Good morning, and a warm welcome to MGI's Third Capital Market Day here in Stockholm. My name is Jenny, and from Europa, and I'm supporting MGI here on the ground, working close together with the head of IR, Sören, you are somewhere around here. Here you are. And also Danesh, Senior IR Manager here in Stockholm, sitting over here. Probably already met him, right? Okay, so I'm gonna moderate today's session, and, as you all know, MGI released the second quarter financial report this morning, and I know all of you are super eager to dig into the financial details. So that's why we put it up really early on the agenda for today. So, what has happened since we met last year? A lot, as always, with MGI.
I mean, management have been super busy with the integration of the software platforms, and of course, also managing the situation we have in the world around us. They also moved their head office to Stockholm. MGI is today a Swedish-registered company, and even though you have 70% of your revenues coming from the U.S., the largest market. That's also why we invited experts to really put MGI's position and way forward into a context in two of the largest markets. We will have experts from the U.S. on the agenda, together with experts from Germany, and I will introduce them to you now. Let's see what we have here on the agenda. Disclaimer. Here we go. Of course, Remco, the CEO and board member, Sameer, Chief Revenue Officer, sitting over here. Paul, CFO, over here.
Rami, General Manager, AI, sitting over here. Jens, COO, also over here. Prasanna, CTO, also over here. Experts, I know you will be super excited about this, because I had a chance to chat with them earlier. Today, we will have Jay, founder and CEO of Digital Capital Advisors. We will have Jeff, Chief Strategy Officer of Basis Technologies, and we will have Oliver, General Manager of Pilot Group from Germany. I will moderate. Danesh will moderate the panel discussion coming up later. The purpose of today is really to give you on business update, financial update, dig into the details when it comes to the strategy and the way forward. Of course, when it comes to technology, innovation, and also to put it into context with the external experts.
But maybe most important is really you have a chance to meet with management. We have the management team, everyone here, so really, it's a great opportunity to dig into the details in the different areas. So, coming to the agenda for today, we're gonna start with the business update and short coffee break, and then we continue with the financial update. After the financial update, we will have a Q&A session, where we also will open up the phone lines. So you will be able, here in the room, of course, to ask questions, but also on the phone, as well as you can put in via email the questions, and we bring them on to the team up on the stage as well.
And after that, lunch, and then we have the external expert sessions and coffee break, and then we will dig into a customer case, really, you know, showing and illustrate what MGI do for, for their customers. And after that, we go for the panel discussion and then the outlook and closing remarks. So with that, and further ado, I would like to hand over to Remco, CEO and Board Member in MGI. Welcome, Remco.
Thank you. Yeah, good morning, everybody. Happy to have many investors here, analysts, other stakeholders, and even more online. Yeah, I would like to guide you through the first part of our Capital Markets Day 2023. Aim is today to look a little bit back to see where we stand today, and also, of course, where we want to move the company. I would like to start, because that's what people want to see, of course, or probably already have seen online, with our Q2 performance. Some highlights. We are really showing, I think, a very good performance in Q2. And this is... If you look at the numbers, you don't see it immediately, but we are working at the moment in an environment that's very negatively affected by the macroeconomy.
Interest rates, interest rates went down, we got the Ukraine war, and a lot of pressure, uncertain consumers, inflation, all those kind of things. The reaction of companies in those situations is to save money, and the first thing where you can save is your marketing budget, your advertising budget, and that's what we have seen. And with that, also, you get pressure on the CPMs, so the price per ad, price pressure, and that's an environment we're working in. We were able to gain market share, so we added a lot of new customers, and we were able to protect our bottom line. I think that's good in this situation. Just to realize, if you have a 20% CPMs going down, 20%-25% volumes being lower, and we were able to compensate that by more customers and doing, let's say, expanding our business.
We increased the volumes, and by that, we overcompensated that. I'll get more into detail what we focus on, what are our strengths, and where the market shares are. But overall, if I look at the numbers, we showed a +1% organic growth in the Q2. We had an overall -2%, but if we compensate for foreign exchange, the US dollar is 10%, let's say, in our negative, in that sense, so we had an overall 3%, growth. We are used to much more from the past. We did 18% organic growth last year. We did 38% organic growth the year before. So we are a bit spoiled, but in this market, I would really say this is a good achievement, and I'm really happy to gain market share.
If you look further into the year, we expect basically also to further be stable. In the beginning of the year, people were more positive about the second half of the year, and everybody was like, "First half year might be difficult, second half year would be a lot easier." The way we look at it at the moment, we see a little bit of light points, some larger advertisers increasing budgets, but still there is a lot of uncertainty. There's uncertainty about what happens with the interest; it still might go up a bit. So we decided to bring our guidance down and to basically say: We think we are doing a good job if we're stable versus last year. But that means stable in revenues.
That also means more volume, more customers, and when the headwinds are over and we're getting the tailwinds, we expect also with this larger volume to really show the growth that we are currently missing. So that's a bit on the outlook. Paul will go much more in detail in the numbers. Yeah, I think we should go a bit further. One thing, sorry, I'm almost forgetting, always read your own slides, I've learned. We did one thing also to protect our bottom line. We decided to do a EUR 10 million cost-saving program, which is implemented, which will be EUR 10 million savings on personnel costs over a full year basis. We are able to do that.
We really built this company via a lot of M&A, and we've always said we integrate what we buy, and integrating the teams goes much faster. We do that basically immediately, but integrating technology stacks takes longer. We were really happy that we disconnected two large stacks in the first half year, which brings us to a lower cost base on technology, but also enables us to be more efficient on the teams, of course. Plus, we have also put some more focus points on certain things. In-app is super strong, so more employees on the M&A part. Managed sales is weak at the moment, so there we reduced some FTE. So we also shifted some employees, but altogether, we were able to really reduce our personnel costs substantially.
So that's, let's say, a few highlights, and there will be much more about this later in the presentation, but I thought it's good to start with this. Yeah, who are we? And for those that don't know us so well, we are an integrated ad software platform. And especially in Sweden, you guys know us from the past as a gaming company. We say also some words about that. But we have built on top of the gaming, and we see ourselves now with the majority of our business also there as an integrated ad software platform. We serve over 550 software clients, and software clients are defined as customers with over $100,000 per year. So we have many more customers, but we look to see really the big customers that are driving our volume.
With those customers, we're running a lot of business. Behind that, there is a basis, a strong basis, a strong backbone with games. We have over 5,000 games. We have over 250 million daily users in the games, and we have over 2 billion connected devices, which you can, in the end, play ads on. Also, there we go in more detail. team, we have reduced the team. As just said, we have reduced personnel costs, so that also means reduction of, of personnel. So we have roughly a bit over 700 employees now. We are in 25 locations, which the main market that we have is U.S., 70% of our revenues, but we have sales offices mostly in other locations, but also some offices where we do technology.
So we're a really global company, and it doesn't make sense to have all your technology people, for example, your engineers in the U.S., that's super expensive. So we have several in Germany, but we also have many in India, and that, of course, gives a cost profile that's much more beneficial for a company like we are. Then I would like to get Jens on stage, our COO. We're going together now for more than 10 years. When I bought Gamigo, a bit over 10 years ago, Jens was in there, and in our first discussion, he told me, "I'm gonna leave this company. This company is not great." He's still here, so I would like to hand over some words and-
Thank you very much, Remco.
Yeah.
Yeah. Thank you. Yeah, it's a pleasure to stand here in front of all of you and present today. If someone would have told me 11 years ago, where really with one step already has been out of the company, that I would stand here and present a listed company in Stockholm at the Nasdaq, I would not have believed it. I would have thought whoever told me that is a little crazy. But yeah, so it goes. As you can see, we have come a very long way in those 10+ years. It's actually 11. Was exciting times. It was not always easy. Sometimes it was also hard, came with hard decisions, but exciting times. Learned a lot.
Here you can see a nice picture from Gamigo event in 2020, 2012, actually. You see less gray hairs than now. Not everything is company responsible, but some of them may be responsible. I remember the first months at Gamigo, actually, how that was. The mood was bad, the structure of the company was totally, the focus was set wrong. People, like, we have seen that very often in the gaming industry, almost everyone in the company worked on the next big thing. So the company was highly unprofitable. Today, we are doing bad, but the next big thing will be the hit, so tomorrow we are doing better. We all know that usually it doesn't happen very often in the industry.
And while the very profitable games were kind of left alone, not a lot of maintenance anymore, not a lot of content, not a lot of events, community management. So what we did, very early days, we exactly changed those structure and changed those processes and said: "Okay, we are really concentrating on our profitable games and make sure that they give us a profitable revenue stream for a very long time," because we all know that free-to-play MMORPGs have a very long lifetime and very loyal customer base.
Then we looked in the market, and we figured out that we are not the only company with those problems or also structural problems or mindset that the next big thing will solve all the problems, and that companies round about 10 million annual revenues were not able to survive anymore by themselves. Competition in the free-to-play market increased, and so on and so on. So we raised some money, and that's where we started to integrate our buy-and-build strategy, and we acquired those companies, actually quite a lot. So far, until today, more than 40 companies, not all in the gaming section, to be honest. We started-
25 in the gaming.
25 in gaming, the rest then later in media. And that's how we really started to grow the company on the buy-and-build strategy. Then we did that really successful until 2018. In 2018, we also started, expanding the team, investing in the team, professionalize ourselves, and we recognized in gaming a problem in the advertising market. It was, for us, not so easy anymore to get users. A lot of offline, traditional offline brands also recognized that online marketing, digital, advertisement, is getting bigger and is basically replacing the offline, marketing, and therefore, brands like Mercedes, BMW, and so on and so on, went into the market. CPAs got higher, there was a fight for, traffic. And we thought, "Okay, what we did with the gaming, we could maybe also do with the media," and we started to acquire media companies.
The initial thought was we start to acquire media companies to leverage our gaming assets with user acquisition, but then we also found out very, very fast that it's the other way around and that we can build basically a flywheel that we can leverage our media companies much, much better with our first-party data from our games, which we now call owned and operated content. Then we did the Nasdaq listing in 2020. Now, Jenny Mundus mentioned it already, we have the headquarters now in Stockholm, and our mission: expand and leverage our USPs to create products and services that make media better. We always had and still have a focus on highly scalable and profitable business model.
Our revenue CAGR, 47%, if you look at that from 2014 to 2022. While you also can see here clearly on the slides, from 2014 to round about 2018, we have growing the company, by the buy-and-build strategy, by M&A. And from 2018 on, you also can see that we added a significant amount of, organic growth, to that. Also, 62% EBITDA CAGR, I think it's very outstanding and impressing numbers.
If you look back to see really that we 30x'd the revenue, starting with roughly $10 million with Gamigo and now over $300 million, it's really nice to look back, but it's even better to look forward.
It is, yeah. It is. One more slide to look back, then we start looking forward again. Yeah, some important milestones. Many part, come and came together as one, as you can see in the beginning. Then, we did the Aeria acquisition, which was, for Gamigo, a really big acquisition. Aeria Games was bigger than we were. It was a huge thing. They were our biggest competitor in Germany. Doubled in size. Then Paul joined the company, also, a huge step, maybe bigger than the Aeria acquisition. Then, then we did the Twine acquisition, was almost as big as Aeria, also doubled us in size again. And then you can see, we started to acquire, with Verve Group, PubNative, we started to acquire media companies.
Sameer joined the company, and then later, you see also Rami and Prasanna, who will present today at the end of the day, some business updates. But then we invested in media. We transformed the company into an ad software platform and with our own content, having vertical integration of everything. And it has been a long way. Again, as I said in the beginning, it was almost always exciting, not always easy, sometimes hard, but we are still not finished, and we have still a long way to go. And I'm still excited to not only to stand here, also to work for the company and hopefully for the next 11 years.
Yeah.
That's my part.
Continue.
Thank you very much.
Thank you. Yes. Thank you very much. Yeah, then let's talk a bit about advertising, opportunities and challenges in the advertising market. For those that don't know us that well, I would like to start with the first slide, which is just showing what we're doing. It's basically super, super simple. If you have a phone, an app, and you open a page, there are advertising slots all over the page, some things, banners, videos, et cetera. At the moment you open the page, a process starts behind it, and it's an auction process. Like on a stock market, we auction the different ad inventories that we have. Sorry, I have to move more to the middle of stage, otherwise I'm not in the camera eye, honestly.
So you open the page, the ad gets auctioned, advertisers are bidding on it, and the highest bidder wins. Few things: you don't want to see an empty ad, so it needs to go super fast. Under 100 milliseconds, we have to do this whole process, because otherwise you see a page with empty advertising spots, or even worse, advertising spots that are still loading. That's not good. So it needs to be fast. It's all electronic, it's all machines. And then the bidding. Who bids most? Why do they bid more? And then we come to the crucial point, and that's data.
The more we know about the person that opens the page, the better we can give information to the advertisers, and the better the advertiser can decide if this is a potential customer, if this is somebody who would like to serve an ad to. So that's what, what happens here, and data is crucial, and we'll talk a lot about data today, 'cause that's one of our key focus areas, something where we're really strong. There's different kinds of data, but that will be covered later in the presentation. Why advertising? Advertising always has a bit of a negative inclination, but some of the biggest companies of this world are basically advertising companies. Google, making its money by advertising. Even in Amazon now, and Apple, huge advertising companies. Why advertising? To promote products, to enable free content, to build a brand, and to create demand.
So a lot of reasons to do advertising. A lot has changed in the advertising ecosystem. In the early days, there were only a few media, so it was easy. You took the phone, or you sent a fax, and you booked your media. We have so many channels now, apps, TV stations, digital TV stations everywhere, so things need to go programmatic, electronically, because you—a human cannot do that anymore, and also not optimize anymore. So that's the reason that the advertising market is really growing. And what you see here is really strong growth. It's a huge market, and it's further growing. And what you see also, I mean, the gray one is the overall advertising market, and you see that the digital part is further increasing, and gaming is a $200 billion market now.
Advertising in 2022 was $850 billion worldwide, and digital advertising is roughly 60% of it. But it's not as linear as it looks. Advertising is very much connected to the general state of the economy. And what you see here is really... And it's going quite a long time back, but already the Great Depression, you see that it was, the GDP was going down, but also advertising was going down. And it's fascinating to see how they almost run in parallel. How really advertising market is connected to the overall state of the economy. And that's a good thing. At the moment, we are in one of the down phases. I mean, before we had the dot-com bubble, which is a bit closer.
We had, of course, the financial crisis, and then we had COVID, and now we have an economic crisis, high interest, less spend, et cetera. But what goes down, goes also up. I think that counts for a stock market, but as much also for advertising. Talk a bit about where are the markets. If you look into the overall ecosystem, we do 70% of our revenues in the U.S., and that's not for nothing. It's a market that's much easier to serve, less boundaries, no different languages, all these kind of things. But it's also 60% of the total worldwide advertising spend that's happening there, so it makes sense to be there. If we add the U.K., Germany, the Nordics, let's say our main markets are 65% or even a bit more of the global advertising spend, so it makes sense to be there.
We also do something in LATAM, we do something in Asia, but those are our core markets where we're serving at the moment. And US traditionally also goes down fast, but also recovers fast, so we also expect to get some tailwinds from that. The market is dynamic, where we're in. There's a lot of effects and things that influence the overall market, the advertising. The mobile phone has come up. How many hours do you spend per day on a mobile phone? So that has changed. CTV is now coming. Linear TV was everywhere, but now it's moving to digital. And with digital, you can much better target people, so there's a lot of changes going on there.
And all those changes that are in the market are challenges, but also opportunities, and especially for a company like us that's growing fast. These kind of disruptions in the market, identifiers for example, also a very interesting one. With more identifiers, existing companies have to change their business models. And those are things where you also, as a newcomer in the market, really can gain market share and can grow. AI, big data, another point. Super interesting if you see what ChatGPT is doing, and that's more to the consumer side, but also in our business, AI is super important. With all those data moments we get, it's very important to target and to see where you put ads and how you optimize it. So that's a bit about the market. Plenty of opportunities, and let's dive a bit deeper into the company.
Oh, sorry, it's good if we go in the right direction. Sorry. Building a market leader, that's important for us. We want to make media better. That's our mission. And I would like to get on stage, Sameer, who is running our advertising business and also CEO of the company. And we both are gonna cover basically what are we doing here? Where are our strategic USPs? 'Cause the market is huge, you cannot do everything, but there are certain points where we really say, "Okay, this is where our biggest opportunities lie." The one is building a leading mobile in-app SSP. So really the exchange, where mobile inventory for mobile phones, for app, for apps and mobile phones is being auctioned. The second one is CTV. I just mentioned that linear TV is moving to CTV, connected TV, and that's very important.
It's a growth sector, but also super interesting to be there, of course. Data targeting, I think that speaks by itself, and then vertically integrated in a market without identifiers and where people are really seeing a lot of intransparency, it makes sense to really have one stack that connects the advertiser with the publisher. Those are the, let's say, four areas that Sameer and I are gonna cover. And Sameer, you start.
Thank you so much. Thank you to all of you for being here today. It's a privilege to be back in the beautiful Stockholm City. We'll take the next few minutes to actually explain what developments we have done, what sort of recognitions we've actually got with the hard work, and also what sort of investments we're actually continuing to do, and we have done in the past, for us to be a better platform and contribute towards the overall ecosystem. As Remco mentioned, we come from a very strong in-app background when we started our media business. We are one of the fastest growing in-app SSP, and we are proud to say that more than 20,000 leading applications today are utilizing our services, our platform.
82% of the group's media volume actually is actually coming in from the in-app monetization today. It's very critical to understand that a majority or a very high percentage of these applications are in-app where the consumers are consuming the contents. They are utilizing the state-of-the-art SDK, and when we talk about SDKs, very few company in the world have actually reached this scale or level in providing a very stable software. We work with more than 85+ third-party DSPs. We work with our own Verve Group-owned DSPs. There are some DSPs, demand side partners, who say to us, "Only send me the SDK traffic." So that's how important it is. It's a very complicated piece of software, but what it does is it provides a lot of stability.
It gives a lot of trust to both the supply side and the demand side partners who are actually a part of this complicated ecosystem. So we continue to improve our capabilities, we continue to improve our fair share, and bring the best towards the entire publisher base and the demand base, and we are also benefited with our own gaming content. Later during the day by a lot of other speakers, you'll be understanding about the nuances and the inputs that comes from all the all the partners that work in the chain with us. But this is, this is really this has really put us into a very leading position for in-app SSP.
This is a very important message, and we are all proud of, because when you put a lot of work, you know, acknowledgment and recognition is extremely important. We work with these partners, and this information was published when the moment actually came up. Pixalate is one of the IVT vendors that we work with, and they publish almost every month about the top exchanges, and we are one proud company that we were the number one exchange for both iOS and Android. This report was published for the month of June, where we stand both in North America, followed by EMEA. So this is a proud moment for the group, and we again continue to invest to make sure that we are always recognized in the top five, top three exchanges, globally.
We also work with a partner, which is Jounce Media. So Jounce Media is one marketing company, which is based out of North America. They work with advertisers, with holdcos, with DSPs, with supply side partners, and they have their own means of actually, working with, with samples of, participating with some very large DSPs and advertisers. And, these are public information which has been, which has been published as a benchmark report, which again places us in an extremely strong position, telling the world that, we are connected to direct premium apps. 79% percentile is the positioning, and we are way above, large competitors and partners like Magnite and PubMatic and Index Exchange.
