means that we making it ready to charge for it, and at a certain point we probably will.
On the other hand, it's in our interest to make sure if you're a merchant on the Adyen platform, that you enjoy the full functionality of the platform. We don't want to put too much obstacles in there, but we know that at a certain point when we are further and when the products are further crystallized out, that we have set them up in such a way that you indeed price for that.
Right. Just one quick one on unified commerce. It's around 11% of revenue today. Any incremental data in terms of the size of customers, where that is skewed to, or in terms of customer penetration today of the customer list, how big that is today?
I think in general, if you look at the point-of-sale customers that come on our platform, they often start with a certain country and then roll it out. We have seen very successful migration paths. That's a good thing of point-of-sale volume. The moment that you have them with a store, they don't have an alternative next to it, which you sometimes have with e-commerce. You get the full volume, and I think we are on a trajectory to see that more and more and more. We prove also with the distributed architecture that we can get to very high uptime. Having a second provider is just not necessary. Yeah, we feel very confident that this is the right strategy.
I think if you look at the type of point-of-sale customers that we board, these are typically either the point-of-sale customers that have a multi-channel strategy. They see the benefits of unified commerce, or they are very international. The problem is if you're international with point-of-sale, the old solution is that you need to integrate with local banks in each and every country, which from a functionality perspective is a nightmare. From a reporting perspective is a nightmare. I think Burberry is a case that we used in the past to prove that they brought down the number of acquirers in Europe significantly because they started to work with us. If you think about the savings in total cost of ownership for running your payments stack, that is very, very important.
Great. Thank you.
Thank you. Sandeep Deshpande here, J.P. Morgan. Quick question, Pieter. When you've, over the last few years, you've introduced various innovations on your platform, such as MarketPay, and then you've got the banking license, which allows you flexibility on settlements. Clearly you have a long-term roadmap from here as well that these are the new features you want to get into, and one of these is the mid-market that you're getting into. Are there other such things that new areas that you want to get into and that would be nice to get into in the next three to five years for the platform? Secondly, because for instance with the mid-market, does it change the cost structure of the company at all, the way you have to address the customers, given that historically you've gone directly to the big enterprise customers.
Now you may have to go through some resellers, et cetera. Does it change the cost structure at all?
Yeah. If you look at the first part of your question, so what else and what next? Obviously we are looking at new regions. Nothing to announce there yet, but that is always on more regions where in our ideal scenario, that means being able to process end to end because we know that that's the most powerful way to help our merchants, has the highest quality, is not error-prone, has the highest conversion rate. That's something we want to expand and also be present in countries to help merchants. That's always something we work on. Of course, there are also ideas about new services or extension of services.
There are nothing I can announce today, but in the back of my mind, if we talk about cost, we want to be able to free up resources, which is in our case actually almost always HR, that have overhead. At a certain point, we had about half our engineers working on point-of-sale. We do make huge investments in time to develop new products, and we see that there is enough room to continue to do that. This is not that we're entering into a phase of small optimizations. We can still make wild new things.
In your last presentation, you talked about your competition a little bit. Has there any competition developed which does this full-stack and thus when you go to bed, you think that this could be a risk to us three years or five years from now, just as we took share from all these incumbents in the past?
I think it's nice to have that we're the only one doing this. Eventually, how we look at it, that is that our safety is in our speed. There is so much going on that we are constantly increasing the delta between us and the rest of the market. If you would now tell me, by the way, this supplier now has full-stack in this region, then my thought line would be, okay, then we start doing that in 2013 or no, in 2011, I think.
We are exactly eight years ahead of you. It's over to the Adyen team to make sure that that eight years feel like 16 years, and that is very difficult to catch up. It's not a single trick that we do. It's the combination of what I said in my summary, keeping things simple, which is very difficult. Every day we get a demand for veering away from that path. There's every day an opportunity to make it more complex. On the other hand, because of that, being able to attract the most talented team to execute it. I think that that's where the solution is. Also, if you look at where we take from, because it's quite spread out between companies which run multiple, which run many, many platforms. Some SME providers, banks which provide point-of-sale terminals.