So a lot of this has happened, and all of this actually happened in less than four years, so it is a real proud moment for us. This really gives a lot of trust to the partners to actually work with us, and we make sure that our marketing message actually goes out very, very clear to this ecosystem. I'll cover some of the most important things, which is investment, right? We have to continue to invest into growth. And some of the examples of our roadmaps are video is one of the most critical means of consuming content today. We all know what happens on YouTube. We are improving our video rendering capabilities. We are making sure that the experience through our stable SDK is actually at par, ahead of the competition.
We are working to standardize the integration of ATOM. Some of the members would actually cover deep details into ATOM. Machine learning, optimizations, algorithmic capabilities to improve the cloud cost, because the bigger the business you scale, you have to be very efficient when it comes to the cost of infrastructure. Support of SKAN, et cetera, Google Privacy Sandbox are important initiatives, and PMPs is a means of delivering the programmatic dollars. We have to make sure we have the best tools which are actually utilized in our standards and our out state. All of these are expected to improve our market position and help us into driving better and more revenues. We are another one lucky company to have a very fast-growing position into CTV.
We did some acquisitions, like Liquid, which was Nexstar Digital Media. We also bought Smaato, and Smaato was luckily, building a very strong CTV stack. We have merged our stacks, but this is just to give you an overview of the fast-growing market initiatives. CTV is one of them. eMarketer report says, it gives a view of how the linear TV viewership is actually declining in terms of how much time actually an individual user is spending, and how it is actually increasing for CTV. CTV ad spends roughly $11 billion, is gonna be more than $33 billion by 2025. Again, we have a lot of advantages, and we are one company which are lucky to have it, and we are contributing towards this.
There are tons of rich targeting options which are actually available. They are very engaging and powerful ad experiences, and there is a lot of investment being done to make sure that, how do you measure the performance of ads? How do you re-engage with the users? Because the Gen Z and the coming generation is all actually moving towards consuming the content through these experiences. And diversified and fast-growing audiences. Remco also mentioned in one of the data targeting capabilities, we have a huge, huge amount of data assets which are available because of being omni-channel in by nature. We get ad requests coming from in-app, we get ad requests coming from desktop, CTV, et cetera.
So we get a lot of intelligent signals, and there is a lot of improvisation on the platform to make sure that the audience targeting capabilities are actually much better, and we can actually offer it as a USP out to the marketers who actually want-- who are engaged with us and the futuristic ones. 14% of the group's media volume is actually through CTV efforts and initiatives today, which is actually growing every day as we speak. 60% of the U.S. households is the coverage that actually we provide today, and you can see some of the big logos placed out here. These are the most premium global publisher base. They're not just in the U.S., but some of them actually have a growing presence in EMEA as well.
They all work with us, and on the other side of things, across the platform, we have the top Fortune 50, 500, 100, most premium demand partners. They are already working with us, and there is a lot of value add that we actually provide here because of the cross-platform capabilities, because of the targeting capabilities and the audiences, and the richness and the enrichments that we actually do. And we are again positioned very uniquely because of our owned and operated assets that we can actually provide as an exclusive way of how the audience targeting capability or the reaching the right user, where and why we can actually do a better job. So this is again one major critical investment in terms of what we are actually gonna do in terms of CTV.
Briefly covering some of the investments in the growth of CTV, as Remco mentioned, that we—when we buy companies, we have to plan so that we are efficient. Integrations is a challenging job, but we as a small group are pretty good at it, and we have already integrated Liquid, which was Nexstar Digital Video, onto a smarter stack, 50+ CTV specific features, et cetera. They've new targeting capabilities. They are all now available through one stack. That is so efficient, not only for our partners, but actually for the internal teams also, because there is otherwise a lot of duplicative work, and it makes everyone's life easy. There are some very important midterm projects.
We have to always make sure that, this whole real-time bidding, auctioning is happening on protocols and some, you know, defined means. We always upgrade ourselves to the latest protocols. We are always investing, putting more efforts into the inventory management system. It's extremely important when you go talk to big media planners and buyers. They talk about: how can you actually give us some tools that how we can actually do inventory management? Because we have to... the guaranteed buying or at scale, ensuring that we will not fail, to deliver their campaign. So we are actually constantly investing into that, updating ourselves with the IAB placement guidelines. So we very actively work with the IAB. IAB is placed in the US, IAB it's in EMEA as well.
We contribute towards with their IAB guidelines and frameworks. CTV still is a newer market in terms of what next set of tools and developments that we actually be doing. Video recognition, finding the content, what sort of automated content recognition, the genres of the stuff that we see. So we are investing into all of these things, and again, all of these are expected to actually help us to get a fair share of the market and also make sure that it is helping in driving the revenues. So there's a lot going on, and I would hand over to Remco to actually talk about data assets and the importance of vertical integration that brings efficiencies.
Thank you, Sameer. Yeah, I think it's really, again, to say, in app, we have really been building a very strong position in the last years, making us a leading player in that field. We will further invest in that. I think that's a really good field, and there's more and more going to mobile. CTV, we're still not very big, but we are one of the leading players there, investing a lot in it to really grow also our CTV part, and with those, we have two really very strong parts in the total platform. But the linking pin is... Sorry, the other side. Data. Data is really the glue, and if you look at the gaming part, where we come from, the gaming part delivers a lot of data to us. Each gamer is registered, and with the registration, we get data.
Those that are not registered or don't give an opt-in, which also can happen, we get a lot of contextual data from them. So data, there's a lot coming from our games. There's a lot coming from our SDK base. So the apps where we integrated, we get a lot of information from that, and we have a lot of data going through our systems. The exchange that we have, there's the so-called bidstream . That's the stuff where the advertising info is passed, has a lot of information. Plus, there's a lot of external sources, and if you combine all those data, you get pretty good profiles, and you can really do good targeting. And that's a project we've been pushing in the last months, quarters, to really get better at this. So really getting data. One thing, very important, consent. You need user consent.
There's so much history in this market, and some of the later speakers will talk about that, about getting everything too chaotic or not consent-based. You need to work with consent. You need to have a trusted service. Tons of data also creates tons of complexity, and for that, we have artificial intelligence. Even though it seems to be like it's almost new because of ChatGPT, it's in every newspaper, it's in every news show, but AI is something that we use already for many, many years, because without AI, you cannot run a business here. Rosana will go into more depth later at this topic, but AI is super important to make something out of it, all those points. The other thing is identifiers are disappearing. Cookies are disappearing.
There are some browsers that already don't pass cookies on, Safari, for example. And there's also browsers which will not pass cookies anymore, and that's changing the world. Also, IDFA from Apple is very restricted now, or IDFA. And that means that you need to also work with alternatives, which is so-called contextual. Without identifiers, without cookies, you still can target very well. Also here, Rami will cover some points later in his presentation. And also a bit about data. We thought it's good, and I mean, it's a little bit technical because when Sameer was just presenting, I saw not everybody looking very happy. But it's a technical world where we are, and it's also good to give you an update on what we are investing in to really build our position stronger. An identity graph is super important.
An Identity graph is kind of a translation tool, which gives us the possibility, if you see in the bidstream a cookie, to see, does it fit to a certain mobile phone, to a TV? Does it fit to a certain IP address? And with that, you can best target. You can also avoid that you get too much ads for the same person on a single day, for example. So that's very useful, very important. So that's something we have a cooperation with Roku Ads that we just started, which is an external vendor, where we have close cooperation to work on, for example, ID graphs. We want to get more data in the bidstream. So at the moment, we serve, let's say, or we get from an app, an ad place that needs to be auctioned.
The more information we pass on, as I said before, to the advertiser, the better, and we can enrich the signals with a lot of things. So that's something we are working on. Proprietary data platform, we collect all the data that we have in the group in one platform, all with consent, again, and make sure that all the platforms that we have, and it's getting less and less because we're integrating the platforms, have access to those data and can optimize those services with it. So those are a few examples of things that we are doing. Data targeting is driving the businesses that we have. Then coming to the last point, which is a bit, yeah, how to say it? A central point in here. If you look at programmatic advertising, it's a chaos.
There's so many parties in this market, and as a result, you see that an advertiser works with an agency, and the agency works with the DSP, demand-side platform, that works with other DSP or an SSP, and there's many parties in there. Very obvious that you have a lot of people that take a share, which is very costly. Also, people having old technical platforms that need to be paid, also very costly. Transparency gets totally lost, of course, in this. And when there's a cookie, it's a bit easier because with a cookie, you can always pass on the information. Nevertheless, it's inefficient. And it even happens that a certain company that's advertising a product, that in one auction for one ad spot, it's competing ads from the same advertiser, just coming via different platforms. Super inefficient.
Talking about it, we have a super scattered landscape. There is thousands of companies in this market, but we see a strong consolidation going on. What are we doing different? We say, we planned, and that's because we also saw as a gaming company, it's inefficient to work with all these parties. So what you want is a single platform that connects the demand side, supply side, and is basically fully transparent in there. Then you avoid the problems that I just talked about, and that's what we have been working on via M&A, via integrating the M&A that we did, or the companies that we did, the platforms that we did, and via making that more efficient and also investing in optimizations, like we gave some examples before.
So this is basically what we stand for: more transparency, direct connections, less costs, and much better bid, bidding patterns, so no competition between ads. The possibility to really make sure that somebody doesn't get served 10 times the same ad in a row. All those kind of things, very important, therefore, we are a strong believer in a vertical platform. Then some things on the roadmap, integrating stacks. As before already mentioned, integrating teams goes fast, relatively fast. Integrating technology stacks is not so fast. And also, if you fully concentrate on integrating everything, you might miss building features, you might miss working for your customers.
So we have to find a good balance between integrating the stacks that we have, in the end, with the target to have only one stack, and we did some really big achievements in the last half year, but we still continue to further integrate stacks. That leads to cost savings, but also to even more efficiency. All our stacks are connected to each other, so we have the possibility to really get traffic from one stack on all the other stacks. So that is a fast measure that we did immediately. Then insights and measurement. It's great to say that you're better, but you need to prove it, so you need to have capabilities like A/B test, which is super simple, but also possibilities to really measure more. We are a multichannel stack, not an omnichannel stack.
Audio is missing at the moment, and that's something that we are looking into, if we also add audio. Last point on this slide, operational excellence. It's complex market where we are. We're doing a lot of things, so it's super important to have the teams really organized and really making sure that we make our roadmaps and that we implement all the things that we do. So that's a bit of an overview of what we're doing and how we're growing this company. Before we go to the coffee break, I would give the possibility to ask questions if there are. So I hope this gives... It's a bit different than we did it last year, but we want to give you a bit more in-depth about what we're doing, with the risk, of course, that we get sometimes a bit too technical.
We tried not to, but I think it's also good to see there is a lot behind just serving an ad, and that our Make Media Better is really something that can be done, and that we work in a market that still has a lot of chaos, where there's still very high margins and where you really can optimize a lot. Jenny, I think you should take over.
So yeah, I see you. So maybe, Jonathan, you can bring the mic, so we get the questions also coming online. Okay.
Thank you very much. Good morning, it's Fiona Orford-Williams from Edison. Just some quick opening ones. Can you tell us a bit more about the competition in this connected TV market and where you stand versus the what technical advantages you have to consolidate that position? Secondly, on the integration of the tech stacks, I mean, you're obviously making a lot of progress. What is the timescale on the remaining work that's got to be done? And thirdly, can you make some comments, please, about current CPM trends? Thank you.
You want to take the CTV one?
I'll take the CTV one, yes.
Thank you for the question.
Yeah, thank you for the question. So, if you really look at who are the biggest players out in the market today, the names, the obvious names comes around is, is, Magnite, who acquired Telaria, who acquired SpotXchange. They also acquired an ad server company called SpringServe. So that is one bucket that comes. Then you start seeing some other competition is basically IAS bought a company, Integral Ad Science, bought a company called Publica, so they have an ad server. But then you really see the market really gets spread across, and maybe Paul has more details to cover our numbers. I, again, we strongly believe that after these companies, we have built a marketplace, which is, which is actually a very respectable number.
With the investments that we're actually doing in order to make sure that EMEA market and other international markets were ready for CTV contribution, U.S. is by far the number one market today. But I would say we definitely and absolutely stand within the top five competitions in terms of the marketplace capabilities. Now, some of the companies have a very different solution, which is an ad server solution. Okay, so we do have an ad server, which is a component that we are still porting onto our Smaato stack. And we will have strategies and plan around actually how we can actually start fair, you know, getting a fair share into the ad server market.
Yeah, we continue-
Can Gert sell it first to answer the other two questions before you?
Yeah.
Or you want to?
Okay, let's go. No, go ahead. Let's talk about CTV first, then.
Hi, Gert Steens, Sterling. Sameer, can we just put some numbers to this? Because you look in the competition.
Mm-hmm.
I was just checking, excluding traffic acquisition costs, so the gross profit tech stack, Magnite is doing $200 million a year in CTV. You said 15% of your volume is in CTV?
Mm-hmm.
Your extra acquisition cost is something like $100 million a year, so you're doing $15 million in CTV. Is that, is that right in terms of-
That is right. The margin on the CTV business, unfortunately, fortunately, happens to be low, but it's the same for the entire industry.
Yeah.
So you are right. We are very close to that, those numbers, yes.
But it's a very scattered landscape still, and that makes also the possibility to really unite it. Because if you really look at the supply side, there's a few very big players, and then there's a lot of smaller ones. There's one that Sameer didn't mention is Xandr, which was acquired by Microsoft, which has, for example, at the moment, exclusivity for Netflix. But there's a lot of rumors in the market that exclusivity is not working, and we have seen that before, so that will also probably be scattered.
Yeah, and there's a reason I didn't mention Xandr. Yes, it was nicely placed, but Microsoft owns this. The strategies will change. They are actually already changing out, as we see right now.
Yeah, then I would say a few words about stacks.
Mm.
So we have acquired 15 ad tech companies, which, if we really look at the main stacks, talk about six stacks , seven stacks. We have prioritized the stacks that were not really technically fit. The Liquid stack, it was like crashing 3x a day. Not when we bought it, but we increased the volume a lot, and then it was really not stable. So that was one that had a high priority. Then the Vlion stack, which had three different activities on it, so it was integrated into two different platforms, was a high priority, and those already have cost savings. But as said before, you cannot do everything at the same time because then it gets fully messy. So we still have to go on the DSP, on the demand side, we have three platforms where we are at the moment.
That's basically the next big task, where one of them is really super fit, which the Dataseat one. But they also have still an earn-out, and so you need to be very careful, of course, what you do then and where the priorities are. But that's a very important one for the future. And then we have on the supply side, it's basically two big stacks, PubNative and Smaato, if if I talk where they come from. And we have, let's say, focused them differently. So PubNative is more on performance, really in-app performance, user acquisition, installing apps, working for big fast food chains, for getting their apps installed, for example, or for big tech players that want to have, apps in an in store or stores, promoted.
And on the other hand, we have Smaato, which we really do focus as a multichannel platform. So there's the CTV, there's web, and there's also in-app, and which is much more focused on brand. Nevertheless, it doesn't make sense to have, on the long term, two platforms that are having similar components or, let's say, more platforms. So we are further working on integrating, and we already have done some steps, for example, with the engineering teams, where we... If you have two teams that build SDKs, for example, one for PubNative, one for Smaato, those teams work together, and you see already a natural tendency to reusing components and to building like that. So that's basically a part of it we do really more organically.
Like, if you get people to work together, they find out that it doesn't make sense to do something double, and you start already to get a lot of learnings from those kind of things. But the nice effect, of course, is that it's getting much more efficient. You need much less people to watch the systems, you need, let's say, less technology to run it, and also features are not double developed. So still a lot to go, but it doesn't make sense to now close the company basically and to do a quick, let's say, integrate everything. We need to go with the market. We need to develop features, so we need to find a good balance. So to your question, estimate, and please don't pin me on it, 3 years -5 years to really have one single stack.
If you do more M&A, you get one extra. I mean, we have seen that also on the gaming side, that we have integrated a lot, but each time you do acquisition, you have suddenly a new platform. So, yeah, I hope that answers the question. Then on CPMs, as long as demand is lower, there's less bidding. It's like in the stock market, CPMs are lower, and it differs a bit per market, differs a bit per category. CPM is less under pressure because it's just growing fast, and there's a lot of demand shifting to, how to say, CTV. Web is very much under pressure, where we don't do a lot, and in-app is also, let's say, following web in that sense, under pressure.
What we see, and it's difficult, I said, to really get a middle value, 25% lower CPMs than we usually have. And there's, of course, also seasonality, because Q4, there's more demand. But even then, if you compare quarter to quarter, a year, then you see that there's pressure on it. And we all were hoping to see some light in the second half year, but at the moment, looking at the numbers, we see a bit of movement. We have some articles also, but Procter & Gamble, for example, announcing that they will increase their advertising. And that's also what's what happened in the past. I mean, market leaders are the first ones to start increasing marketing budgets to just gain market share.
Yeah, so we used this period very well, I think, by gaining market share, but it will take a bit, and it's... Yeah, crisp, I don't have a crystal ball. Nobody has to say when it's going up. I hope that answers your questions.
... More questions?
Okay, let's see, what more do we have here in the room? Everyone longing for coffee? I think so, right? So, we are going for a coffee break, and we are going to be back at 11:20 for the financial update, with Paul.
Perfect.
That's it. Yes.
All right. Thank you.
Thank you so much.
... A warm welcome back! Without further ado, welcome up, Paul, CFO, for the financial update.
Thank you, Jenny. So welcome also from my side, and good to see so many faces again this year being interested in the MGI stock. Yeah, I will start today before we go into the accounting numbers with some commercial KPIs, which is basically our core business, and showing that we take a lot of market share. That's what we see in the ad impression growth. So we delivered now 181 billion ad impressions in the second quarter where there's 161 billion ad impressions in the previous year. That's basically with every ad which we auction, we have our revenue share. We also see that the ad impressions are growing. So we had 6% growth year-on-year in the first quarter, 13% year-on-year in the second quarter.
That said, what we also saw is, and I think that has been also already reflected by some of the presenters earlier, is that CPMs still being depressed on the SSP side, which also has an impact on our revenues. And that means we need to make more volume, drive, get market share—yeah, gain market share, basically, and adding new customers. And that is what we see within the software clients. So we added now year-on-year 46 new software clients, so that's the large ones, which are doing minimum $100,000 revenues per year. And that's a growth of 9%, we are actually being very proud of.
We also see on the customer revenue expansion and retention rate, that the retention rates with 96% also remain very strong, which means the customers stay on the platform. We are basically onboard a lot of new customers, and that is leading to a positive organic growth. Despite, and that's what we see in the net dollar expansion rate, despite decreased CPMs, decreased budgets. While this number was well above 100% in previous years, we now saw constantly that this is going down to 82%, which is again, lower CPMs, but also reduced budgets, from on average, the advertisers. But as they stay on the platform, we basically can expect a very nice acceleration once the overall market recovers again.
That we take a lot of market share is also reflected in this report, where we see that on the mobile SSP side, for Google, we are in North America and in Europe now the leading player. Since the second quarter, also on Apple, so iOS, in North America, having a market share of 17%, while within Europe, we are the third largest player. That means basically, if an advertiser wants to do advertising in mobile in-app, we are the company to go to, and we have really built a very strong foundation, as we really said mobile first, and additionally, also now expanding further to the CTV business. Looking at the revenues of the second quarter, it's a bit more complex than in previous years. We saw a 3% growth from our core business.