They cannot all respond in such a way that they are better in holding onto their merchants. This is not a market in where I say if another company would start looking like Adyen, then still the total addressable market would be EUR 21 trillion and then still we only captured EUR 159 billion of that.
Thank you. Sanjay Sakhrani from KBW. I had two questions. One was, there's obviously been some consolidation in the U.S. acquiring market. I was just wondering how you think it affects Adyen. I have two for Ingo, actually. One is just, do we need to consider anything as eBay volumes roll on in terms of take rates or operating margins, especially as we move into 2021? Secondly, just excess cash usage, sort of how we're thinking about utilizing that cash, whether it's capital management or M&A.
How we look at the consolidation that's going on in the market is, it's not our strategy. We never did an acquisition. We are not looking at companies. The reason why is that we feel that it would make it very complex to make the payments market a functionality play. We make it a functionality play. That's how we win. By consolidating with many platforms, the dream is always that you then move all that volume on your new fantastic build. To do that in practice turns out to be so difficult that you don't see examples of companies actually executing that. I've worked for a few years for Worldpay. I've lived it.
I know how difficult the management decision is to say, "Okay, we're going to switch off this platform." Either as a merchant, we give you a year's notice, you are on our new platform, or we just tell you today you have to be migrated away. In practice, everybody goes for RFP if you have to reintegrate. You're going to see enormous income cuts and lost customers. It's from a management point of view, enormously unattractive. Especially if you run the company with a little bit of a shorter horizon. We run things on a long horizon. We feel it's much better to spend our efforts to boarding merchants on our platform, make sure they're really happy and that they hang around with us for a long time. Then you about eBay.
About eBay and margin, I think if you look at eBay where we are, we're really happy so far with the progress that we have made of eBay integrating into our platform. If you look at their earnings call, they also announced that they were quite happy with the performance so far. Until the summer of 2020, they're stuck with their PayPal contract. We'll see more volume over time. I think individual margin, we never give any indication on individual customer contracts. Of course, if it's a big merchant with high volumes, you might expect a decline in take rate of that single customer. I think more importantly is that it is a profitable contract. We don't have loss-making contracts on our platform. That's never been the strategy. We always make sure that contracts are a net addition to the profits of our company.
On excess cash, I think it is really helping us that we have this strong balance sheet. I think the combination of the eBay deal and the IPO last year got us into a position that we're now negotiating even bigger deals. We're at a table of really big companies talking about payment strategy. Having this very strong balance sheet just makes the conversation way easier. For the foreseeable future, I don't see any change in our dividend policy, which is basically everything adding to earnings or to equity. Of course, if that continues in the future, we will revisit that. Now it really helps us to execute our strategy, and that's why we continue doing this.
Hi, it's Jamie from Susquehanna. Thanks for doing this capital markets day. I know you like the long-term. I was just wondering, though, over time, do you think your prices go up or down? At least if you could help us with a structure about how to think about that. Thank you.
Okay. I think there are a couple of effects on our platform. If you just take take rate as a starting point, because that's in the end how most of you like to look at us. There are a couple of effects on take rate. Getting more full-stack volume on our platform is a positive because we make both margin on the processing fee and settlement fee. Getting enterprise merchants is a negative, and the same for point-of-sale. That's for the same reason, because they typically are very large entities with tiered pricing structures. The more volume they bring, the lower the price per transaction is. That also has a negative take rate effect. The fourth effect is the mid-market, mid-market comes with higher pricing per transaction because they are in lower volume tiers. That has a positive effect.
The question is, how is that whole mix panning out? It really depends on how fast we can grow certain areas. That's also why we don't look too much to the take rate, because it doesn't really say anything. If we are competitive, absolutely. If the feedback we get from customers is if we go for RFP, that we're typically more expensive than the competition. We can do this because we bring additional functionality. We make it a functionality game instead of a cost game. If it's purely about cost, we would walk away. So this is not a race to the bottom for us. It's about adding functionality and getting paid for that.
Good afternoon. Michael Del Grosso with Jefferies. Two rather high-level questions. The first is, what percentage of your customer base are you currently processing on a global scale across your unified payments platform? Just a high level percentage of what wallet you currently have of your merchant base. Then, I guess the second question is more on the technological capability. What's the typical authorization rate improvement that your technology has versus maybe one of your competitors, or when you onboard a merchant, what's that improvement typically?
I'm afraid it's really difficult to give generic answers to both questions. I think in general, if you just look at share of wallet, it really depends on how far we are with the customer in the implementation. If it's early stage, it's probably a few percent. If we're working with them for a couple of years, it's trending towards above 50%, I would say, specifically for point-of-sale. I think point-of-sale in a region, if you take Nike as example, we work for Nike in both U.S. and Europe, so we have their full volume. That, I think, gives sort of an indication what our strategy is in this area. The same for authorization rates. Of course, there is a Forrester report claiming that we do about 1.4% better than the competition, but it also depends on the implementation.
If you sell to Brazil out of your U.S. acquiring license or your U.S. acquire connection, you probably get to authorization rates of around 30% or 40%. If you switch to local acquiring with Adyen, you probably get to 80% or 85%. That's an enormous improvement. Much depends on the setup. I think the two metrics that really matter to me, the first one is our volume churn. Our volume churn is less than 1% over last year, and that's a very consistent number over time. The second thing is that if you look at customer feedback, if you look at our account managers working with our customers, of course, our customers measure very precisely authorization rates. As more than 80% of our volume growth comes from existing customers, it gives an indication that they are happy with our service and give us more over time.
Hi. Anneka Treon from Kempen. I think following up from the previous question, what are your thoughts from your merchants' perspective with regards to processor diversification? You sort of laid out your competitive landscape earlier on in one of your slides, and I think we would all agree that especially as you fast-forward things, especially from a sort of unified commerce perspective, you could argue that the banks that are playing a part today would be increasingly less relevant. Also the SME providers, especially for some of the larger, more omni-channel merchants, also less relevant. If we think about the world in the next 5- 10 years, and perhaps it starts looking a little bit more oligopolistic from more single-platform players, what do you think is a sensible pie split for your merchant base?
You mentioned that, for example, Nike U.S., you guys are pretty much exclusive. Do you think that's a reasonable working assumption? Do you think merchants feel uncomfortable with an overreliance? How do you look at that?
I think it's really a matter of trust, building a trust relationship with your merchants. I think typically if you provide sufficient added value, we see that merchants give us more and more volume, and you have trust in the relationship. I think there are a lot of companies that fully trust on one ERP system. Why couldn't you trust on one payment system? It's more working together with your merchant, like how you prove that you are sufficiently resilient and that you're not there to, if they're dependent upon you, that you try to fool them or try to make profits out of that. I think that's, of course, traditionally a problem in this industry, that there was a lot of intransparency about pricing. That's exactly what we're trying to bring. We're trying to bring transparency to our customers and build that trust relationship over time.
I certainly think that share of wallet could grow to 100%. That's not a problem at all. It's just not the standard traditionally in this industry. Philosophically, I see no problem to get there. Do you agree or?
Yeah. It creates additional cost for the merchant to not do it. You see it changing in the more traditional company you talk to. They sort of use arguments like "That's the way to keep you honest is to have you competing with others." Whereas the more advanced company that you speak to, they are more rolling out per region, and you never hear that argument. I agree with Ingo. It's partly a mind shift because that's how you used to run it, because traditionally in payments, you just charge whatever you could get away with. High forex rates, weird additional fees, reporting fees, all sorts of fees that you could throw to the merchant. On the other hand, we need to do perfect execution to build that trust. Brian, anything to add or no?