We divested companies, closed some games, which had a revenue impact of EUR 2.7 million compared to the previous year. We also had to have FX headwinds, EUR 1.4 million, which means, while we growing on 3% with our core business, we had a reported revenue decline of 2%. While the growth of the core business is especially coming from mobile as well as CTV, and we also saw these numbers earlier. Looking at the media volumes, 82% is coming from mobile, 14% is coming from CTV, but CTV is currently the strongest growing media volume stream, with a growth of 34% year-on-year. That brings us to the financial highlights of the second quarter, and here we see stable numbers with positive organic growth.
We also see that we have been able to further improve our profitability with EBITDA margin increase from 27% to now 28%. And despite the reported revenue decrease, we have been able to keep our profitability stable with 1% growth, which I would rather say it's stable. We also see that we have a strong operating cash flow, which is close to the EBITDA of EUR 21 million, and a very solid interest coverage ratio of 3 x. Therefore, currently, we see a lot of softness, which is driven by overall macro environments, lower CPMs, but we also expect that to recover in the future.
Looking now at our long-term financial development, and here we see that we have shown for the last 10 years now a very strong, profitable growth, always growing revenues, and EBITDA year-on-year. We have a CAGR of 55%, since 2018, and when we started to further invest into the media business in 2019 and 2020, we really diluted our EBITDA margins quite a bit by doing so. But afterwards, after we integrated the companies, we realized a lot of revenue synergies. We could really bring up our EBITDA margins again to a level between 25% and 30%. And we are very stable now at 29%, and we also expect and being very confident that we can maintain these high margins also in the coming periods, despite short-term headwinds.
What can we expect in the second half of the year? The programmatic advertising business is a cyclic business to a certain extent, but there's especially a very strong seasonality effect. So in the first quarter, where advertisers allocate their budgets to new channels, work on their strategy, doing certain testing for the year, we see during the year that the advertising budgets are increasing, and the fourth quarter is always the strongest one, where you have Christmas, Thanksgiving, Black Friday, all the events where the companies drive a lot of big marketing budgets. And that means for the second half of 2023, we still expect a growth in revenues and EBITDA compared to the first half year.
Looking now at our operating cash flow and CapEx development, and here we see that we have printed now EUR 101 million in operating cash flow on a last twelve-month basis. We had a pretty strong working capital, in fact, in the second in the first half year of 2023, which is also in relation to our securitization program, where we used less of the program, as well as, yeah, some accrued expenses after the restructuring of the games part, which were still, yeah, already recognized in Q4 last year, but now running basically through the cash flow statement. We also see that we have increased the interest expenses on a LTM basis to EUR 32 million, compared to EUR 23 million by end of the year on a LTM basis.
We expect here that this goes up to EUR 40 million, maybe EUR 42 million in terms of cash interest expenses. But we have also hedged 65% of our long-term debt, and therefore mitigated already during Q1 the further increases on the interest expenses. On the right side, we also see that following the strategy with less priority on M&A, we've really reduced expansion CapEx tremendously over the last years. We're also able to further decrease our maintenance CapEx from EUR 10 million in 2021 to now EUR 8 million on a LTM basis. And therefore, giving the lower CapEx, we basically have been able to drive more free cash flow after interest expenses.
Important to mention also that we're still investing into further growth, which is also reflected in a EUR 3 million investment, which we did in our AI platform and identity graph in the second quarter. There will be much more on these topics and how that influences also the future targeting for our advertisers, as well as will also drive more CPMs and the better monetization for the publishers in the future. That brings us to our net leverage and interest coverage ratio. In these times, obviously, it's important to keep an eye also on the balance sheet.
We see here that since a spike in our net leverage at 3.7 x in the second quarter of 2022, where we had some earn-out payments, as the companies which we have acquired were performing extraordinarily well, we have been able to decrease leverage year-on-year to 3.2 x, and also expect until end of the year to be again, on an annual basis below 3 x, and within our financial targets. On the right side, we see the interest coverage ratio decreased, but it's still very strong and solid with 3 x. That brings us to the MGI balance sheet, and here today, I would like to emphasize basically on three things.
We have a very strong cash position still, with EUR 108 million, which also enables us to invest into further growth of the company. It also enables us to take advantage of certain opportunities, while, and I think that was something we always have communicated, also want to keep an eye on our net leverage ratios. The second thing is that we have a well-diversified, long-term financed debt profile. So more than 90% of our maturities within our bonds are maturing in 2026 and 2027, and therefore, we already have taken care of all the refinancings over the last quarters, and being therefore well prepared to navigate all the cycles. Looking...
The third thing, the last item I would like to highlight is that we still have quite a bit of earn-outs on our balance sheet, so EUR 109 million as of the first half year, and that includes EUR 86 million cash earn-outs, as well as EUR 26 million payable in shares. And here, due to the current macro environment, limited organic growth, which is then also directly impacting the earn-out liabilities, organic growth stays limited. We expect that we don't need to pay this to a large extent, and therefore, we're also obviously reevaluating it when we have certain trigger events. That brings us to our updated guidance, and for the first time since we listed, that we're not increasing the guidance during the year.
We currently see a certain softness in the market, which we didn't expect by the beginning of the year. What we basically now guiding on is a stable year-over-year performance, normalized for FX and divestments. That is what we see here on this bridge. So we really wanted to make it as transparent as possible. When we start at the EUR 324 million revenues, which we saw in 2022, then we taking the current FX rates, we have headwinds, especially in the second half year of EUR 20 million. Then we have the divestments of EUR 9 million, and that brings us basically to a normalized year-over-year performance of EUR 303 million. That is what we see also in our updated guidance.
While on the EBITDA side, and I think, that's also important to emphasize on, we have a very resilient profitability. We have a very strong natural hedge, between the revenues. They have FX headwinds, but also therefore on the cost base, and expect, therefore, to be able also to be at the same levels, like in the previous year, taking also into account the cost savings, which we already initiated now during Q3. That brings us, to the last slide of the presentation, the financial key takeaways. So we have a positive organic growth, and a very high profitability. We have a strong customer and ad impression growth, so we're really building the further base.
We are the market leader now on Android and on iOS in North America, and really can grow from that base in future periods. We have a very good natural hedge in our earn-out liabilities, which is basically hedging the current macro environment. We've initiated a cost-saving program to really mitigate lower revenues and enable further stronger free cash flow, as well as deleverage in coming periods. The full effect we will see in 2024, but we will see already effects starting this year. And last but not least, diversified bond maturities. I think we have really taken care of refinancing the long-term debt, and therefore being well positioned to navigate the cycle. That brings us to the Q&A.
To the Q&A, yes. And we are now opening up the phone for incoming questions. So, and the web is always available to send in questions, and I have received a few of them, so we will take them here in this session. And we're gonna start with the room, of course. So please raise your arm so I can see where to guide Jonathan. So we start here. And please, when you ask a question, introduce yourselves to help all listeners online.
Hi, Niclas here from Oberon Family Office in Stockholm. I have three questions. First one on your working capital securitization program. That was quite a big bump in Q4, and you've seen some release in Q1 and Q2 from that. Could you speak a little bit about why this has reverted? If that's a lot of seasonality effects or if you just see it less sort of sensible in this environment. Second quick question on your interest rate hedging, just what your sort of average maturity on that is. And third, on the EUR 10 million cost out program, is that all going on the EBITDA line, or do you see some of that going on capitalized R&D as well?
Thank you very much for the questions, Niclas. Starting with the securitization program, so it's a quite flexible program. We can always decide basically every day, how much do we want to securitize, and basically sell off our receivables. We have made less use of that in the first half year, as we did not need the cash. But as also the receivable portfolio is growing basically together with the revenues during the year, there's also the possibility to even securitize more, by end of the year and make a true sale, basically, and therefore also decrease leverage, and maybe also doing more bond buybacks and all these kind of things. Things which we have already initiated also during the first half year of 2023.
Second question, that was on the cost savings or?
No, on the rate hedges.
The interest rate hedges. So we have now hedged 65% of our long-term debt, and we have hedged basically EUR 100 million out of the 2026 bond and EUR 150 million out of the 2027 bond, and that gives us a very good risk mitigation for further interest rate increases. While we also kept a certain amount open to make sure that once interest rates decreasing again, where based on swap rates, that's expected for the second quarter of 2024. No one exactly knows when it's happening, obviously, but therefore, we also have a certain upside from that and have mitigated the majority of the risk.
Talking about the cost-saving program, that's largely driven by personnel headcount and, certain amount is also going to the capitalized expenses, so we expect to further decrease also our maintenance CapEx, on that. A little bit is also expansion CapEx, as we also have certain streamlining on projects, while we're further investing into future growth, which I think we also see in the EUR 3 million investment, into the identity graph, in the second quarter. But the majority will basically go into the EBITDA line a while.
And that's also important to mention, we still increased personal expenses this year compared to the previous year, and we're basically going back to a level where we have been in Q1 2022, and therefore, a certain amount of the effect will already be in this year, especially in Q4, while the full effect will then be realized next year.
Okay, just a quick follow-up. So the EUR 10 million, that's the total, including the CapEx line and the EBITDA line?
Exactly.
Thank you.
Okay, so we continue here in the room. Yeah, please.
Hello, it's Philip from Warburg. Can you talk a bit about the change of the audience, given that we... Well, basically, the gaming content is really an important part of your business, and while in the past we used to—we had a vigorously growing gaming community, and that has now changed a bit. So how do you look at this change of audience dynamic, moving from a tailwind to a headwind and hopefully again to a tailwind?
That's a very good question, Philip. So 50% of our supply is coming from gaming companies, so therefore it's our largest supply. And what we're currently seeing is that gaming companies, mobile gaming companies, being rather flat year-on-year, and that also CPMs were further decreasing. Also, according to the latest Newzoo report, they expect also the mobile games industry, revenue-wise and ad spend, rather being flat this year, while there is a good rebound expected for the coming years, so for 2024. And obviously, we-- yeah, somehow also floating together with the supply which we have there. That said, we also have further diversified over the last years.
We actually started with 90% on the, on the game side, and we also have added other channels and verticals like weather apps, social, as many others. And therefore, yeah, we have also mitigated that risk to a good extent. But yes, 50% is still coming from mobile, from the mobile gaming industry, and therefore, when we see a recovery there, we also will see a further recovery of the organic growth of MGI.
Hmm.
Take the microphone.
Yeah.
Take the microphone.
Remco, come on up. Come on up.
Coming up. Okay, otherwise I get in trouble.
Right now.
Now, we have seen in the gaming, and I think because Sweden is a very gaming—there's a lot of people invest in gaming also. We see from the advertising side, of course, also some changes. So hyper-casual gaming, for example, which was a substantial part of ad spend, has basically totally disappeared, because the CPMs were down. The business model of a hyper-casual game, which is basically spending advertising money to get a user and to make more money with ads with the user than what you spent because of lower CPMs, has declined. So that's where we see a whole sector, basically, not spending advertising anymore. Then what we have also seen is that especially the private equity-funded companies, smaller companies, have also spent a lot less.
You have a lot of smaller game companies that just spend less at the moment than before.
Hmm.
To just add it.
Hmm. Okay, I think you can stay, Remco. Might be questions, some more questions for you. So, anything more from the room? I think I have— Yeah, over here. Yeah, Garrett, please.
Thank you. Gert Steens at Sterling. Just, in Q1, you gave us numbers for the games and the ad tech business. You said the ad tech business had grown 8% in Q1. Can you give us the number for Q2 as well?
So we-
I was a bit surprised it was not in there.
We saw a further decline in the games revenues year-on-year, also due to the divestments which we did there, which is quite similar to the ones which we saw in the first quarter. While on the programmatic side, we still were growing organically, but at a low single-digit growth number there.
Okay. Thank you. Then on the, so the change in the guidance, you've taken EUR 35-45 million off your top-line guidance, which is, what? 25%, 20%, 25% of your H2 revenues. Where did it come from? Is that from disappointment on the game side, on the third side?
It's basically organic growth. So we have still expected a high single digit to low double digit organic growth from the overall programmatic business. We expected that also that CPMs will see a faster recovery already in Q2, which was also somehow what the market expected together with us, and that also the games industry is recovering faster than initially or what we're seeing now. Given that June was still a bit weaker than we expected, and also July from a CPM perspective was not recovering, we just said, "Okay, let's update our expectations for this year." Also taking the FX headwinds and everything into account, obviously.
Therefore, yeah, we expect now a stable revenue performance year-on-year by taking these effects out.
I think you see it overall in the market. I mean, we were—when you look in March, April, everybody, and May, everybody was much more positive about the second half year. We were also, there were some signals, but if you really look at it, especially June, we had something like, "Okay, this is not going the way we would like—would have loved to let it go." And we are now rather safe than sorry, which means really we have reduced guidance. It can change, but at the moment, we thought it's better to guide lower, instead of then, yeah, to come later.
And maybe to add one thing, I think, yes, revenue guidance has been reduced, but looking at the EBITDA, I think it's very resilient, and we have also a very strong natural hedge here. It's not just the revenues, it's also the cost base, which is very variable to a large extent. Then on top, we are also, through integrations of certain tech platforms, just in the second quarter of 2023, we reduced tech costs by more than EUR 3 million year-over-year. That's also now in our report reflected. There we can really see that the integrations of various tech platforms drive a lot of efficiency, and that really builds a much stronger base now, also for the future periods again.
One thing which we also see in terms of raising the market share is that the ad impression growth is also due to a reason that we're really adding or we're sending out much more bid requests now, which in the old days would have meant that we're increasing our tech cost tremendously. Now, we actually, with our new setup, also using hybrid cloud and these kind of things, we're really being able to scale the ad impressions much more. And that will enable us, as well, once the CPMs are coming back, to really drive much more organic growth for future periods.
I think, yeah, streamlining the business now, being very resilient and making sure that we have a stable business for the time being, maybe also further upside in the second half year from organic growth, but we rather want to be a bit more cautious now and therefore updated the guidance.
... Yeah, happy with that? Yeah. I'm gonna throw in a couple of questions from the web related to the questions we got here. So the first one is: Do you see the first signs that the advertising market in which MGI operates will develop better in the coming year, 2024? I don't-- Not sure who I'm gonna throw it at.
I can start.
Yeah.
- and then you can add to it. I mean, we don't have a crystal ball. I think this-
Just want to say, as long as you don't steal my crystal ball.
I learned from you, so therefore we don't have a crystal ball, so therefore it's hard to predict what happens in 2024. One thing what we see already, without giving too much details now, but that is, while July was also still quite weak, then in August, we see some signs of recovery. How confident this is also for a longer period of time, we will definitely see. But I think also with the CTV business, for example, where we're also doing further investments, we have a very good base to further grow the company, even in a bit more challenging environment.
2024, there's a lot of expectations from many market participants, analysts, et cetera, that we will see, maybe even in the second half year already, a stronger recovery, and that 2024 will be a good year. But who knows, basically? It's very hard to tell, and we rather want to be cautious, and I think that's the right thing to be in that environment and underpromise and overdeliver.
Yeah. And I would turn the question back, actually, if the person who asked it can tell us how the economy will develop the next quarters, we can tell how advertising will develop. No, it's really difficult to say. There's many variables, of course. Also, interest rate, I mean, is one of the things that will affect us at the moment. It looks also like we have a negative there. But looking at the market, the big question, I think, to everybody is, what will happen with the consumers?
Mm.
Because we see on the company side, they have increased margins, time, increased profitabilities. But you hear everywhere that the... Yeah, the sales are not, the volume sales are, the volumes of the sales have also been reduced. And the question is, what happens due to higher energy prices, the inflation, high mortgages, et cetera, to the consumer, and are we getting into recession or not? So there's so many variables in there. We remain optimistic, but the only thing that I can tell you for sure is we will further work on gaining market share, even if CPMs stay long, if low, even if, demand stays low. But, we hope also that in, let's say, 2024, we will see really some tailwind.
I think we got it that, right? So we continue with three more questions, and it's more around the guidance. So the first one is, is. And it's coming from the web. "Is mid-term outlook still plus 25% revenue CAGR? How do you plan to achieve this M&A?" In brackets, question mark.
Should I start, and then you can-
Mm-hmm.
So, yes, the midterm financial targets are still being 25%-30% CAGR on the revenue side, 25%-30% EBITDA margin, 15%-20% EBIT, as well as keeping the leverage below 3x. And with the organic growth, I think we saw in 2021 with 38%, that this company is really built to drive organic growth. Yes, we have some short-term headwinds, but we really expect when the CPMs recover to a good extent, that the net dollar expansion rate is also going above 100% again, and that automatically has a very strong impact also on the organic growth drivers of the company. And on top, yes, M&A is still part of the story.
We have been not putting as much priority on it for the last 1.5x years due to the environment, as well as, yeah, the target from the 3.7x to delever to below 3 x again. But there's definitely a lot of opportunities out there, we also look at, but it needs to fit the balance sheet and the risk profile of MGI. But I would say, yeah, 25%-30% is still our mid-term financial target, and we are very confident that once everything recovers a bit more, we are also being able to achieve that or maybe outperform it.
Mm.
I would like to quickly add, M&A, also reason for not doing M&A was that, public company valuations have come down very much. Private markets have not reacted that fast, but we see also there that the valuations are coming down. Also companies that have problems with the financing now, so we see a well-filled pipeline. But money has become much more expensive, so we need to be much more picky in what we do.
Mm.
So that's actually on the M&A side. But I'm convinced with everything we have here, we can also do the full growth organically now, and especially now, if we get, the winds from the back and not from the front-
Mm
... you will see for sure when it recovers with higher CPMs, et cetera, that it really can grow very nicely.
Mm-hmm. And speaking about money, next question from the web, coming in then, is: "What are the plans regarding bringing down the debt?" So maybe just come back to that.
So we have now EUR 108 million cash. We also expect to increase that until end of the year due to seasonality and the effects which we discussed earlier on the working capital. We definitely also want to keep a certain liquidity basket to take advantage of certain opportunities which you cannot plan. So, and especially what Remco just mentioned, there might be companies which have refinancing issues, where you actually can really, yeah, take that as a good chance, to combine both companies, get additional cash flows in without putting much more risk on the balance sheet. And therefore, it makes sense to have a certain liquidity buffer, and that's also what we learned over the last five years. The best deals come within two weeks or three weeks-
Mm
... and you need to have the financing in place. And therefore, yes, it's a lot of cost which being associated with it. And we also want to further buy back bonds as we did already in the first half year, but we also want to keep a good flexibility to really be able to take advantage of the opportunities which arise.
... Yeah. Okay, a last question from this investor analyst from the web. I don't have the name. Sorry about that, just see the questions. So, by what percentage have the CPMs decreased compared to 2022 and 2021? I don't think-
Yeah, I answered, or let's say I talked about it before.
Yeah. Yes.
Difficult to say because different segments, different countries, et cetera. But what we, let's say, what we see in our systems and also comparing to, what we hear from peers, talking about a 20%-25% reduction.
Mm-hmm. Okay, good. So, if you want to ask questions from the room, just raise your hand, and I will continue with the web because we get questions from the web, but you need to let us know. Okay, so from the web, and this is, yeah, yeah. Yep, yeah, more, more, yeah, I take it. Alphabet and The Trade Desk seem immune against the advertising cycle, yeah, cycle-
cycle. Yeah. Thank you, Chris. Thank you. This year, what is your explanation for that?
What we see, at least currently, I mean, we can't look in every detail-
Why are-
But we have-
Why are you different-
Yeah
... so to speak? Yeah.
No, what we see, I mean, we're talking about it. It's a digital market where larger parties with larger shares can easier, let's say, expand margins.
Mm-hmm.
We see also that the DSP side, so the demand side, has a bit more flexibility there than we have, which was mostly geared to the supply side. And what we are assuming, without knowing all the details, is that, first of all, they have increased margins, which shows, let's say, more revenue then. And secondly, that we also see that the larger parties, the, the Procter & Gambles, those kind of guys, are faster in raising their budgets, and they work with the big holdcos and mostly do the budgets via Google and via, Trade Desk-
Mm
... while we are more mid-market, which we are serving-
Mm
... and that is, what we assume is the difference.
Mm-hmm.
Additionally, what we said earlier as well, 50% of our supply is coming from the gaming industry. That's completely different than The Trade Desk, for example.
Mm.
Therefore, we floating much more with the gaming industry, which according to Newzoo, is rather flat this year.
Mm
... and recovers the next year.
Mm.
So therefore, yeah, in the details, there's a lot of differences.
Yeah. And from the same investor analyst, then just clarity around who are actually, you know, your competitors? Who do you seen as your core competitors?
Yeah, we had a nice slide before, which shows, like, hundreds of companies.
Yeah, but-
But if we really look in the sectors where we act, and I think we talked about CTV before, we talked about the Magnite and Xandr and Publica, basically.
Mm-hmm.
And if you talk about in-app, we have an InMobi, we have an PubMatic, which are strong there. And let's say, if you then look at, for example, at Google, which is super strong on on web, that's a part where we are not even that strong.
Mm.
So there is a bit of split in the markets, who's serving what and where.
Mm.
Those are some of the main competitors. There's tons of competitors, also a lot of smaller competitors.
Mm.
But maybe also important to add, we don't regard them always as competitors. It's basically also partners to us. So we work on the demand side with a lot of demands and platforms like The Trade Desk, so they bring a lot of additional demand also to our supply side. While even on the supply side, yeah, there's also other companies which don't have the direct integrations to the publishers, which then use our connections, et cetera. So on both sides, we're working with a lot of partners, and in this industry, the more demand and the more supply you basically drive is better for everyone, and therefore you can increase overall revenues and volumes.
I think, especially in the open internet, where we have a lot of transparency, yes, it's maybe harder to increase take rates, or it's almost impossible, basically. But overall, I think also in the long run, transparency always wins, and therefore, yeah, we are very confident with our business model that we can really sustainable grow that long term.
Mm.
All the names that I gave, we are working with.
Yeah.
So it's really-
Yeah
... they are partners like ours.
Okay, thank you. And then, and I will continue. So, and, now comes a question. It's a game, game-related question. So are you planning on making a mobile version for one of your PC console games? Maybe-
Jens is the right-
Exactly
... person to answer that.
Jens, welcome up.
Give it.
The short answer is, maybe. There's not at the moment, but we are looking into opportunities. I mean, we have in our portfolio with Wizard101, we have two games where we believe they could fit very, very well on mobile. What we now do in the very short time, and therefore it's, yeah, it's maybe also a yes to answer that question. We have some casual games in our portfolio on mobile. One game to mention, it's called Pirate Freedom, where we're currently in the process, and I can mention that already because it's done within the next 14 days, which we're gonna reskin to our Pirate101 license.
So gonna using the KingsIsle license on one of our already existing casual games. But that we are planning now from scratch, starting a development of mobile games, that's nothing we have done in the past, and that's, at the current stage, also nothing which is on the table. But for sure, always looking in those kind of opportunities, but more in the direction of reskinning already working apps instead of developing from scratch.
And also bringing the games on more platforms.
Exactly. What we are working on to expand the question, maybe what we are working on and what we're going to see in the next 6 months - 12 months is, we're going to expand our portfolio to distribution platforms, which is not only including when you talk about platforms, you only think about mobile, Xbox, Sony, but distribution platforms also means Epic Store, Steam Store, Humble Bundle, territories like Southeast Asia, Middle East, and so on, and so on. That's what we are actively working on. So we are really much more focusing on our internal portfolio at the moment to expand our portfolio. There's still a lot of opportunities which we have not maybe taken the last two or three years, where we see a lot of opportunities, and that's what we're actively working on.
Yeah, further development of the games that we have. New features, new worlds, new additions.
Always bread and butter.
Yeah.
I think one example to close that maybe is also the multiplayer, which we are planning, for AxesInMotion , which is a very cool feature, I think. So there is further development, but as said, no fully new game development for the time.
Not announced yet, but therefore, multiplayer AxesInMotion now announced.
It's in the report.
Okay, I think we got it. So do we have any more questions from the room? I look at Jonathan here. No? We have no more questions from the web. So I'm thinking that Paul and Remco maybe wrap up the financial update, and after that, it's time for lunch. And after lunch, we will be back here at 12:30 P.M. with the expert sessions. So yeah.
I would leave the wrapping up to our CFO. So, up to you, Paul.
Yeah. Thank you very much for joining for the financial update. I think today we have many experts here, and the coming sessions will be even more interesting. And today, we try to keep it not too technical, to really make sure that our stakeholders really understand what we are doing, and that's the target of the day. And yeah, welcome us for the upcoming sessions, I would say. Thank you.
... A warm welcome back, and it's time for our expert sessions, and follow on the expert sessions, we will also have customer story presented after that, and MGI story, just to let you know what's coming up after. And then we will have the final with the panel discussion. So coming up first is Jay, Founder, CEO, Managing Partner, Digital Capital Advisors, right?
Bingo.
Plus 25 years, deal maker, banker. I don't know what more to say.
But I can see, though, on your CV, all the deals, of course, you have made. You can see it up here. It's quite impressive. I can also see you were coming back from retirement in 2010.
I did. I founded the firm back in 2010. I used to own. I'm actually an old media guy. I started a magazine out of college, owned a number of different media companies. And, and I always found that, the bankers used to call on me smart, financially oriented, terrific, but not one of them had ever run a company before. So I thought there was a really interesting opportunity to create an investment bank by entrepreneurs, for entrepreneurs. So that's why I came out of retirement.
Mm. This is going to be interesting.
Well-
Mm-hmm.
Let's hope so. As my wife said when she saw this picture, she said, "You need a new one. You don't look like that anymore." Thanks, sweetie. So, Remco and team, thanks very much for inviting me. I appreciate being here. Appreciate everybody in the audience. This is the part of the presentation you're gonna lean back. I know you leaned forward when the financials were discussed. This is the banker talking now. You can relax, have a cup of coffee, and enjoy. So, I do have perspective relative to advertising. First of all, you know, ad tech, right? Ad tech, advertising technology. It's really two different scenarios combined to one.
The advertising side is really quite interesting in that the technology has really revolutionized, as Remco was talking about earlier, but I think all of you know, has really revolutionized the buying and selling of online advertising. It's been very unique, it's been very disruptive, and it continues even today. But it's not perfect. It hasn't been perfect from the beginning of time. Advertisers are still asking the same question: "Why aren't my advertising dollars being 100% effective?" And it sounds like that's new, but it's not. John Wanamaker, one of the famous retailers back in 1919, said, "Half the money I spend on advertising is wasted. The trouble is, I just don't know which half it is."...
Today, what I'd like to do is I'd like to talk to you about a little bit about media then and now, some shifting demographics, which I think is important. One of my goals today for this little conversation we're having is to ultimately bring it up a notch. You've been in the weeds with the numbers and the granularity of the business. I hope to put a little context around the whole day for you as it relates to this. Programmatic advertising, which is obviously core, vertical integration, consolidation, and then, you know, a view on what's coming up. Media then and now. I think Remco had already talked about... Media used to be very simple. Frankly, it was pretty simple when I got out of college. There was no internet, right?
There was only a few mediums out there. In fact, my kids today can't believe, I go, "Dad, you didn't even have an ATM back then?" But, you know, there was radio, you know, 100 years ago, radio, there's mail, there's newspaper, there were some magazines, magazine. There, The Atlantic, I think, is 170 years old. TV was just one, you know, four channels, black and white, became color, billboards, a little outdoor. But we know the world has changed significantly today, right? When you take a look, I mean, I don't need to tell anybody in this audience, take a look at the music business. Spotify, your hometown company, has revolutionized and democratized the, online music business. Out-of-home is becoming a very digitized, medium these days. TV is no longer just a single channel.
You've got cable TV, paid TV, you've got a number of just streaming and a variety of different things. So it's a very complicated world today, and that's just talking about the online side, as Remco and his team talked about earlier. So has it made-- You know, is, is the ad tech landscape a lot simpler than it was? Not really. You saw the slide. I'll show it to you again. This is the one that, Remco had put up a little bit earlier. I mean, this is the landscape of online media. You've got agencies. The agencies themselves have trading desks, you have DSPs, you have ad networks, a variety of other point solutions like measurement companies, data companies, data suppliers, all of them coalescing into what is called the online digital landscape.
Very complicated, but the other thing that people really aren't talking too much about is the shifting landscape of the consumer. Right? In the United States, the Hispanic population has grown 88% since the year 2000. Think about that, 88%. It's projected. Right now, 20%, it's projected to be in another 10 years -15 years, 23%. To put that in perspective, African American population is 13%, Asian is 5%. In 15 years -20 years, the Asian population is expected to grow in the United States to 10%-13%, or 10%-13%, Hispanic to 20%-25%, and the African American community is gonna remain relatively flat, according to the statistics.
That means the changing face of the consumers that these brands are trying to reach and the technology companies like MGI are facilitating the advertising to these individuals is changing, and folks need to be aware of that. Same thing's happening in Europe here, right? There's been a lot of folks that have come over. The demographic. If you take. Let's take a look at the Muslim population in all of Europe in 2010 was 3.8%. It's projected to be in another 20 years, 11.5%. That's not good or bad, it's just different. And you need to understand what your consumer is doing, who they are, what they're doing, what their, you know, what their buying habits are, because the old consumer is very different.
Of course, this is what's done it, right? This is what's helped facilitate that responsiveness and the ability to directly target the individual consumer, not the group, but the individual consumer, who then can make a decision whether he or she wants to buy the product or service that's out there, that's being advertised. So the top line is pretty much the same as it's always been. Advertisers want efficiency, transparency, and performance. That's what they've always wanted. John Wanamaker was complaining about it back in 1919. But programmatic advertising has allowed for the level of speed to be able to make that buy, that decisioning buy, is unheard of in the perils of advertising. Number two is they can make, with analytics, they can make decisions to target hyper, hyper audiences, right?
They can get down to the granular level as it relates to audiences. On the transparency side, the slide I showed you before, the one about the LUMAscape, every one of those guys, when they're involved in a campaign, is gonna take some money. It's called the ad tech stack - ad tech tax. And brands and agencies are tired of the ad tech tax. They wanna know about it, because oftentimes it doesn't happen above board. So the transparency side of programmatic advertising has helped to allow that to be more visible. And on the performance side, you can monitor performance, and this is something really unique to, advertising, right? Which is you can not only monitor the performance, you can change it, optimize it real time. So has it improved the landscape for efficiency, transparency, and accountability? Well, the answer is yes, right?
Just like the Swedish stock market, the London stock market, Nasdaq, and so forth, everything used to be by hand. Very inefficient, right? Very noisy, too, if you've ever been down in the old stock market where everybody was shouting with pieces of paper. But, you know, looking at the world today, buying and selling stocks, that's what technology does, particularly programmatic side. It looks for inefficiencies, large groups of inefficient inefficiencies, and it comes in and solves that problem. And it's done that pretty much for advertising as well, because as Remco quoted some stats earlier in the day, the advertising market is huge, and the digital advertising market is still growing.
I want to just pause and kind of go sideways for a second because, you know, the guidance has been lowered, but one of the things that's really interesting is online advertising continues to gain greater share of dollars, of those advertising dollars, and that's going to continue because it's more targeted, and it's more efficient, and it's more effective, and hopefully becomes even more transparent than it is today. All right? Again, as I said, an incredibly cluttered marketplace out there of these point solutions. Point solutions, in case you don't know what that phrase is, is somebody who has an idea for a particular specific product that they will get venture capital backing for.
DCO, right, or dynamic creative optimization, which allows for real-time change of creative directed to different audiences in milliseconds, right? That's not a, that's not a billion-dollar business. But somehow, some of these companies have gotten venture capital backing, and they're in there, and they're going to partner with one of, one of the DSPs, they're going to partner with a, with an MGI, and they're going to take their take, right? A lot of these companies shouldn't even exist, and one of the reasons why there's been mass consolidation that continues today, vertical integration. The reason why? Because, because there's opportunities out there to pick up some of these companies at a very spectacular price and value that's going to assist the bigger platforms like MGI and others in terms of being able to capture more advertising dollars.
So vertical integration, I started talking about a second ago. It is a messy landscape out there and has been for a while, right? And consolidation is helping that. The first ones to do it really were obviously the big boys, right? Google, Facebook, Amazon, Apple. Amazon, as you probably know, their advertising business is $30 billion this year. Three zero. What's really interesting about that is it's really small, small revenue of the total revenue of Amazon, and particularly AWS. But regardless of that, Google bought DoubleClick about 10 years -15 years ago, an ad-serving technology platform. Amazon bought Sizmek. They've been building their own technology in-house. They've had an advantage, which is their walled gardens, right?
They've got their own audience that's already identified, that's first-party data, and now in addition to the regular business, they're able to sell them advertising. Small side note: my son happens to work for Walmart Media Group. Walmart's trying to copy Amazon, as all the big-box retailers are, because they recognize they have an advantage, they have a captive audience, they have customers, right? It's interesting in that regard. But one thing I think everybody ought to keep in mind that wasn't talked about earlier today: why? Vertical integration, why? Well, because scale matters in the advertising world. Scale matters in the advertising world. Size matters. It doesn't matter whether when the winds are behind you or in front of you.
In fact, they really, against, against all odds, when the winds are in your face as a business that's an advertising business, when you have scale, that matters. Amazon, Facebook, Meta, and Apple garner 70% of the ad dollars of, of, of the digital advertising dollars, which means everybody else is fighting for the other 30%. Go back to scale matters. You get big, you're able to... Because, because the other interesting trend is that agencies, particularly when times are difficult like they are right now, they don't want to talk to 4,500 vendors. They want to talk to a few. They want to do business with a few. So vertical integration like, like MGI has done and continues to do, and others as well, is really critical because scale matters.
It allows you to increase your market share, it gives you downside protection, and downside protection is basically a moat. Now, the challenge is, it's not all easy, and I think you heard, you heard Paul and everybody else talk about the integration of their technology or the companies they bought. MGI happens to do an extremely good job of integrating, but integrating technology companies is an extremely complex and very difficult proposition, right? So it's not easy to just go out and acquire somebody, a tech company, and roll them into you. There's also, obviously, the cultural differences as it relates to integrating a company in. Of course, obviously, all the governments around the world are paying attention to the monopolies that are being built because scale matters.
So most of this I've already talked about, kind of the drivers of consolidation, but clearly, the death of the cookie is one, right? Number two is AI, data, and analytics. A lot of companies have had, like MGI, have had AI, you know, artificial intelligence as a platform inside their platform for some time now. But the reality of it is, and one of the things that excites a lot of the analysts in the United States, is they think that AI is going to really revolutionize and take ad tech to the very next level. So we'll see. Hopefully, I can come back in a few years and see if it's changed, but the reality of it is it's something that, that's gotten everybody excited because the opportunities are endless. The macroeconomic environment is conducive for picking up other companies in vertical integration.
Remco was talking about a lot of these companies are running out of money. Right? And then the shifting regulatory landscape is something that's also driving vertical integration. This is a relatively redundant slide, but one of the, well, the interesting piece is that DSPs and SSPs are coming together, which is something that had not happened heretofore. They'd always been apart with somebody in the middle. Now, now they're coming together. Let's talk a little bit about the M&A landscape. It's been very robust in 2021 and 2022. Let me back up a half a second and say, the interesting thing about the ad market is, there's always opportunities, even when things are down, right? When the markets are down. When COVID hit, everything across the planet stopped, right? But what happened?
What was the one industry that was, like, quadrupled in size? E-commerce, right? Because everybody had to go, everybody had to buy something. They needed their essentials: food, clothing, whatever, right? Alcohol. And, and what'd they do? They bought everything online. So e-commerce during COVID exploded, right? So following COVID, you had an incredibly active, in the ad tech market, M&A marketplace. When things started coming down economically, higher interest rates, borrowing costs going up, all the, all the, items and issues that you guys know that have put pressure and, and reduced the stock prices across the board of most companies, has created an opportunity in the M&A market for private equity groups.
Because the strategic buyers went away, because they were all working on cleaning up their own companies, making sure that they were tight on their own budgets, reducing ad spend, sorry to say, right? But private equity had been sitting on the sidelines, and now they had no competition. So what's interesting about private equity is, I, I call it the private equity flywheel, right? Because a private equity company... Let's look at Equativ, Paris-based ad tech company, right? Bridgepoint Advisors just bought them. Generally, or we can even talk about Simpli.fi.
Generally, what happens with a private equity company is, they'll come in and buy a relatively large company, a platform company, then they'll put dollars behind that company to grow organically, and then they'll give them money to buy inorganic, to go grow inorganically. So as that happens, then they get bigger. Then what happens? Private equity doesn't hold companies forever. 3 years -5 years, off they go, and by and large, in the last two years, what's really interesting, what's happened? They've been selling to other private equity groups, bigger ones, and then the cycle continues. Simpli.fi is a classic example of that. It was a small, programmatic, advertising platform for local. They were, they were bought eight years ago by Frontier Capital, small private equity group down in Charleston, South Carolina. They grew organically. They grew inorganically.
Four years later, they were sold to GTCR, a much larger private equity group. GTCR did the same thing. Eighteen months ago, KKR bought them. So even when valuations are down, there's always somebody out there, particularly in a market that is so fragmented, to pick up and consolidate in a vertically integrated way. These are more notable M&A things. So to wrap up, if you wanna, you can look at those deals online in the presentation at another time. But looking ahead, so what do we think is gonna happen for the rest of the year? Well, you guys know, the IPO market is cracked. It's closed. Maybe it's opening a little bit. You're seeing a few companies get out now. But it's relatively closed and will remain closed for some time.
There's gonna be accelerated development of AI and ad tech, as I said before. Sadly, we don't believe that ad spend's gonna increase in any dramatic way over the second half of 2023. I think we're more bullish about 2024, for one particular reason. It's been a while since we've been in this, don't call it a recession, but economic slowdown, and advertising has been reduced, right? It's been a while. But at the end of the day, brands are what? Their job is to sell their product, and they begin spending at some point. They're going to spend. Now, they may not spend if the winds don't shift. They may not explode in their spending, but they're gonna spend more than they are right now.
If you look at the history of online advertising, both online and offline, you'll see that in any prolonged downturn, brands started spending because they have to sell their product. There's gonna be a continued move towards vertical integration, and one thing that's certainly not going away is obviously government focus on data privacy. So with that, I will say I will change my picture. I promised my wife. Thank you very much.
Thank you.
You're welcome.
And, you will get opportunity to ask questions to Jay during the, panel, discussions later on. So we're gonna move ahead with the next expert, and, yeah, let's do this. Yeah, I would like to welcome you, Jeff, up.
Thank you, Jenny.
Chief Strategy Officer of Basis Technologies. And I know you have been, you know, founder, CEO of, like, a start-up, entrepreneurial-
I have. I've... I'm far older than I look, and thankfully, this picture was taken a few weeks ago, so my wife isn't scolding me for looking inappropriate.
Yeah.