It's Adam Wood from Morgan Stanley. I've got a few as well, actually. Maybe just as a follow-up to that U.S. consolidation point. You're alluding, Pieter, to the fact it's very likely they end up with a lot of technology complexity and problems integrating platforms. Is that a workstream from a sales point of view that you can go after specifically and try to target some of the bigger customers that are working with those companies that are going to go through quite complex technology changes over the next few years? Secondly, just on PSD2.
When you speak to the big merchants, I'm thinking particularly here, the very kind of tech-driven big merchants, is there any risk that they try to do their own payment system and create their own payments model that would actually have, apart from the tech, a zero cost because they can use PSD2 to do bank accounts, they would view it that we're a tech company, we can actually enable that and make a structural change to our payments cost. Is that a discussion you have with any of the big merchants or see as a risk? Then finally, maybe just trying to understand the shift to full-stack and the impacts on take rates. We saw that 60%-70% shift in volumes going to full-stack.
Is that just the areas that you don't want to acquire are not growing, so the airlines and travel industry? Are there other factors in there that would change that mix, like getting new licenses in new markets? If you could just try and help us understand the gives and takes on that'd be really helpful. Thank you.
Very good. All right. I think in general, changes in the industry usually work to our benefit. Maybe it's nice, Brian, if you say something about PSD2, how that works to our benefit.
I think there's two sides of PSD2. There's one which is open banking, which you allude to. The other is the requirement around Strong Customer Authentication. You can see out in the market, we're covering ourselves in both of those areas. I spoke briefly to our 3D Secure 2 implementation, which is exactly an implementation our merchants are using to prepare for the Strong Customer Authentication element of that. We're already in that space. However, that wasn't really your question. Your question was more around the open banking APIs and sort of where we see that position. First and foremost, we're already live in the U.K. with open banking API aggregation, and we offer that via our platform, right? I think you need to really distinguish between what a company is capable of doing and what they have the business incentive to do.
The farther something gets away from a company's core business, the less likely it is that they'll invest in technology to do that. We see ourselves as a payment technology company who can fill that space. I think as you also probably know, no space immediately 100% moves to anything. It becomes part of a greater mix, right? Like any payment method mix in any given region, we see open banking as one more thing in the mix. We think that this actually plays to our strength, that the more competition and more complexity there is in a market, the more you need technology players like us to simplify that complexity, and therefore, the more attractive our offering is.
The question on full-stack from 60%-70%. That's mostly the fact, of course, of rolling out globally, so adding the additional acquiring licenses. Of course, we think that we further can build this. The 30% is mostly acquiring for, or the volume for airline that we do not acquire or just a gateway because we don't want to acquire it. Of course, over time, if the merchant base grows, the airline share will decline because if you look at total, the EUR 21 trillion market, airline is not 30% of that. That will decline, and of course it will be, I think the full-stack growth will be accelerated by the fact that we further roll out internationally.
Two questions. I think we all understood the benefits for merchants, can you talk us through a little bit the process that a large enterprise would go through to switch? The timeline, the cost, and how they think about it, right? Again, when you're convincing them, the benefits may be there, what is the process and how intensive is it for them? Second question, somewhat related, although different, again, these speculations that Uber is applying for their own banking licenses and they'll start processing their own payments. How do you answer to that?
The sales process. Unfortunately, it is true that if a merchant has mentally already decided to choose for Adyen, you tend to be part of a rollout of something. Often, something else also changes. They might change to a new cash register. They might change to a new system for their online sales, and often you're part of that. That is a timing when you get in, and often even that combines with a new region. When we go live with that platform in Australia, we'll do it with Adyen. That can mean that sometimes we speak to merchants for a long time before they go live.
There are the cases where they commit full countries or regions where they say, "We're going to do this with U.S. and Europe with you," and you're looking at full rollouts where you get a lot of volume over a few months. We didn't use the word lumpy yet today, payments is lumpy. That means that there are large deals that you're working on which are very difficult to time, which could fall in a certain year, could also fall a year later. We have seen cases where it took 10 years from where the merchant was already saying, "Yes, we're going to do that," to actually doing it.
The cycle can be very slow and it often depends on other processes. The pure integrating effort is not so much work. You could do that in a week or two weeks, even if you have a very deep integration. The disruption to the business is material, because think about simple processes like how do you do a refund if you change provider because that new provider doesn't have the original transaction in its system. It does disrupt your business significantly to move from supplier to other suppliers. That is good once we have them on our platform, and that can be a reason why it's difficult to board them, because it's not a decision you take lightly. You had another question about Uber. Uber is doing that for their payouts.
There are all sorts of rules in Europe that as a marketplace, you are not allowed to touch the funds unless you adhere to rules, and they solve that with their own banking license. It's not intended to do acquiring, to do their own card acquiring, which we do for them. That does have impact.
Good afternoon. It's Nooshin from Deutsche Bank. I have maybe two questions. First is on your cross-border transactions. If I understand correctly, you can make this to not appear as an international transaction, but as a local transaction. I'm wondering how you can comply with basically all the network rules and so on. If you can just give a little bit on that for my understanding. Then again, on the same on PSD2 and the licensing of your big customers. If I understand correctly, most of this complexity of processing the payment comes from processing the data and basically to put aside the fraudulent transactions and so on. Some of these big customers, they have no problem to process those data and this would not be a complexity for them.
What would you think about them more and more going for own license and processing their own payment or becoming their merchants of record and so on and so forth. How do you see that as a risk to your business and where the whole market is going, basically, here?
Let me answer the second half of your question, I'll ask Edgar to do the first part of your question. Second part regarding where is the cutoff for merchants to do things themselves with scale. Originally, we thought that it's something that the largest merchants might do themselves. If you see how the payment market develops, it becomes less and less attractive to do that because the resources a merchant needs to commit, the size that you need to have as a merchant to do that, then you could say, yes, Amazon, U.S. domestic, probably yes. Even for a company like that to do it in other regions, probably not.
If you are a size below that, it doesn't make much sense if you see how much investments you need to do in payments to have that run properly, how much you pay for it to outsource it. What actually the benefit to the company are to do it themselves. We don't see that as a threat. We don't see that our large merchants. It's unattractive also to split out a payment method yourself because suddenly your data is in multiple places and the overhead assigned to that is huge. We have seen examples of the other way, though, that merchants were doing it themselves, and were running that over time and then finally say goodbye to that. It is actually a strategy which I think is now more and more proven to be not valid.
Whereas there are large merchants moving now to us and sort of saying, "This is ridiculous. We have 150 people working in our payments department. This does not make sense." I found out that each time that I mention a merchant name, I regret it later. Also this time, I'll stick to that rule. Wanting to be helpful, but it somehow never works out. Edgar.
To your other point regarding local acquiring, you're completely right. On the card scheme rules, we can only offer local acquiring when a merchant has a local entity. We just make that process very simple and offer them the same solution in the markets where they have entities. We help them find the optimal payment processing setup within the lines of compliance and card scheme rules, of course. It's not localizing where a merchant is not having a local entity. I hope that answers your question.
Do you have to have a local entity or an issuer?
To qualify for the full benefit of local acquiring, to register them as a merchant in, let's say, France or the U.K., they need to have an entity in that market. Yes.
Hannes Leitner from UBS. I got also a couple of questions. You mentioned that in 2018, most of your growth came from existing merchants. Can you give kind of a guidance for 2019? Maybe also in terms of new merchants, can you break down a little bit the pipeline so we get a feeling about the ramp-up, how far you are, and how it compares to last year? Then maybe something on the industry How often do you actually see merchants churning their merchant acquirers on the POS terminal and then in the e-commerce? Then in replacing your existing, or when you win a new contract, those replacements, can you say if you usually replace banks or other payment companies? What do you see out there in terms of dynamic?
I think in general, if you look at evaluating how our sales evolves, we look at sales pipeline, we measure that internally, and we see how that evolves also compared to growth of the sales team. That's a very important internal KPI that we will not publish outside of the company. We, of course, find that very important to make sure that number is right, because otherwise you might run into problems in later years. I think the fact that it is a matter of building a trust relationship with a merchant explains why most of the growth comes from existing merchants because you work with them. For the new ones, you need to build that trust relationship. I think that's on the first part of your question.