No, I-
... I began in the enterprise software world, and I've been a hardcore data geek most of my career. I've been very fortunate. I got into the advertising space around 2007, for a very large publisher, one of the large niche social networks that still had some scale before they were essentially crushed by both the 2008 financial crisis and another little social startup called Facebook, which ate our lunch. So I went off and built my own data company, which was eventually acquired by another larger advertising company called Quantcast. And I got to spend several years on the demand side, going out and buying supply. And honestly, I became the data guy at Quantcast, which is a very large data company.
At that point, they were the second largest analytics platform behind Google Analytics. We saw roughly 45% of English-speaking internet traffic and about a third of the overall internet traffic. So it was a tremendous data asset. Following that, I went to the supply side and was also very focused on data and identity for one of the largest mobile supply side platforms in the world. Worked with Sameer briefly, and then somehow I got persuaded to move off into the land of investment banking and venture capital. And I actually helped Nexstar Corporation divest their Liquid asset to a fantastic company called MGI. So I go way back with the team, and I'm the speaker today that's going to try to not bore you to sleep with a whole lot of technology and jargon talk.
But I think there's some critical points coming up in the next few years on the identity and privacy side that I'm frankly surprised I haven't heard more Q&A about from this audience, because I think there's some existential threat risks in what's about to occur. And usually, when I do these conversations, like a real deep dive is 2 days -4 days, a typical board overview is three hours or four hours on this topic. I'm gonna try to compress that down to the CliffsNotes in 20 minutes. I'm gonna try to keep it fairly jargon-free. So thanks, Jenny, here we go.
Yeah, and before I let you go, I can promise you, you will think twice before you open the apps in your phone after this presentation. So here you go.
Thank you. So you've heard almost every speaker thus far, point to the incredible importance of data in this ecosystem of advertising technology. I wanna leave you with some key takeaways that the advertising industry today is entirely powered by data. Data is entirely dependent upon identity. Without an identifier, you simply don't have any way to link all of these valuable data assets to any of these ad impressions. And supply-side platforms like MGI rely on extraordinary data assets and technology to help enrich the value of that ad supply to maximize the revenue and yield for their publisher partners. On the flip side...
On the flip side, the demand-side platforms are also exploiting data as much as they possibly can to minimize the total number of ad impressions that their advertisers have to buy by ensuring they're buying the right audiences, the highest value audiences. Even if an advertiser spends a significant amount of money to purchase data, to leverage, to most accurately identify those audiences, if they're able to reduce the total volume of ad impressions they purchase by an order of magnitude, it's incredibly cost efficient and ROI positive for those advertisers at the end of the day. Ultimately, the evolution of advertising over the past 50 years or so, it's been entirely enabled by data. I mean, in the past, advertising and marketing was largely focused on building the intangible asset of brand value.
Coca-Cola wanted you to think of them and a polar bear and an oddly shaped bottle anytime you were thinking about a sugary carbonated beverage. And it turns out, most of you, when you see this slide and you saw the polar bear, you've got a very high likelihood of associating that with Coca-Cola because marketing works. But the lack of data and the real lack of measurement around brand and how to value brand has left marketers with a lot of challenges for a long time. Then, digital emerges, and every single person is connected to the internet by a device that has a device identifier. They have individual user identifiers, subscriber identifiers, a huge variety of different elements that allow you to very precisely identify individual people and associate a lot of interesting data with those individual people.
It really enabled a critical shift, and that was, I like to say, Moneyball came to marketing. Instead of trying to build this large, fuzzy, intangible concept of brand value, marketers and advertisers got to focus on the bottom line and to really crisply define return on advertising spend. That actually enabled a whole bunch of interesting businesses. Direct-to-consumer businesses wouldn't exist if they didn't have the ability to very concretely measure their customer acquisition cost and estimate the lifetime value of that, and then go and present that positive ratio to venture capital to flood their pocketbooks to go buy those users and build great businesses like Warby Parker or BritBox or what have you. It's been a tremendous shift, and I think it's really opened marketers' eyes to the realm of what's possible in the future.
And now, the most innovative, and these tend to be very large brands like Nike or McDonald's, or Procter & Gamble. They're beginning to carve off experimental portions of their budget. Some of these companies are even buying their own data science and data optimization teams to bring in-house to maximize their own capabilities in this area. Because if you have the ability to precisely measure and highly optimize your touch points in paid media, why can't you extend that to all of your media, your owned and operated websites, your owned and operated mobile apps, how you're optimizing the content that you push out into search and earned and shared in other channels? And that's really the long-term future for the industry.
Frankly, I think the next batch of multi-billion dollar companies in the ad tech or martech space are going to be the companies that figure out how to enable marketers to do that type of optimization across all channels in real time, but most importantly, in a privacy-safe way. That's really the segue to the rest of the conversation. So, you've seen a lot of slides today about how the ecosystem works. I really wanna focus on these purple boxes, and bear in mind that the entire value proposition of supply-side platforms like MGI or demand-side platforms like The Trade Desk is that they can bring extraordinary volumes of data to bear. And we're not talking tens or hundreds or even thousands of potential elements of data.
We're talking hundreds of thousands of elements of data that can help you precisely identify, as a CPG, if I'm trying to advertise a new diapers product, am I actually sure I'm only reaching parents who've had a child or are about to have a child? Because if I'm not reaching them, why would I bother buying advertising places on the wrong people? And to put this, to put it in a scale, I mean, you've seen that MGI themselves are handling hundreds of billions of ad impressions each month. Globally, total, across every platform, the best estimates are that we're currently in a digital realm of around 1.7 trillion ad impressions each day occurring. Now, Google doesn't have all those, Facebook doesn't have all those.
A huge proportion of those are in China, which is a very interesting and very challenging market for a wide variety of reasons, including the Chinese government, who I hope are not watching this on camera. So you're talking about trillions of events per day, with hundreds and hundreds of thousands of data parameters involved, in an industry that, in many ways, is largely run out of Excel spreadsheets today in the agency world. So there's really a spectrum for how things evolve. But underneath all of this data, you have end user PII, Personally Identifiable Information, at the root core of how all of this is empowered and enabled today.
If you visited a website or installed a new mobile app recently, you're going to notice that there's been an exponential increase in the number of pieces of content that require you as an end user to register and validate your email address before you can access that content, before you can play that mobile game, before you can read the news, before you can go look at the dating opportunities in Stockholm when you're visiting there. That's simply because for those publishers to enable all of this to work and maximize the value of the ad impressions on you while you're consuming that content, they have to have your email address to plug you into this ecosystem. Once they have your email address, you are in the system. What's that look like?
Well, as a guy who has bought and sold tens of thousands of audience segments, I can tell you that if I have your mobile ad ID or I have your hashed email address, it's no problem for me to call up vendors like LiveRamp or Oracle Data Marketplace or a wide variety of others and identify thousands of attributes on you. The big data catalogs at the big data marketplaces out there today are offering ballpark 300,000-400,000 different audience variables. And all of that data, that's it. That's the key to all of this enrichment that's possible on the supply side and the very high fidelity, very granular, very precise targeting on the demand side. So big takeaway, the ecosystem today is built upon and is entirely reliant upon PII, Personally Identifiable Information. They need your IDs from your phone.
They need your hashed email. The 10,000-12,000 companies in ad tech, martech sectors, in almost every way, are dependent upon PII. Which begs the question: what happens if those identifiers go away? What happens if PII information is highly regulated or restricted? Well, it's happening. It's been happening for quite a few years now, beginning with Apple introducing a variety of privacy protection tools in Safari. We've seen a gradual progression, and most recently, probably the most significant event in the turmoil of privacy and regulation coming to the advertising ecosystem, has been the changes that Apple made in iOS in April of 2021. So I was an investment banker at the time, and I...
As a banker with a lot of connections in the mobile space, I called up about 30 companies in mobile and just asked CEOs and CROs, "Hey, how prepared are you for this? Are you ready?" And I was frankly shocked that, and I built a spreadsheet, so I actually know the numbers. Close to two-thirds of the companies didn't believe Apple was actually going to do this, and were very poorly prepared for what was about to occur. And what occurred? Beginning with iOS 14.5, Apple required that if you were going to capture a user's device identifier, the ID for advertisers on Apple, or ask them for their email address and use it for audience targeting or advertising targeting purposes, you had to ask the end user to opt in and consent to those use cases.
So they required you get the user to consent, and you had to fully inform the user of all the use cases that you were going to deploy their data for. Now, my question to you is, would you do that? Did you do that? Because during the course of the next 90 days, as people with iPhones upgraded to iOS 14.5, the apps that were installed on your phone, the next time you opened them, they would pop these opt-in consent pop-ups. And I see some of you nodding your heads like, "Yeah, it was actually it was a really annoying 90 days," because it was like, "Oh, my God, I have to register for all these apps that I've been capable of using for free? What the hell!" But that's what was driving it behind the scenes.
Additionally, all of that beautiful data and measurement and metrics capability that enabled marketers to go in and play money ball and very precisely measure the actual efficiency of at a very granular level got completely broken. Apple said, "Nope, no longer. We're going to take any data about ad campaign measurement. We're going to aggregate it ourselves, and you're not gonna see any user identifiers, you're not going to see any PI information on that reporting, and we're only gonna give you it in aggregate once or twice a day." So you can't even get that data in real time anymore and utilize it for real-time optimization, which opens up a wide variety of different challenges.
For example, if I've just paid for the product you're advertising to me, the advertiser doesn't know that, so they might continue buying ad impressions on people who've already bought the product that they're being marketed. In the first 30 days after Apple released iOS 14.5, these companies in mobile, who I'd spoken with and said, "Jeff, it'll be a nothing burger. We expect maybe we'll lose identifiers on 10, maybe 15% of mobile ad supply." Well, they were in for a very rude awakening, because it turns out that you, and I, and the rest of the world that has an iPhone, really didn't feel there was a fair exchange of value for their email address or huge volumes of data about them to be utilized for advertising for tools that they'd previously been able to use for free.
Roughly 90% of people turned off or did not opt in or consent to the use of their data in any way by those apps for audience targeting or measurement. That, of course, had a very interesting ripple effect very quickly. Within the first couple of weeks, marketers literally saw their measurement simply stop working. These incredibly sophisticated tools that captured data from all these different channels, aggregate it in real time, enable real-time optimization of all of your advertising, they simply broke. Marketers were suddenly faced with the serious challenge of, "Oh, my God, how do we understand how these significant advertising budgets on iOS are performing relative to all the other things we're doing in marketing? Because we can still see our return on ad spend on Android.
We can still see our return on ad spend on search, and social, and all these other channels." And of course, what happens? They pulled back their media budgets on iOS, in many cases, simply canceling budgets altogether. Those budgets moved. Dollars flow where they're happiest, so those dollars flowed to Android. iOS CPMs plummeted by roughly 50%-60%. There were pockets, particularly around video, that retained CPM a little better. Android CPMs grew in a 90-day period, 35%-40%. CTV benefited. Search benefited. This interesting new product that Apple had just begun promoting to the market, called Promotion of Apps on Apple App Store, went from a $1.9 billion business in 2020 to a $19.6 billion business in 2022. So was this pure generosity and protection of consumer privacy by Apple?
Hmm, I think they probably also enjoyed the 10x revenue growth on their own advertising product in less than a 24-month period. Some big takeaways out of this, and one particular lesson learned out of this that I think is super important. Compare and contrast how Facebook reacted to this environment and how TikTok reacted to this environment. Facebook very publicly said, "Nope, we're not going to play this game. We're gonna fight Apple. We'll even threaten to build our own mobile operating system. We're not gonna play the game." TikTok, on the other hand, said, "You know what? We're the new guys in town. People are scared of us because of China. We're gonna rapidly, proactively embrace this thing.
We're gonna try to be the easiest platform in mobile to work with vis-à-vis Apple's ATT and SKAN." A year later, Facebook has to come on to their quarterly earnings call and point out they've lost an estimated $20 billion in projected revenue, and they directly cite failures around ATT and SKAN on iOS as the root cause. TikTok, on the other hand, has a completely different tone in updating their guidance and noting that their already strong revenue growth has been dramatically accelerated, and they directly cited their tight partnership with Apple and how easy they had focused on being to work with around these new privacy changes in the landscape. Why is this so relevant? Because it's about to happen again. Google has announced, several years ago actually, that they'll be implementing nearly identical policies to those implemented by Apple in iOS.
The first step being the requirement for opt-in consent for usage of identifiers and hashed email and other PII-related information in Google Chrome that will ship, projected to be second half of 2024. They've also begun giving initial indications they're going to follow a very similar path for ad IDs and other related data in Android operating system, likely in a 1 years -2 years after Chrome, so 2025, 2026. What's astonishing to me, as a guy with a ton of friends in the ad tech ecosystem, is there are still people who are going to conferences and writing articles in the ad tech press saying, "Nah, it's, it's a joke. Google can't do this. They're never gonna do this. It's gonna dramatically impact Google's own business. How could they shoot themselves in the foot this way?
That's never gonna happen." If any of you use the Chrome browser, I encourage you to enter this URL in right now, and roughly 50% of the users in the world of Chrome right now are in a test trial. So these changes, called Google Privacy Sandbox, have been in test in the real world since November of last year, and they're progressively scaling up those tests so that there are no big surprises to their own systems and platforms when they flip the switch on and make this a mandatory requirement for all Chrome users in 2024. This same Google Privacy Sandbox set of capabilities is also in very early testing on Android OS right now, but to a very small percentage of users. I think it's about 4% or 5% of Android users globally that are in a test right now.
So what do we have to expect when 2024 rolls around and Google flips the switch? I think it's largely going to be a replay of what occurred with Apple. I think you're going to see measurement break, I think you're going to see CPMs plummet, and I think you're gonna see media budgets go through a tremendous period of chaos and turmoil, and I think you're gonna watch media budgets shift around dramatically, very quickly. Who stands to benefit? The platforms and the partners who can provide high granularity, targeting, and measurement, particularly those who can do so in a privacy-safe way. Alternative means of targeting and measurement are going to be required. The second big challenge facing the entire industry is the fact that we have an increasing number of regulatory bodies implementing some forms of regulation and requirement.
GDPR has been in effect for many years. GDPR's core element of concern for the ad tech ecosystem is it essentially says, you're not allowed to use these identifiers or this personally identifiable information without end-user opt-in consent and fully informing the users, so they can appreciate whether they're getting a fair value exchange for the content and services you're offering them, in exchange for all this rich data and opportunity to monetize their eyeballs when they're consuming your content.
And really, another big key takeaway is that I pointed out earlier, most of this data and identity ecosystem hinges off the central linchpin of hashed email, and the use of hashed email is essentially prohibited by GDPR, by the CCPA, and a wide variety of state-level legislation in the US right now, and by the recent regulations passed by the Chinese government, which is the second largest market in the world. So where are we heading? What is going to change? What has to change? What do marketers need? Well, it seems like the only obvious outcome is that this entire ecosystem has to evolve and change in a way where we overcome our addiction to the easy solutions that have been available to us for years and years and years and years and years.
Was there ever any conscious choice to go and build all of this off of hashed email? No, there was no open advertising standards body that said, "That's the ticket. Let's do that." It just happened organically because it was super easy. It was super available. There were no regulations. It was the Wild, Wild West, and it worked, and everybody could print money very happily. Now that the adult supervision is showing up, things are going to change, and they're gonna change pretty quickly. So of course, we're hearing a lot about AI, and you're gonna see massive investments in AI in the ad tech and martech ecosystem.
Contextual targeting and other alternative mechanisms of targeting that aren't PII information dependent, are going to also see tremendous levels of new investment and tremendous levels of new acquisition by marketers, because they're going to need those alternatives to continue conducting themselves the way they have been for years and years and years. So very quickly, just to ensure that we're crystal clear on this, deterministic PII information. I go down to my local auto dealer group. I just bought a new Toyota Tacoma. That auto dealer group registers me in their database, which flows up to Toyota's central database, which flows to their ad agency, Saatchi and Saatchi.
They hit me with follow-up surveys to capture more information about what I personally remembered about my shopping experience, and what advertiser or marketing had the most influence on changing or helping me decide to purchase a Toyota. That's gonna go away. I mean, it's, that's all PII-based information. They're using my email address. They're using my actual purchase history. That's exactly the type of thing that's going to go away. What's going to replace it is, remember how I pointed out that there are trillions and trillions of events happening every day? All of those things are an opportunity to evaluate what's shaping a user experience. If I'm going online, and I'm reading reviews about Toyota Tacoma trucks on a popular review website, does, how much does that correlate to that user eventually purchasing a product or shopping for that product?
And it turns out, data at scale, observed in real time, scored continually, can progressively build into very accurate, probabilistic estimation of behavior, which marketers and, and their partners can use to correlate to who are the best audiences and the best pieces of content to purchase for specific advertising goals and outcomes. It's very different. It will involve a significant amount of investment, but it is absolutely perfectly possible today. Another key element of this is, it's not just one event. If you can see patterns of events, you can begin to add those together and say, "Well, user XYZ 789 was reading reviews of Toyota Tacomas at this website.
The next day, he was looking at some other website about auto insurance." Those things all add up to give us a sense of a very high probability that the next time we see this user, we should be sending them Toyota ads or competitive T-truck ads or auto insurance or what have you. Now, to try and take trillions and trillions of events per day and build high granularity, probabilistic estimation, that sounds like a really complicated thing to do. It turns out we're already doing it, and we've been doing it for quite a long time. Certain spaces like quant finance in Wall Street, in your world, began looking at big data-related projects back in the mid-1990s. I happened to be selling statistics software and got to watch what happened inside of D. E.
Shaw and Morgan Stanley and UBS Warburg, and a number of other great clients who were early investors in applying neural networks and machine learning to various areas of what we would now call programmatic trading. Of course, we're all familiar with ChatGPT, and I think the biggest change for what OpenAI did when they implemented ChatGPT was that historically, the big machine learning projects at Google or Amazon's internal recommendation engine or a wide variety of other projects had a very thin number of layers. So for example, it's not public knowledge, but the perception, especially if you have friends at Google, is that their core search algorithm's neural network is about 8-12 layers deep, and it varies depending on some categories, et cetera. And that it's got hundreds and hundreds of thousands of nodes.
Well, OpenAI, after having beaten chess and then played and beat the best in the world at Go with machines, we're sitting on $1 billion plus in venture capital and simply said: Well, if we really wanna do this pattern matching, and we really wanna do it exceptionally well, why not go to a completely different scale? We've got this killer cheap deal on computing cycles from Microsoft. Let's go spend $100 million and train something. And that's where ChatGPT 3.5 came from. They literally said, "Let's take a neural network, but let's make it thousands of layers deep. Let's have billions and billions and billions of nodes." And it turns out that that enabled ChatGPT to provide all of these unplanned or unexpected behaviors.
ChatGPT, it turns out, can play chess at a grandmaster level, even though it was never tuned or optimized or specifically trained to be a chess engine. Its pattern-matching logic is so good because of the complexity of that probabilistic engine, that it's just, it's out-of-the-box, crushingly good at chess. Now, it can't add mathematics beyond even simple three-digit numbers. If you went in and asked ChatGPT 3.5, "What's the square root of nine?" It would actually jokingly answer 42 a lot of the time. So it's not perfect, I think is a key takeaway, and a lot of these, call them foundational models like ChatGPT or Facebook's Llama and a variety of other, fuzzy animal-related model names, they're really looking at large language-type patterns.