I think replacing point-of-sale is not that easy because to go into store operations is not what retailers typically like to do. There has to be a good reason. Well, standardization is a very good reason. That's also why we see that certain roll-outs can take place very quickly because they get standardized procedures throughout Europe, U.S., or other regions where we are active. To replace it again is, yeah, why would you if it works really great? That's, I think, what we, yeah, need to prove or what we continue to prove to our merchants that we are very good in delivering our performance and that there is no reason to churn. Yeah. And so far, in point-of-sale, yeah, churn is not an issue. I think the overall churn in volume is less than 1%. That includes, of course, point-of-sale. Yeah.
Hi, this is Wei from Whale Rock Capital. Two questions. I guess the first one is, in terms of all these other value-add products, when they're inside reports from all the data you collect, are you still planning to keep that as part of the bundle for your current processing take rate, or are you planning to monetize that in the medium term or long-term future? The second question is, with PSD2, the two-factor authentication, if the two factors getting rolled out, would that theoretically decrease conversion, or do you have some ways to kind of offset that to improve shopper experience and increase conversion?
On your first question, I think for us, it is most important to grow quickly with our current merchants. I think the reason why we build trust is because we provide a lot of value-added services. Of course, we will continue to ask ourselves whether we can find something, but we want to stay away from intransparency. We want to keep transparent pricing, and if a product proves to be a real value add, of course, we will start pricing for it. We have done that in the past with a product like RevenueProtect, but also now with 3D Secure 2.0. These are things that we price to merchants. If we see good opportunities in the future for other products, we will. At the same time, we of course, want to make sure that we, yeah, keep our customers happy on our platform.
To your question of the, let's say, the authorization rate risks around Strong Customer Authentication, absolutely, it is a risk for the merchants. Depending on the merchant's business model, anywhere from 20%-80% of their transactions, perhaps even higher, will require Strong Customer Authentication in Europe. We actually see that, however, as an advantage for ourselves. Most payment processors are using a white labeled 3D Secure authentication solution. You now have a scenario in which 20%-80% of your customers, your merchants' transactions, are now leveraging someone else's technology, and it'll be a crucial part of the authorization flow.
Actually, our ability to very quickly pivot into our own built and certified authentication solution well ahead of that September timeline, has turned into a real asset for us in the market and has turned a risk on the merchant side into an opportunity for us and less of a risk for merchants who work with us.
Hello. [inaudible] Thanks for taking my question. My question is for Brian, I think, with regards to ShopperDNA. One of your competitors, PayPal Braintree, they very much emphasize their, what they call, two-sided network, where they also have 250 million consumers with a PayPal account. They say that helps them very much because that gives them a lot of insight in those consumers. It helps them with authorization, marketing. I think similar things to what you were talking about in terms of ShopperDNA. My question is, do you think it's a pity you don't have your own digital wallet, or do you actually think it's maybe an advantage? It gives you additional freedom not having your own consumer base? Maybe a related question, how do you think about privacy in terms of ShopperDNA?
Great questions all around. I'll not go into speculating on the advantages of moving into consumer-facing. All I will say is we work with merchants, and therefore, we do have visibility on data of their consumers, and we feel that that's the place in which we're most comfortable. Our customers are our merchants. To the more macro level question of what then does that make in terms of the difference between capabilities between us, PayPal Braintree, it's different data. At the end of the day, one can have access to more data, but then they might not necessarily do things with that. For example, you see that they've had to acquire companies in the risk space in order to have a competitive product, of course. I'll sort of leave the speculation there. There was a final part of your question I'm forgetting.