What's going to be required in ad tech are bespoke models that can utilize a lot of the same tools that have been open-sourced and are freely available right now today to develop these exact same types of things to solve ad tech's core challenges. So what do I think wins in the future? I think you're going to see these systems be developed. They're being developed now. And you're gonna need to see management at companies take a proactive stance that we have to embrace, that this is the future, and if we wanna win, if we wanna have an outsized potential for the future, embrace privacy early and often, try to be the easiest to work with, don't fight it. Don't continue to be addicted to what has been easy to do to date.
Embrace the brave new world and the challenges of the future. Combine that with a wide variety of different capabilities, all of which can ultimately cumulatively add up to the best possible performance and outcomes for your marketing clients, because ultimately, that's what dictates who will win and who will lose. Because marketers are entirely focused on performance and return on ad spend metrics. Crush it for the marketers, you will prevail. And long term, I think the next batch of multi-billion dollar companies in the space will emerge from these types of capabilities, and I think you're going to see ad tech and martech continue to blend and overlap and compete more frequently.
I think the companies that figure out how to closely, tightly partner with marketers and enable these types of privacy-safe systems to be employed by companies for their own websites, for their own mobile apps, for their own sharing of content, I think that's where the next big wins will come from. Thank you, and I, I think I kept it under four hours this time.
Thank you, Jeff. You will get an opportunity to ask questions to Jeff as well during the panel discussions. Moving on. I would like to welcome Oliver up on the stage. We are moving from the US perspective, with bankers and digging into the details of data.
Yes.
Oliver, I mean, first, maybe I should say a couple of words about you. I mean, you are the general manager of Pilot Group , one of the leading independent agencies in Germany. You have pioneered the market, and I guess-
Sorry
... you're gonna understand, you know, where I'm coming from when I, when I'm saying that. So, but I'm super interested now to understand, what's your take today?
So first of all, I have to say, I realized three things today. First of all, how are the clients feeling when they don't understand what is presented? So all the acronyms, you know, when we are talking to our clients, we have so many acronyms in our technical world. Thank you for that, first of all. Second of all, I'm the outsider today. Neither am I a banker, neither am I an investor, neither am I the techie guy. I'm a historian, came into the agency world, yeah? And the third one is the most funny one for me: I don't have text today on my slides. I'm the one with the least, I would say, just a few words on the slides.
So yeah, and our take is, from the German perspective, looking at the industry, I mean, Jay, you told it before, the agencies don't want to work with the scattered market. But I have to correct you, we love to talk to the scattered market because this is our business model. We are consultancies for our clients. Yes, please. But what we don't want to do is to work with all of them. That's a little bit too much for us. We can't handle that. But, yeah, and, today I want to present you why we are there, what you hear today about the cookies, the cookie collapse, and so on.
Yeah.
This is our perspective.
Looking forward to. So over to you. Yeah.
Thank you very much. So let us start the journey with the Big Bang. In 1996, they invented the advertising cookie, and it was DoubleClick back then. Thank you for that, DoubleClick. We almost came to the point that we achieved the perfect world. We could talk to the right person, at the right time, with the right message, and with the right frequency. But instead, to stay at this route, as it always happens when the market is not regulated, we ended up by having more than 1,000 cookies per day, per person. What for? I mean, the question is, 1,000 cookies per day per person, that's a lot of data that is in the market right now. And one of the reasons why we are ended up with this is because the market developed itself.
We saw the slide multiple times, but now I'm showing you the development in time. We started 2011 with 150 companies working with cookies. Now, 2023, we are at more than 11,000 companies working and based on cookies. So that means nobody in this room, neither me, even when I'm talking to my clients, a lot of time about this, knows exactly what is with the cookie world, who's using it, how, what for, and where are they store the those data sets. And then something else happened. A telco company in the U.S. came to a good idea to help the market and create the super cookie. So the super cookie was an invention to track everything in terms of who you are, what you like, what you're seeing, what you're reading, and so on and so on.
The only thing they forgot was to ask you for your consent. So 2016, they got fined for that. What they did is they lost the trust of the people in terms of why should we use data still, and why we should use, why we should use the cookie. But if we look at the real world, as in comparison, everything that I said is not new to us. Think of the loyalty cards, yeah? We give our data to somebody, but we get something in return. Think of governmental data. You can't hide. They exactly know who's married, how many kids, where are you living, and so on, and so on. But we are not combining this with advertisement, so it's not disturbing us, and we can't run away of it, from it. So, and then, Jeff, thank you for being a data collector, yeah.
So data collectors on the other side, and I make my, let's state it this way. When I was young and went to Miami to rent a car, I gave them my real address, my real telephone number, and my real email address. What happened? I got so many frequent calls from I don't know where, and I got so many junk emails, but it wasn't an issue or problem for me. I could solve it by myself because it was easy: block them or get rid of the phone number, and so on, and so on. But this is common and frequent in our world right now. So let's summarize quickly the cookie world. We forgot who's using it for what and why, and also, we lost the trust in it.
That's the reason why a bunch of guys in Europe came together and said, "We need to end this. We need to end this." It's not about the business. They don't thought about the business. They clearly thought about how to change it. That's why we developed in Europe, the GDPR back in 2018. Some guys in California said, "Hey, that's cool. We want to have the same." So they came up with the CCPA in 2020, and both of them are really saying that they want to get the active user consent for everything that we are collecting as a data set. If you don't do it like this, you will be fined. This means that they wanted to put us, as a user, into the charge of deciding to whom to give our data sets and what for.
But as history repeats itself, we got now into a phase where we see that everything develops likely like the cookie business right now. We have multiple, multiple ID solutions in the market. We have multiple consent solutions in the market, multiple CDP, and now I'm throwing acronyms at you, yeah, just to be a little bit fair. No, but, we have multiple technologies who are working on behalf of this. And with this said, this means that the complexity, again, for us as a user, is much higher than it has to be. So again, I need to know where did I give my data, to whom, and what for, and so on, and I can't just de-click it and say, "Don't, don't track me anymore." Yeah? And then also, Jeff, thank you for that.
Now, we are changing a cookie, which is anonymized, anonymized to my email address. Who in this room changed his email address in the last 12 months? I would say nobody. Who of you changed the mobile identifier? Jeff, you should now raise your hand because I know you changed it, but recently, nobody. So this is how relevant it is to know where you give your data and how to deselect out of all these data sets. But a lot of people in the business are talking about what to do. How can we merge this? How can we get past of this issue? And they are saying, "Oh, look, our easy-to-use cases are now vanishing." I mean, the cookie world was easy to use for us, easy to target, and so on.
Now, the ID world, it's easy to use it, but it's damn hard to scale it. Yeah, scaling is a different part than just using it and knowing how to use it. But in this world, we need to wake up and realize that we have so many data points right now in the universe, so many data points we can work on, and that's why I love to have this point on the Verve presentation, that data is one point of the invention that you are seeking for. From the agency perspective, I just wanted to show you some of them, which we are using, and then I will present you some numbers on cases that we achieved already.
Today, I brought with me the contextual data targeting, first-party data targeting, like IDs or like cohorts built upon the first-party data, behavioral data, which is probabilistic data, and then also regional data. Regional data, I want to stop there because a lot of people think regional data is vintage data, outdated, you don't use it anymore, and so on. Today, I will prove that it isn't. But let's start with contextual data. Think of Porsche, and think of Porsche doing a performance campaign. So they really want to sell a Porsche. And when you use contextual, you would say you want to be on any site where somebody is reading about luxury cars, and you would be on any site where somebody is reading about Porsche.
On the left side, where you see the green and the red squares, this is how the campaign would target Germany. Everybody who's reading about luxury cars would get an advertisement of Porsche for the performance campaign. Sadly, I have to acknowledge, not everybody can afford a Porsche... but everybody in Germany loves to read about luxury cars. With that said, you need to think of it, that somebody who hasn't the money to buy a Porsche still gets the advertisement. When you think of it, that you can now stack a layer system on the contextual targeting in the terms of taking regional insights that we have, who is living where?
Also in Stockholm, I think you have parts of the city where the people don't have the money to buy a Porsche, and you have a lot of parts of the city where you have people who can afford a Porsche. So rather than taking all of Stockholm and target those who read the article about luxury cars, go there where the people live who can afford the Porsche, and give just them the advertisement. It's more performant, it's better for the budget that you're spending, and this is one way how the layering could work and be much more efficient in terms of how to use data sets. I brought a case that we did with Verve, and just to show you how you can change something that is obvious and do something that nobody did worldwide right now.
And this is, we took, with a partner and Verve, first party data out of the CRM and built in real time. Oops! I hope your phone is still intact, or you have a good insurance. So think of it. Instead of doing DCO, Dynamic Creative Optimization, we changed to Adaptive Creative Optimization. Why do we say adaptive? Because we look in real time in cohorts of the CRM. So we are looking who's buying what, from where, at which time, on which day, and so on, and so on. And the information that we are matching to that, we are taking out of the bidstream of the SSP, for example, from Verve. So we take all the information that we have, match it to the CRM system, and then we are looking the possibility and the likeliness, so the index of the buying possibility.
So instead of serving everybody the same advertisement, in this case, the red one, we are looking who has the highest likeliness to buy the white one. And then adaptively, we correct the advertisement, and we send to this guy or to this girl. I don't know how to do it in English, we have this in Germany, to be always correct in terms of guy and girl, but you got. And we send them out the white one. Why? Because the likeliness that he will buy something is much higher. And this is what we proved with several cases that we already did.
So we proved that we can achieve an overall conversion rate uplift with 4.9 percentage points, and we got. When we look two years back in the data sets of the clients that we did this case for, for one of them, we, on average, get 506 new clients, and this is not a small number for the client. 506 new clients just for a period of two weeks. So the campaign that we tested here for the one was two weeks long, which is a good result. And then a case for you with regional data. What we did with Verve was: how can we enrich regional data with geographical affinities and intents?
So we looked at the room, like the Porsche example, and we wanted to see what else can we get to the table so that we know exactly how to improve regional data and the performance of our campaigns. And this is what we do for 40 clients of ours already, and proving, and those numbers that I'm showing here are just the average numbers that we are achieving. But what we can see is that when you take out of the bidstream , who's reading what in which region, if you take out who is opening which app in which region, and if you take out, this is something that we also do, who's getting which email in which region? You're getting insights that you never got before, and this is where you can really dig out the performance of any campaign and ramp it up.
By saying this, we really achieve better campaigns, not only in the digital world, from CTV to display campaigns, but also on search, also on social, which is really powerful. So the data sets that we are now working with are really cross-channel and can be used cross-channel. Thanks for Verve, we also achieved to do something that will vanish, as Jeff also said today. So the one point that you always need to keep in mind, nothing matters if you can't prove it. Nothing matters if you can't show it to the client. So if you can't show that your attribution or your thing that you build it up is working, who's believing you? I can tell maybe a good story, but it's not showing the numbers that the clients want.
So we achieved attribution, which can show you how well spent is your money and how well it could work for you within detailed regions that we are analyzing and measuring by the ad impressions that you're serving in those regions. With that said, this means that... Think of two rooms. I will give you an example. Think of two rooms. You have room A, where 100 people live, with 53 people from your target group. And think of room B. You have 100 people living there, but just two of them are from your target group. And now you have 100 EUR that you distribute equally to room A and B. EUR 50 in room A has a mathematical possibility of 53 person to hit. In room B, just two person. Where would you like to spend your money?... Yeah, yeah.
But the fun part is, nobody's looking at this detail. Right now, nobody's looking at this detail because they don't have the data sets as we have with Verve. Thank you for that. But also, nobody knows about that. So in terms, and I can tell you just the internal numbers, because I'm really digging deep in those numbers right now. For example, social with Meta and also Google spends more than 30% in rooms that are based like B. But they have so much performance power right now that they can serve you the whole campaign as a success, and you say: "Yeah, wow, that was a great campaign!" But your budget could be more efficient, and this is what we are doing as an agency right now.
We are trying always to be on the side of the client in terms of how can we use technology, how can we do different solutions in the market? And yeah, that's where we come to the end of today, of my presentation, where I have to say, I'm really thankful, thankful for that, that I have such a partner right now on my side, which is nimble, fast, understanding the data that they have, sharing the data, so that we can also be innovative. And thank you for that. Remco also, Sam, all the Verve Group, thank you for having me.
Thank you, Oliver. So coming up is a coffee break. So I'm looking for... Yeah, I think we actually go for a coffee break and be back at 2:00. So we are a little bit ahead of schedule. So we take the coffee break now, 15 minutes or 15 minutes -20 minutes. Be back at 2:00, and at 2:00, we will have the CTO Prasanna up on stage, and we will also have a customer success story from Rami presented. And then we are going for the panel discussions, where you also get opportunity to ask more questions. So see you back at 2:00 P.M. ... Okay, a warm welcome back. I hope everyone got some coffee, and it's a time for the CTO. And let me introduce you to Prasanna, CTO, MGI, and yeah, welcome up. As this is your first-
Yeah.
Capital Market Day
First, yeah.
Maybe, you know, share a little bit about your background before we get-
Yeah
-going.
Yeah. So I'm, I'm, you know, fairly, fairly new to the Verve team. I've been here for about three months now. First time in Stockholm, first Capital Markets Day, and super excited to be here, and super excited to also be the best dressed person in this room. So looking forward to this. Yeah.
Yeah, looking forward to hear more. We will welcome Rami up as well. Hi, Rami, and you, for you as well, this is your first Capital Market Day?
Oh, yes, it is. Yes, it is.
Yeah.
Great to be here. Great to be here. Yeah, my background, I'm, I'm running the Audience and Insights team here at Verve. We are responsible for the data for all of our business units, and it's a short way for me to come here because I live in Åbo.
Yeah.
Not too far.
That's just why, right? Yeah.
Yeah. Yeah, yeah, yeah.
Yeah. Yeah. Okay, we will get going. So, yeah.
Thank you. Good afternoon. Good afternoon, everyone. Super excited to be here and, new to the company, so I'm still, you know, learning a lot of things, but I have, you know, 8-9 years of my, you know, professional career has been in the ad tech space, so I understand the space well, but I'm loving, loving the challenges that, you know, the Verve and MGI team is bringing, to the table. We will talk about what is probably the, you know, the hottest topic right now in, in advertising tech.
Jeff and Jay and Ollie set a very strong context on some of the challenges that, you know, we are going through as an ecosystem, and, you know, we'll recap some of that, but more importantly, we'll also spend some time, talking about all the investments, all the effort, all the thinking that has happened within the MGI and the Verve team to continue to keep advertising relevant, and non-intrusive with the new world of, you know, privacy that's coming in. We know about this. I think Jeff, you know, Jeff explained this, you know, pretty, pretty well. The reality is that digital identity and addressability as we know it, is going away, right?
I think that is, that is the reality that, you know, we all have to accept, you know, being a part of the, being a part of the advertising industry. Users, publishers, regulators, platforms, all of them are increasingly more aware, increasingly more cognizant of personal information. And the digital promise, you know, which, which came in, you know, two decades back when open internet, you know, came into advent about, "Hey, we can actually deliver advertising in an extremely precise manner.
We can have very strong feedback loops to measure what campaign is working and not working." A lot of that reality is being reset, you know, as we speak, with the increasing awareness and consciousness about the world of personally identifying information and ensuring that users are fully opting in, fully consenting into what information about them, you know, gets used, you know, gets used by advertising tech. If Apple and, you know, we spoke about Apple making a sweeping change in, you know, 2021 when they launched, you know, iOS 14.5, when they took away the concept of, you know, IDFA and identifiers and made that fully, fully opt-in.
If Apple is any indication of, you know, the gravity of this problem, then we know that in the next 1 years -2 years, specifically in the world of web on Chrome and in the world of Android as well, we know that 70% of the users potentially will completely fade into the dark. And fading into the dark here means we will not be able to precisely, you know, identify and target, you know, 70% of the user base. What that means is, what we will be left with is, you know, the population, the user population, that, you know, advertising is catered and optimized towards, will organize itself as a pyramid. A pyramid very similar to, you know, what you have on screen right now.
There'll always be a, you know, minority of the population, maybe 15%, maybe 25%, who continues to opt in, who continues to be okay with, you know, their data being used for personalized advertising. And that population previously was close to 100%, right? So what was originally the entire pyramid is now shrinking and concentrating itself to the top, and it'll probably be a 15%-20% of the market share, you know, going forward. What is coming into existence and, you know, probably not relevant in the world until yesterday, is the world of inferred identifiers and the world of completely anonymous users, right?
So a lot of the solutions that, you know, companies like ourselves are investing and building, historically and certainly in the future going forward, is how do we continue to keep adtech addressable for the middle and the bottom of the pyramid, outside of augmenting what we already have, you know, for the top of the pyramid? The Verve Group has a very, you know, interesting vantage point. We are one of the few non-walled gardens, you know, companies that operates across all screens. If you look at what are the predominant screens that users today spend their time on, there's the mobile phone, there's, you know, the desktop or the web, and there is, of course, Connected TV, which is becoming more and more important.
Verve Group is one of those companies that kind of, you know, exists across, across all the screens that users spend any of their, any of their online time on. In addition to that, we also have the advantage of not just playing across screens, but also playing across the entire supply chain. You know, we have seen in some of the previous slides about how convoluted and how complicated the ad tech supply chain can get. As Verve Group, we have, we have multiple shots at the, you know, at the same pie to the chain, and that allows us to significantly provide verticalized solutions that are not possible if you're specializing on one side of the market.
Because a lot of the kind of companies that exist in the ad tech have chosen to specialize either on the advertiser side, on the publisher side. And as a strategy, the fact that we have presence across the supply chain puts us in a very uniquely advantageous situation. We will continue to you know address the top of the pyramid. The top of the pyramid is you know this is where the entire world was yesterday. It's obviously shrinking, but we have advanced solutions to kind of target behaviorally, where identifiers continue to exist. And while that might diminish, we will continue to invest and augment on you know the precision of those solutions.
But what is really probably gonna set us apart in the coming few years is, you know, the first three boxes that you see there, right? I think we're one of the few companies who've seen this coming, and I'm... that while, you know, while I'm three months old in the system, I take immense pride when I look back and see how, how the Verve Group has actually perceived some of the changes that are happening in the ecosystem and prepared ahead of the competition in providing solutions for the world of anonymous, for the world of ID-less. Moments.AI, that, you know, Rami will talk about, is our ID-less solution for how do we keep advertising relevant in the world of web.
ATOM, which is an SDK-based solution, is our attempt to ensure that we continue to keep the in-app audiences addressable even in the absence of identifiers going away in Apple and soon going away, you know, in Android as well. And Household Targeting is, you know, how we use the data assets that we have and target CTV users in their home with what we think is the most, most, most relevant advertising. The interesting part of, you know, some of all of these, you know, products that we have is, The Verve Group has grown inorganically, significantly. A lot of additions have been added, you know, to the family over the last few years.
The addressability suite that we have and that we will continue to build on provides you know, off-the-shelf services for every future product that we build, every future company that we acquire. We will probably have one of the you know, one of the biggest data lakes out there with all the assets that we have that will allow us to keep advertising relevant, not just for the products and companies that we have acquired so far, but potentially the investments going forward. What we want at the end of the day is we want the benefit of the collective union of everything that we have, right? We don't want to be in a situation where, you know, we have Smaato, and Smaato uses Smaato data.
We want to be in a situation where Smaato, through the data lake, gets, gets value from all the data that we have, all the assets that, you know, we have as a company. You know, we, we saw this, you know, we saw a version of this slide, you know, earlier, earlier in the presentation. You know, AI is great, but AI works, when you have good data, when you have high-quality data, when you have unique data that we believe, you know, a lot of our competitors don't have. And at, at the heart of some of the, you know, research and development efforts that, you know, my team will be looking at over the next 6 months -12 months, is how do we get an interesting, or an advanced understanding of privacy-enhancing technologies, right?