Oh, yes, indeed. I think you should make a significant distinction between the sort of flows that we talk about in ShopperDNA and the other data-driven flows that we provide in unified commerce. ShopperDNA is part of our risk system, and therefore it generally falls under other aspects of GDPR and data protection. However, we never expose multiple merchant data to a single merchant. We have a strong delineation between showing a merchant to their data and their transactions. Then we do, let's say, PSP-wide aggregations, but those are always highly anonymized, tokenized, and then we never show specific information to the merchant. It goes into the modeling and helps them with their assessment, but we never give them any information about a customer that's occurred on another merchant's transactions.
I do think that it would be a tricky choice for Adyen to start a wallet and starting to work with consumers because our merchants choose for us because we chose their interest, and we don't seek a relationship with the consumer. Whilst ruling nothing out, it would be a major choice if you say that from today, we're going to interact with consumers. We have one customer, that's our merchant, and we try to help them to sell, and we don't try to build a relationship with their shopper. A large part of our merchant base wouldn't appreciate it if we start reaching out to their shoppers to sell one of our products.
Hi, thank you for taking my question. Tammy Qiu from Berenberg. First part of the question is about a lot of your growth is coming from relying on high growth enterprise and large enterprise. You obviously have captured all the right ones for the time being. How do you make sure that you have all the right ones for the future or forever? That's the first one. The second one is, I noticed that you guys actually do gambling business for acquiring a gateway service. Can you disclose what is the % of your total volume coming from that side of the market, and how you view as a mix for the total volume going forward? Thank you.
If you look at the merchants which we have versus the merchants that we're working on, you constantly see. What typically happens is that in geography and in industry, if you have a merchant, that the merchant which is close to that, be it in products or be it in geography, is boarded onto the platform. There is, in the beginning of building Adyen, there are certain biases. Over time, you see that moving away. If you now look in the pipeline merch that we have, and we spoke about quick service restaurants, you see that we are developing whole new businesses, and that we make sure that our product really works for them, for the future. That we don't end up being very highly biased towards certain industries.
On the other hand, the ones which we boarded are, of course, the merchants which are most open for change. I think it's just a matter of time, I don't have the feeling that we are only interesting for those merchants. It's more a factor of how quickly can you board and how important is it for you to make those changes, or do you feel you can wait a little bit longer? There was a second part to your question.
About gambling business.
Do you want to?
We indeed have licensed gambling on our platform. It's low single- digits total volume percentage. We have very strict measures in place to make sure that it's fully compliant. For instance, that U.S. shopper can't use sites, et cetera. We feel that we're not exposed there.
Yes, hello. Good afternoon. [inaudible] Econopolis. I was wondering, you being a technology company, there are some other secular trends alongside payments, like Internet of Things. I was wondering whether some of your merchants would want to go beyond just the terminals you have there into connected devices, smart devices, in order to allow payments, and what the consequences might be of that, because I guess mostly it will be micropayments in that case.
I think overall you would see that as a positive trend for us. I would speak back to the themes that we talked about of being channel agnostic and having sort of a universalized tokenization system, right? If you look at initiating a payment from an Internet of Things device or a voice-driven device or something like that, to us, that's just implementing what we would call an online transaction. It's the same API interactions, it's the same tokenization system. From that, we're better positioned than I think most in our industry to deal with integrations into those new types of channels. I even spoke to things like, could you have a token attached to your vehicle, things like that. I do think that there are some trends that you could consider around that moving things towards micro-transactions.
That being said, that's a dynamic that already exists in the, let's say, the App Store ecosystem, and you tend to find that large players tend to bundle together transactions, aggregate them, and send them in bulk. There's a lot of different things around that. I don't think that would have a substantial change to our fundamentals, however, in how we treat that.
Thank you. [inaudible] . A bit of a more general question. You mentioned that it is possible, philosophically at least, taking the 100% of wallet of a merchant. In the long-term, how do you see this current market situation where it is filled with a lot of providers of payment services and acquiring? Do you see that there is a scenario of a winner takes all in the long-term, or is it going to remain with the many players in the market? Secondly, you used the research of, I think, Nielsen when evaluating your total market, and at least total addressable market and the growth of that market going forward. I think it is based mostly on card payments, and when we think about it, should we also think about other payment types?
Would this add up to the total addressable market or would you still remain at the same number of around EUR 21 trillion currently? Thank you.
Yeah. I think the question on market size is a question that we don't spend too much time on it, because the market size is just so big that we see the more important limitation is how fast we can grow the company in a responsible way. Market size is not limiting us. To your second question, a winner takes all scenario. I think this market is so big, that's highly unlikely. I think there will be a difference, and that's a key thing about, do you see payments as a functionality game or do you see payments as a commodity? I think there is plenty of space for people that see payments as a commodity, and they will fight each other to death about pricing. If it's a functionality game, you can actually bring new revenues.
I think there are only a few players that can actually do that. I think that's why we win a lot of the deals with the more advanced players because they see the payments as a functionality game. They want to pay a premium for it. I think in that particular part of the market, we're super well positioned.
Yeah. Hi, this is Mark Udis, Camden Capital Management. I was wondering about the sales process in the mid-market SMB business. You mentioned that you both use system integrators but also inbound. Can you tell a bit about those two parts and also about the sales organization around those two sales methods?
Yeah. In the mid-market, it's important to be integrated with the platforms that mid-market customers use and to build a good relationship with those platforms. If you look at how we do account management at the corporate side, that means that it's a one-on-one relationship. Although as an account manager you could have multiple companies. As a company, you know who is your dedicated account manager. In the mid-market, it's more a team. You do have a level above just customer support. There is somebody who can look at your business and be helpful. That's a pool of people. There is a lot to gain for us in working together with the platforms because for them, having multiple payment partners is also for them suboptimal. There's a wish for them to reduce the set of companies they work with. That's the objective for us if we work with those platforms.
What platforms are you currently working with?
Oh, we are integrated. Yeah, I'm never so good at those lists.
Magento, Salesforce Commerce Cloud. You can imagine the smattering of different e-commerce platforms. I think for the mid-market there tends to be more of a regional variation. I think part of the build-out that we have this year on the product side would be a smattering of the regional specific ones.
[inaudible] ABN AMRO. Few questions. I know that you don't drive the on take rate, but was just wondering, for clarity's sake. Your take rate on point-of-sale versus online, is there differences within a single merchant of that? Secondly, are there maybe geographical differences in your pricing schedules due to competitive pressures being larger or less in certain geographies? Well, if you could give some insight in your wallet penetration, that would be great. I think that question has been asked already multiple times.
I think indeed the question about online versus offline, it really depends on, I think, how a retailer works. We prefer to work on one pricing deal because it makes it also easier from our perspective. We have a good discussion about what a merchant wants to accomplish, it shouldn't really matter how volume is brought in, whether that's online or offline. Of course, sometimes, if a retailer is set up in a more traditional way with separate teams for point-of-sale and online, you could get two different pricing models for both channels. It has not our preference. Of course, there are regional differences also because if you look at certain markets, they're just way bigger than other markets. U.S. market is a big market where there's a lot of volume for some merchants.
If you look then at what kind of tiers we end up, it is a different tier. With better pricing than, for instance, the average in Europe, which is a way smaller market. That's the type of difference that you see geographically. I think in general, we see with merchants that they like for straightforward pricing proposals. In the end, if they start to realize that they really have one platform, you also come to a one pricing discussion.
I do understand the differences in volume size in the U.S. and that you will be higher up in your tiering schedule than, for instance, in Europe. Would that also then mean that the height of your tiering schedule will be different so that your, let's say your top tier would be 20% lower in the U.S. than in Europe, for instance?
I would say not necessarily. I think it really depends on the type of deal and also the type of added value that we can bring. I think that's a key thing where all sales processes start with, like what kind of value can we bring to a merchant and how do we price it correctly? That's mostly volume related and indeed what kind of additional value we can bring as a company.
Thank you.
Okay.
Anyone else with questions?
No. Okay. Well, thank you for being here today. It was a great pleasure to have you here in Amsterdam.
Yeah. Thank you.