It's called PETs. There's a lot of development in the world of PETs in the last, you know, 12 months. How does it apply to advertising, clean rooms, secure multi-party computation, homomorphic encryption? How do we use all of that to ensure that we continue to keep our partners' data safe, but at the same time allow them to deliver highly effective ROAS and highly effective campaigns? Edge computing is a big part of our strategy. The best way to keep user data safe is to actually not transmit it outside of the device that the user already controls, right? So, we are investing significantly into ensuring that we have on-device solutions. We already have a great SDK, you know, distribution and presence across, you know-...
20,000 apps that exist in the app ecosystem. How do we augment that with solutions that are able to draw inferences about the user on the device without having to transmit, you know, PII information outside? And then how do we use all of that back into serving what we think is the best, most relevant, you know, advertiser, most relevant campaign for the user? AI is, you know, it's a key part of our strategy. You know, we have been massively amping up on our investments in this field of AI. You know, we are working with some of the cloud partners that, you know, we already have deep relationships with Amazon, Google, using some of their machine learning platform and AI capabilities.
And again, how do we use all the assets that we have, the assets that we have through our owned and operated systems, the assets that we have, you know, gathered over a period of time through acquisitions, the first-party data that we get access to both on the supply and demand side, because remember, we operate an SSP and we operate a DSP. And then how do we allow AI to eventually make, you know, the best decisions for the user? It's a key, key part of, you know, what, what we're trying to do. I will be handing it over to, you know, Rami, for the second part of the presentations. There's one thing that I wanted to close with.
I think there's in general a lot of skepticism in the world of, you know, advertising about, you know, how data is going away, how new age privacy will potentially, you know, regress the world of digital advertising. At the Verve Group, we are actually very excited about it, right? And we've been excited about it because we have seen this coming. A lot of the acquisitions that we have done, a lot of the investments that we have done in the last two years is catered to the world, that will be born going forward. So the world is skeptical, but we are super excited about, you know, the portfolio of solutions that we have, that will allow us to keep, you know, advertising relevant for the publisher, for the user, and the advertiser.
Thank you, and over to you, Rami.
Thank you, Prasanna. Thank you so much. Very good. Back in 7 years -8 years ago, when I started my Beemray data company, we already saw what's gonna happen with the advertising technology. We saw how the advertising industry was exploiting personal data, and we saw that it couldn't continue. Then GDPR came along, CCPA came along, and the other acronyms that we saw already on the slides. Now, what my team, engineering and data scientist team, what we wanted to do is to build a natural language processing platform that understands content so well that we can actually address and find the right people just by looking at the content and the context of where these people are. And that's what we started building back in the day.
Now, maybe two years ago, Sameer gave me a call and say that, "Well, can you guys join us?" And, the answer was fairly obvious. Verve provided us the global presence and also the massive amount of, of data. So today, there is no longer Beamray, but the platform is now called Moments.AI. So what is it then today? We've been fine-tuning the platform ever since, improving the models all the time. So today, we can say that we fairly intelligently classify content to understand, the sentiments of it, the meanings of it, even the images on it. And we do it at speed.
So during our 20-minute presentation, more or less, we have classified about approximately 2-3 million new pages globally released, to a state where a marketer can now address and use these new pages as their new inventory. So what are then the results? The results are that we can provide our clients hundreds and hundreds of segments, like, fairly basic ones like sports or news or politics, whatever that they can use for their targeting purposes, but down to very, very granular segments as well. Custom segments included. We do visual intent segments with Getty Images. I'll go through that a bit later, and some brand safety categories. All these available now, and soon we can see how our good friends at Match2One are using this.
The Getty partnership is something that we started maybe early this year, and the idea there is that we initially looked into how we can analyze the images ourselves or videos on the internet ourselves, and we can do that in fairly good manner, I would say. But Getty Images collaboration allows us to go way deeper. The quality and accuracy of our classification is the key, which is why we then chose to go to the market with Getty Images. This is now also available globally. Contextual classification of the content is not just about targeting, it's also about getting insights. So now that you know that we are classifying pages every single second, new pages, millions of requests per second, we know exactly what people are reading in Sweden or Norway or Germany or wherever.
That gives our clients good understanding on what types of content they should be targeting, some emerging trends and so forth. So it's really important for the planning purposes of those campaigns. But it's also important for the measurement part. Now, this is super interesting, at least for me. Now that cookies are dying, so what can a marketer do? They can use our contextual measurement and see what kinds of content actually resonate with their own cookie-based audiences. So they can use this measurement tool to convert the old cookie-based audiences into something that is future-proof. We can also obviously do other things like verify the brand safety vendors and whatsoever with this measurement tool. Then what about results?
Let's have a look at the results, first from the technology perspective. So what we have found out during these years is that roughly 70% of all pages that people are reading today have been released during the past 24 hours. So in other words, the pages are very, very new. So now, if a marketer thinks about their contextual provider, data provider, they need to consider this, because if the contextual data provider cannot perform in lightning speed, they will miss 70% of that supply, available inventory. So in other words, speed is of essence. Now, this was kind of proved when TPA Labs, a UK-based research company, they did this research back in June. So they compared Moments.AI plus two industry-leading vendors, contextual vendors.
So the idea there was that they set up a campaign that was only for 24 hours and a sports content. And I'm sure you can read what the results were. 96% of the URLs that we found were published during the last 24 hours, compared to our dear vendors, that were down to 5%, more or less. And then, of course, we were also accurate in terms of the content itself, you know. Some of those published were released in the same day, yeah, but they had nothing to do with sports. So what can I say? I'm extremely proud of our technology and obviously my team. And then, let's have a look at results from integration perspective.
We've been talking about vertical integration today quite a bit, and there's an excellent DSP founded here in Stockholm. Guys are sitting there, Mikael there. So Mikael prepared us a video of how Moments.AI is now integrated with their DSP. So I would say that Match2One DSP is now a contextual DSP, correct? And also what this video will prove out is that cookieless programmatic campaigns is not-
Hi, my name is Mikael Kreuger. I'm the co-founder and CEO at Match2One, that joined Verve Group back in September 2021. I'm here today to talk about the integration that we made with Visual Intent and Moments.AI, making it accessible and available through Match2One's UI. Since we made integration, which is a couple of months ago, we now have eight out of our 10 largest clients actively using the contextual segments as we made available through the platform, which is a very positive uptake. Now, I'm gonna take you through a demo of how to set up a contextual campaign through Match2One, and how to follow the reporting and the results that it has delivered. So the first thing we wanna do is to set up a new campaign. We then contextual, we then select the ad format, so whether it's display or native advertising.
We then select Contextual as the model, and we choose the contextual categories, whether those are visual intent or moment-based content categories, such as education. Then we select countries. So for this one, we'll have Sweden. We will use all the device types as targeting settings, mobile, tablet, and desktop. $500 in a daily budget. We're gonna drive it towards that website visitors as a CPC goal, with maximum delivery as an automated goal value. SEK, as we are in Sweden, we do this so we can follow the revenue that is produced through this campaign, and then we launch the campaign.
When the campaign is launched, we'll get some very interesting analytical data that we can use to further optimize the campaign, or at least learn what the audience is responding to the contextual segments we are now targeting with those ads in this campaign. The first thing we see here is how the campaign is performing over time. We get a glimpse on how many impressions, clicks, conversions, all the way down to what revenue has been produced. To get some more insights into the contextual segment, specifically, we go to the audiences, where we can see a share of what categories, and what interests, and what moments has led to the performance in this campaign, including the CPM, CPC, CTR, as examples.
We can also follow the specific conversions all the way from what ad was shown on what website, where was the user based, and what order value did that lead to. So I hope that gives you a sense of how this looks and how this works in Match2One's platform. Thank you very much.
Very good. Thanks, Mikael. What a beautiful interface! I still like it so much. So let's have a look now from a case study perspective. What happened when Match2One integrated Moments.AI? There's a fairly famous OTT company called Viaplay, very famous here in Scandinavia, at least. What they wanted to do is to drive new subscriptions in Norway with contextual targeting. The target was, of course, to target new content that is brand safe, where consumer and identify consumers' interest and intent to get those new subscriptions. And I was astonished to see the results. Remember how I said to you that we wanted to build a platform that matches the performance of cookies?
So based on this campaign, every single dollar that Viaplay invested in the campaign, they got $8 back in terms of new subscriptions. Excellent results. Excellent results. Needless to say, that they are collaborating with Match2One still. Good. So what next? It was great to see Jeff's presentation about data, and I think we have quite a few responses, answers, to your challenges that you mentioned. This is like, for my team, this is like being a kid before Christmas. We started a project early this year, and started looking into the entire dataset that we have in the company, not just from our exchanges, but maybe even more importantly, our owned-and-operated data. So all of this data now allows us to come up with answers that Jeff raised.
What's gonna happen in the world now that cookies and identifiers will be gone? Now, having this owned-and-operated data will position us in a place where we always have access to a certain level of IDs. So and that helps us then to fine-tune our models, fine-tune our accuracy and quality, even in the contextual space. And also, what's very, very important, because it is our data, we have made sure that we have legal consents to all of it. So the regulations, we support each of them, and we can access the data, use it on behalf of our marketers, and performance, obviously. So interestingly, so what we then do, we have all these different types of data: contextual data that, where I'm coming from, we have the ID-based data that is coming from our Gamigo data, mostly.
We have a lot of regional data or census data. Oliver was explaining this really well, and I'm really happy about our collaboration in that sector. We have on-device data that relates to ATOM, the new product. So all of these are... Think of this as multidimensional wheel in a way, where each of these support each other, and that's why I believe that we have ways to address and engage with consumers at any place. And it definitely unveils affinities, so we can find certain types of personalities from all this massive dataset: travel, high-net-worth income people, college graduates, and so forth, that the marketers can then engage with. Now, because we are sort of hiding the complexities of our data, and we enable clients to target and find their consumers in an omni-channel environment.
So let's say we want to target a business decision-maker. We find this person when he's browsing an iPad at home. We find this person when he's on a train to the office, in the office when he's using the desktop. Maybe in the afternoon, he passes by a digital out-of-home screen, we find him again, and later in the evening, in the CTV environment. All of this, with or without IDs, but we have the capabilities to find such a person all the time. To summarize, what data products we have today, the Moments.AI, including the visual intent for contextual targeting, ATOM, which is kind of Moments.AI, the contextual thing, but in a behavioral manner, but anonymous still, because we can follow the patterns of the user, behavioral patterns.
We have the ID-based audiences coming mostly from the Gamigo datasets. Then we obviously do a lot of insights to improve our own optimization processes and to give those insights to our clients. How the Verve Group does makes money, the revenues with this, basically three things: curated packages, exclusive brand-safe private market packages that we can sell as deals to DSPs, marketers, agencies out there. We do managed services, which is basically in-house, expert-led services, where our experts are managing the campaigns on behalf of the client. Then thirdly, we're also adding value to our SSP signals. When we are enriching those bid requests with our datasets so that we provide more value coming from our SSPs, and this obviously in an omni-channel environment. That's about it. Thank you all for following this. Thank you.
Thank you, Rami. Okay, so next step then, panel discussion. So I'm gonna start with inviting the experts up on the scene again. So a warm welcome, Jay, Jeff, Oliver. And then we're also gonna have the management team up. So Sameer, please welcome up, Jens, Prasanna, Remco. And then we are waiting for Danesh to get his microphone on, because Danesh is going to do the moderation of this panel discussion. Yep, I think we get all of you up, right? Yeah.
... No, we're missing Danesh. Danesh, the moderator. Yeah, he's coming within shortly. Yeah. Okay, keep it like that. So ready to go, Danesh?
Yes.
Perfect. Then I hand over to you.
Thank you, Jenny, and welcome to our panel discussion. So this panel discussion will focus on the topics that were presented by our external experts. And to quickly recap, Jay spoke about vertical integration and consolidation, Jeff spoke about identity and AI, and Oliver spoke about database targeting. And as promised, we will also open up for questions both from the audience here in the room, but also from the web. And let's jump into the panel discussion, starting on the topic of consolidation and vertical integration, and then we will do a Q&A on each topic. So first question to Jay: consolidation often leads to a smaller number of major players dominating the market. What effect does this have on competition, innovation, and diversity within the ad tech industry?
I didn't hear the very first part of that. Pardon me.
Consolidation often leads to a smaller number of major players dominating the market.
Right. Right. I guess I would say that, the ad tech business is never going to be, ten players. There's always gonna be... There's too much venture capital out there supporting way too many companies that frankly don't belong to be in business, because they're great technologies, but they're not real businesses. I don't think that's going to change. It hasn't changed in ten years. Now, what's the good news on the flip, on the flip side of too many point solutions and too much funding for these small companies? They're really, they're really innovative. They're creating amazing new technologies that larger companies like MGI and others will acquire, because they're not real businesses themselves, but they definitely will fit in to the platform at some point.
So I think that, I think overall you're gonna basically see a continuation of funding for small startup companies on the technology side, and you're gonna continue to see larger companies acquire the technology, 'cause it's really, it's the old adage: buy or build. Take the time to build, or you can just buy something very quickly.
Feel free to add to this, if anyone wants to add something.
Sure. Look, I think just to extend what Jay is saying, we as an organization already have been actually doing that. Our early on conscious call to actually make Moments.AI as a part of our family was like we knew that this is gonna happen, so we got to plan it before ahead of time, rather than actually panicking at the last minute. Another similar decision was when we decided to actually make Dataseat as a part of our organization. These are the state-of-the-art solutions which have been built for the future, and vertical integration and consolidation, they go hand in hand. And we still very firmly believe in how can we add more value towards what can be done better for the CTV environment or the video, the recognition capabilities of images, et cetera.
So I second what Jay says, and we have been actually with those principles.
Yeah, so, a follow-up to that, speaking about dominating the market, so walled gardens, you spoke about those, Jay. They account for roughly one-third of the time spent on internet, but they attract two-thirds of the ad dollars spent globally. Why is that, and do you think that will change?
Why is that? Well, I think they've had an advantage because they've had First-party data for a long time, so they've been able to target their customers, their existing customers, in a way that the open web wasn't able to. That's changing. You know, it's not gonna change, I mean, the reality of it is those are from very large, dominant players who've got a lot of funding behind them. And that's not gonna necessarily change right away, but the open web is going to continue to grow in size. It's gonna continue to grow in complexity, but in efficiency, that's going to allow people like Ollie and others in the agency side to have a choice beyond just the Walled gardens.
Can I add a little to that?
Sure.
I think it's worth noting that confidence and trust in these companies is at an all-time low. I think nearly everyone I've interacted with at the C-level in ad tech has read the U.S. Department of Justice filing against Google. And they've seen... I mean, literally when I read that document, and I read it twice, cover to cover, the day it was released, and I had at least a half dozen light bulb moments where I suddenly understood why Google had been doing things in such an obtuse, or half-assed, or obscure, opaque way. And you have to realize that there are dozens of very well-documented accusations in those filings where Google is playing variable take rate games, and they are not seeking to optimize yield for their publisher partners. They are not seeking to optimize return on ad spend for their advertiser partners.
And marketers need better alternatives. They need trustworthy alternatives, and Facebook, of course, has quite a history. I think we've all watched from afar with astonishment at the rapid growth and emergence of Amazon Media Group as a titan that literally is defining the retail media sector, which is arguably the second hottest element of the advertising space behind CTV. However, if you manufacture products, if you manufacture diapers or toothpaste or T-shirts or shoes, increasingly, I don't think most people realize Amazon owns and operates 3,000+ different brands that are vending product across amazon.com today.
Amazon diapers, Amazon toothpaste, Amazon clothing, Amazon et cetera, and they have huge product groups inside of Amazon that are continually harvesting all of Amazon's data about what people are searching for, what they're shopping for, what they're purchasing, at what price points, and that's all market research for them to power their next generation of products. And if you sit down with people at P&G, if you sit down with people at Nike, they have exceptionally deep, strong reservations about continuing to send their data and their media budgets to those types of partners.
But it's only the last few years-
Right.
- because exactly this trust has built the big giants, part of it at least. Because when we bought, for example, Platform161, Verve DSP, it's called now, we had... The problem was they didn't get customers because customers had something: "Yeah, if I'm a marketing manager, I'm going to Google, I'm going to Facebook, because if something is maybe not efficient, I choose the big one, the big brand." Now you see really that there's more and more awareness coming up, that it's more about quality, and that especially the large ones are sometimes doing other things with your data than you would like them to do-
Yeah
... or playing with your margins and things.
I would do. I would add two points to that, and, first of all, the view from a client perspective and not from a techie perspective. I mean, we are all, I would say, quite good at what we are doing and exactly knowing what is happening. The Bernanke case you are referring to, I don't know if somebody in this room knows about it, and the clients, I can tell you, yeah? To be honest here in this, round. And, and the main reason is the complexity of what we are dealing with right now is too high, and the management worldwide, globally, if you are running a global company, do you want really to have a DSP in Germany?
And, I mean, take Europe, how many countries we have? We are not America, yeah? We are here, Europe, with a lot of borders and so on, different cultures. This is the main reason we are fighting against, but it's... You know how it is. It is a difficult fight to take, and I think that, Jay, what you stated before is the right answer to it. Innovation will never come from those big silos because they are seeking just to earn more money. So we will have a lot of new and small, nimble companies that will emerge into the market. And then the biggest point is, we need companies like MGI to give them more scalability, because they have maybe the, the right solution, but if you don't scale the solution-
Yeah
... what do we do?
A question, Rem, directed to you, Remco. Where does MGI currently stand in terms of vertical integration?
We are—I hope that became clear today, we are vertically integrated, and we did it via M&A because that was the fastest way to build it. If you do this organically, greenfield, it's going much lower. If we then look, yes, we are vertically integrated, but the majority of our revenue still is coming from the supply side. And yes, behind every supply dollar, there's also a demand dollar. But we still need to do more towards the demand side, and that's what you have seen before. Match2One, for example, is really leveraging the strength that we have on the supply side, the strength that we have on the data side, and bringing that closer to advertisers. And that's where we have still a bit of work to do.
But we are capable to run a full vertical stack with all the advantages, less margin taken, less cost overall in the chain, more transparency, and we can do that multi-channel.
Just quickly on this question, because we want to take some questions from the audience as well. But if the open web consolidates and vertically integrates, will it not just end up like the walled gardens in the end? Well, how will the open web do it differently?
Some of the open web players might become closed gardens. We have seen some trials to that, in. But-- And then also, Facebook started pretty open, actually, to be say, in the beginning. Apple started open, and they are almost more closing their gardens. There will be some open web players that might also want to do that. And, yeah, we always say we're getting stronger. We're also a bit of a walled garden. We have a lot of stuff that we can do in-house, but we say we want to have the bridges, we want to have the transparency.
So I think it depends a lot on where you position the company, and we are strong believer in making media better, as we also say, and making media better is meaning you make it more transparent, you make it more efficient, and in the end, you help advertiser and publisher to both become better.
Anyone want to quickly chime in on that?
I can add to this. Look, Danesh, back in the U.S., there is a evolution of some new walled gardens that we actually see, and I'll give an example. Disney, right? It's about so much data that they have, that Walmart is another example that we closely see, that's so close to Amazon. Now-
... end of the day, it's about a company's belief and principles. What kind of solution do you really want to bring out? Is it only about making money, or it's, you know, as Remco mentioned, let's make media better? What kind of solutions you really want to bring out to your customers in this complicated ecosystem? So yes, there will be some, but again, ultimately, it's the trust factor, it's the alternative solutions, it's the complexities of the ecosystem, and companies like us are very fairly positioned, because we are actually really trying to solve these complicated problems in the sense that we are still an open internet supportive company. So I do expect some new names for sure.
Yeah.
I would also add something to that, but Jay, please.
I said quickly, yeah.
Quickly, yeah, quick. We have time.
Yeah, we've got, we got 49 11.
Yeah.
That's not going anywhere. The. So I would argue that it's actually gonna be--would be very hard for somebody to create a walled garden because you need scale. And even if you talk about somebody like a Disney or something like that, now you're talking about something that's a cul--Remember, Disney's theme parks, film, it's all these variety of entertainment businesses. Oh, and they, by the way, they have an online business, right? Amazon got big because Amazon started selling online, right? They're selling books, so culturally. I mean, all you need to do is really look to what's happened with the telcos, right, over the years. I mean, all the telcos, they tried to buy... They said, "Wow! We've got, you know, we've got AT&T, I've got, like, 45 million, you know, mobile IDs, mobile customers.
Now, I can sell advertising to them." Well, it's a very different culture, the online advertising world, than it is from a telco. You've seen it with, you've seen, you know, Singtel buy Amobee, then they divested it. You've seen AT&T buy Xandr, they got rid of it. AOL or, Verizon bought AOL, then Yahoo! They got rid of it. They just. And it's not necessarily that the technology didn't work, the cultures didn't work.
Yeah.
So you really need to factor that in, is my only comment as it relates to the new walled garden, so to speak.
Yeah, I think you're gonna see an increased number of CTV platforms continue to pursue walled garden strategies. Roku, proprietary Roku ID, don't share any IDs outside of their walled garden. Disney's following the same track. Netflix is following the same track. NBC Universal is following the same track. It's going to be prevalent, and it's largely because, I mean, look, 30 years ago, they were the kings of the castle. Like, they dominated advertising spend, and those were advertisers going directly to those companies. And I think their walled garden strategies are largely an attempt to return to that. They want to eliminate the intermediaries as much as they possibly can and harvest the optimal total number of dollars and margin take rate themselves, plain and simple. Is that possible in mobile or open web? I think it's exceptionally more challenging.
For scale, for the breadth of traffic, et cetera, it seems far less likely.
Yeah. Let's open up the questions from the audience.
I wanted to-
Oh, sorry, yeah.
I just wanted to add to all of this discussion. Scale is one point of the side, but the other side is, what's the worthiness of this silo that is emerging? Because look at Netflix right now, they tried it, but they have no scale. So what is the worthiness of Netflix as a silo right now? You know. And the same is with Amazon. You have something in your hand. Whenever I'm doing a campaign, a performance campaign on Amazon, I know exactly how much I will sell of my items. And then the cultural thing comes also, kicks in, because Walmart, for example, yes, they have a lot of data, but do they really know how online is working? I mean, now I would need to talk to your daughter, but, I mean-
Me too.
Yes. You know what I mean. You always need to have in mind that silos aren't that easy to use. And in TV, in Germany especially, we have it. We have two companies that built a common force. They called the d-force. It was really funny for them to be called like that. And they created ATV as a silo. But do you know how they are scaling? They are scaling limited.
Mm.
They never passed a critical mass in terms of it's skyrocketing. So do they have fun with the silo? I don't know. Is Google still having the better silo? Yes.
Yeah.
Not to forget, we're talking about a market that has $hundreds of billions of size-
Yes
... which will allow several business models and which will allow several parties. But what we see, you need a certain size to innovate because this market is changing, and you need to adapt to it. You need to go into contextual, you need to do things with your data, you need to verticalize. So you need a certain minimum size, but then with a good business model and good customer acquisition strategies, you can really gain market shares in this market.
Okay. So now let's open up to questions from the audience and, on the topic of vertical integration and consolidation, and also from the web.
Yeah. So I'm gonna take one from the web.
Yeah.
Maybe a little bit more helicopter perspective. It's about the value of data, and this one is for you, Remco. So question coming in is: can it be the case that the value of your data and your access to various datasets is higher than your enterprise value for a strategic acquirer in the sector? Would you be open to that idea, or would you prefer to remain independent?
I've learned one thing as a CEO and also a board member, you should not talk about the valuation of your company. So that's, let's say, giving already part of the answer. I think that the data part that we're building, the data that we have in the company with the gaming part, but also all the data that we get from our exchanges, our SDK base, is a huge value. But as we had a lot of, how do you say, raw materials, you need to get the value out of it, and that's what we are working at. So just having data, there's a lot of parties that have data. It's really what you do with it, and that is what, what builds the value, and I hope that today we are making clear that that's what we are working on.
We have a saying for that at Pilot, and we are saying: "Having data without inventory is nothing worth. It's like having the best fuel on the planet, but you don't have the rocket," so.
Any questions from the room? Don't be shy.
Thank you. It's Fiona Orford-Williams from Madison. We're talking about Google and Meta having these existential threats going on. They're not gonna sit back and let people take their lunch away from them. How do you see them reacting to the changing environment?
Remco, why don't you? I mean...
One of the things they do is buying a lot of companies, so they're really buying technologies. If you see the acquisition, the number of acquisitions that in Google and also in Amazon and all those guys are doing, that's huge. On the other hand, they are also... The bigger you get, the slower you get, and they have not been proven that often that they're good in innovation from within. So again, like before, I think the market is so big, yes, they will, of course, fight for it. But you see that some of the things that they are doing is yeah, increasing their stake of the cake, margins. It's fully intransparent, and it's really super interesting reading the DOJ files, what they do to increase margins.
What we see, I think it emphasizes what was said before, a lot of advertisers are getting a bit fed up with working with black boxes, working with companies that use the data for other things. So there is an opportunity.
I'll have another one since no one's taking the mic away. In terms of, sorry, the Google cookie situation, do you think everybody, all the advertisers have really taken on board what happened last time and have actually got their act together for next year?
Jeff?
No.
Need a guy.
I'm astonished how many ad tech companies don't have their act together in preparation for what's coming. If you look at some of the market metrics that have come out in the past 12 months, a lot of the tactics that are being followed are trying to apply band-aids to the current operating model and extend the current way things have operated as long as possible. You know, frankly, disclosure, as a guy who was a VC and sponsored an investment in a company called ID5, that's one of those best band-aids, I love it. But does it look like a genuine, thoughtful, deep response that addresses the fundamental challenges of being entirely reliant upon cookies and hashed emails and everything that we've discussed today? No, it doesn't.
I think you're going to look at a majority of the ad tech ecosystem will try to apply band-aids and continue operating as they have thus far. And personally, I believe they face a significant amount of risk for choosing that type of strategy.
I can add a comment here. Look, we hired a very large marketing agents marketing company back in the US to help us position ourselves in front of the whole coast and the large agencies. Remco and myself had the opportunity to actually meet a lot of executives of the agencies. Apart from some smarter agencies, a lot of agencies, they are not prepared. They have been waiting. They've been waiting because they rely eventually on a lot of solutions. There will be a panic situation, but again, companies like us, who are nimble enough and actually are slightly ahead of the game in terms of our planning and in our thought process, and there are more companies, it's not just MGI or Verve Group is doing this.
So the agencies who represents these big budgets, they are not completely ready. I'm sure Ollie can add from the perspective of not just partnering with us, but how do they feel? Because they have to represent these customers who are unaware of how January 1, 2025 is gonna look for them.
I think just to add, I think it's unclear what the new normal is. I think there are a lot of people, you know, there's a lot of guesswork happening right now on what would the world without cookies look like. You know, we have our take, and we have our expertise over the last few years, and we believe we have a compelling solution, but then there are other parties out there who are also thinking on those lines and trying to outguess on what that new normal is. So I actually think it is the onus is largely on, you know, technology platforms, you know, like MGI, to define what that new world order is gonna be.
I think agencies and advertisers, you know, will catch up, but the onus and the heavy lifting has to be driven by, you know, some of the tech platforms out there.
And also from looking at M&A cases that we have to table, I'm shocked how badly prepared a lot of the companies in the market are. It's a bit like, yeah, what we saw when Apple introduced IDFA, people couldn't use identifiers anymore as much on Apple, so they shifted their marketing budget to Google... but if Google also has no identifier more, anymore, it's getting a bit narrow. Yeah, we have seen it with GDPR introduction as well. A lot of people closed their eyes and thought it will not come, or it will be delayed, and it will be delayed, but at a certain date, it's there, and that's the time when a company that is prepared can really gain market shares.
Thank you for the question, Fiona. That segues nicely into, yeah, starting the discussion on identity and AI, which we already did now. But, a question to Jeff: "So when do you think we will have completely reached a new world where there are no non-consent-based IDs, and there are... Yeah, are there other changes on the horizon which will have similar impact as ATT and IDFA?
Well, if you take Google at their word that they intend to turn off cookies in Chrome in 2024 and Android ad IDs a year or two to follow, which again, many companies in ad tech do not believe is going to be the case, which again, astonishes me. I do believe, and I've heard from senior people at Google, that it is a concrete fact. I think you're looking at the last remaining channel, which at present has widely available free identifiers, is CTV. However, as I noted earlier, a lot of the very large players in that space are pursuing a walled garden strategy, and they are moving from transparent flow of identifiers and related information to opaque.
They're replacing open IDs with a proprietary ID in an effort to consolidate demand back to them directly and eliminate intermediaries from the ecosystem. I think collectively, in terms of the media channels with largest allocation of spend, I think, yes, but in the next 3 years -5 years, I think we'll be looking at a landscape that is predominantly either opt-in consent managed or dictated by the underlying platforms to be operating on a proprietary identifier with, with a different, although related, set of challenges.
And a follow-up to that: So, what effects will this have on CPMs, CPMs long term? So in your presentation, you mentioned that CPMs decreased on iOS following IDFA, and but it increased on Android, right? Just because they were-
Yeah
... they gained relative strength. But when we move to a world where there are no, basically, identifiers, what happens to CPMs then? Does it level out when the competition is equal?
I think 2024, 2025 are going to be years of absolute chaos and turmoil. I think marketers will face a tremendous set of challenges trying to figure out and map out for themselves where to reallocate their budgets to continue to harvest optimal performance out of advertising supply. I think it will force marketers and agencies to test and experiment very, very rapidly, a lot broader set of alternatives than they may have tested previously. At the same time, I think they're going to default to safety as well. I mean, simply put, dollars go where they're happy and comfortable. So is Google going to grow search revenue as a result of turning off cookies in Chrome? Yes, full stop.
Is Google going to grow their own Android app store, ad promotional product revenue, as a result of turning off cookies and turning off Android ad IDs? Yes, full stop. I think companies that are positioning themselves to enable marketers to very rapidly test across a wide variety of different channels, to most accurately measure or estimate where the new pockets of opportunity for performance and attractive ROAS exist, will become dramatically more attractive to marketers and agencies than they are today.
I would like to add something to that. Do you really think that, publishers and so are really ignoring that point, that they can't track? Or isn't it maybe also so that they are hoping for it, because where no measurement is, they can start to do business as they did before? I mean-
Yeah
... TV was, and is still, we are talking about digital today here. The biggest media spends are in a channel that we can't track, and we have just some implications of technical solutions to track it. So we always need to keep in mind that none of us hopes for the worst-case scenario. Most people are hoping for the best-case scenario, and that is going back, yeah, back to the future. It's like, in the film.
Yeah. I think, as Jeff mentioned, and as we have learned from the history of switching of dollars to different operating systems, we'll have to learn from the media planners. Maybe the dollars are shifting too fast enough to DOOH. Scale would absolutely matter by then, so we'll continue to make sure that we acquire more and more scale on this, on our platforms to help the customers. But, it is sort of in a slightly ambiguous situation for some period of time. Yeah. But who will come and solve it? I think all of us have to together solve for it because we represent both the advertisers and the publishers. We can't just sit, putting our hand on hand, let's wait and watch, right? This industry together has to solve for it.
If you go high level, people need to advertise their products.
Right.
People have their budgets, or let's say companies have their budgets, and the money needs to go somewhere... and those that offer the better solutions and also the better measuring of the, of the results of those solutions will take the budgets.
Mm-hmm.
Part of it will be world cards, but there is a big world outside of that.
Yes.
That's what gives a lot of opportunities.
That's an important point, because I think CPMs, you know, looked at standalone, or CPMs regressing, if you look at it as a standalone metric, you might, you might have concerns. But as Remco said, the budgets I still think will continue to grow, right? I think today if, if an advertiser has enough data about a user to say, "I'm willing to pay $10 because I know you, you're gonna buy this car from me," tomorrow, maybe the advertiser actually has to you know, potentially spray and pray and look at four users and are willing to pay $2.5 each. So it looks like the CPMs are coming down, but there's still $10 that, you know, the advertiser has that he's looking to spend. So yeah, I think they go, they go, you know, hand in hand and yet they're orthogonal.
I think CPMs might slide down as, you know, data starts, you know, disappearing, but I think budgets will continue to grow.
Yeah, and also it's often an opportunity. If we see with Dataseat, where we do contextual targeting on iOS, on Apple, if you buy your inventory 20% cheaper and still are good at targeting because the CPMs are down of unidentified targets, you buy cheaper, and if you then have the right ways to target contextually, you have a win on both sides.
Yeah.
So in the land of the blind, the one-eyed man is king, right? So I guess relatively, if you have more data.
We try not to be too blind.
No, we're the one-eyed man. Okay, so any questions from the room on the topic of data and AI or identity and AI?
Yeah, hi. Just probably trying to sum that up a bit. I think there's general consensus that first-party data is getting more valuable, but what does it actually imply for you? Actually, do you see it more as an opportunity if the cookie goes away because you can win market share? Or is your DSP exposure, in your view, too small to really monetize it enough? So just kind of a summary on that one.
We see it as a clear opportunity. In a market where everybody's playing with the same, weapons, it's the wrong word now, but, where everybody has the same means, it's difficult to gain market share because you're just competing head-to-head. But if there's disruptions like there are here, where part of the parties doesn't invest in solutions, where part of the parties is just sleeping, there's a much bigger opportunity to gain market share, and that's what we see already, that a lot of people starting to test out more, let's say, lately. Also, the market being under pressure, people are willing to switch and try new things. That's the positive of it. But we are still also a bit, how to say it? Surprised that a lot of people don't.
Like ATOM, a solution which we already announced, I think, four years ago, the take-up of it and really getting people to use it, to understand what we're doing with it, took much longer than we expected, but it's now coming. You see now really that companies need to prepare for this, advertisers need to prepare. And the awareness, especially by, I would say, the larger advertisers, starts to really come. But there's a big field of advertisers that don't even are not aware of actually what's coming. And the same is also with a lot of agencies, they just continue doing what they did before because it's easy. But and here compliment to Pilot, I mean, that's the reason we really like to work with them. They are super innovative and really looking, what can you do with data?
How can you optimize? Those are the players that will win and gain more market share.
Hopefully.
Are you working on it?
Yeah, I'll quickly add to this. Look, again, it all depends on what sort of solutions you take. It's the most difficult part today is, and was always, to ask for the first-party data from the advertisers. They are not willing to share it with everyone. So in the coming times, the level of confidence, the solutions that you actually go towards, and just alluding towards what Remco mentioned, like Dataseat, we have the ability to actually go find users where 80% of the IDFAs are gone. So it's the trust that you build in and talking and working with those advertisers to actually, you know, "This is the value we bring on," is...
It's gonna be a combination of things, but we believe that if your solutions are good enough, the advertisers should be willing to share the first-party data. It will be the most valuable thing.
Do we have any questions from the web?
No questions for the web at the moment.
So then we can move on to the topic of database targeting. And, Oliver, you presented us with some solutions for the new world, so contextual, first-party data, behavioral, and regional data. How do you see this landscape shaking out? Will one or two of them prevail, or will it be a mix?
The quick answer is, it will always be a mix. You can't just bet on one thing, you have to combine those things, you have to think of new solutions. We are talking so much about new solutions. For example, in Europe, we have this solution that's called Utiq, where the four telco companies, Vodafone, Telefónica, Orange, and Deutsche Telekom, sit together and created a new identity graph, for example, to solve a lot of issues that we are discussing today. But the question is, what will help you to scale? This is the question that you need to ask yourself, and then you use the data with which you will scale and get the best performance.
Speaking about first-party data, I mean, we're talking a lot about how important that will become. But won't this make the walled gardens even more powerful, or will it - won't it be harder to dethrone them?
They are already powerful. I mean, the question is not will they be more powerful? The question is: how will we act on it? Either we follow them blindly, or we play the game copycat. But as a historian, I really don't know any case where you are a copycat and then you surpass the one who invented it. Or will you be innovative? And I think the last one is the most powerful in getting new market shares also.
I'll, I'll get back to one of great examples that Oli shared today. Like, if you have two rooms and you have X number of users in one and X number sitting in, you know, if an advertiser uses a very large platform, they would say, you know, "Facebook or something really works for me." But, but the mind shift and the thinking towards where the right users and how you actually go and reach out to them in a identity-less world, is gonna be a win-win situation for companies. That actually empowers those solutions.
It's also an opportunity. I mean, think of it. They get rid of the first-party data. In the market, we have a lot of first-party data solutions. As I mentioned, LiveRamp, ID5, netID, and so on and so on. I can make a list here for 20 minutes. But this means also that if Google is not betting on the identity graph, there comes an opportunity with it, new market shares that the outer world of the silos can again gain for themselves. It's all about the story that you're telling, so.
We spoke about-
I also... Sorry, one-
Go ahead
... last point. I, I do think that trust and transparent operations are going to be a challenge for the walled gardens.
Yeah.
I think it will be an increasing challenge for them. I also think that they're distracted in a variety of ways right now, which where is Google investing the majority of its innovation R&D budgets today? To fight the new threat of ChatGPT and preserve their search dominance, and that is now a red flag, all-hands-on-deck initiative inside the Googleplex. Facebook has had 12+ different red flag, all-hands-on-deck innovation initiatives in the past 5 years , 6 years, the most recently AR/VR and now AI. I think AI is an extraordinary new tool. I think it's gonna serve as a distraction that can open a window for smaller companies to compete against them more effectively in various of their revenue streams.
I want to add to that point. This is something that we already also saw. I mean, just think of Alexa. When Alexa was introduced-
Mm-hmm.
Everybody was, "This is the new thing." Yeah, everything will shift to that. "I will talk to Alexa. I will not anymore search on Google. I will do blah, blah, blah." Fun fact, Alexa never scaled. So, I mean, yes, Jeff, the thing that you said is true, but let's see where we will come to. I mean, it will change our habits, and the use is changing fast. Take TikTok. What is the biggest threat for Meta? Is it really the ID, or is it a new player-
Yeah
... that is doing completely something different?
Mm-hmm.
What was Spotify here in Stockholm? What was Spotify doing with the media, music business? We consumed a CD, we bought a CD, we bought a something and, and listened to that one month. Ask your kids, how long are they listening to an album, or even are they listening to an album or just one song? So everything is changing. Media habits are changing that fast.
Oliver, from the perspective of a large media agency, managing the budgets of many advertisers, why did you decide to work with Verve?
The last slide said it all. Nobody can repeat it. Maybe just a little bit more information about it. As an agency, you have a lot of talks with different players on the market because everybody comes to you and says, "Hey, I have the next big thing. I have the next solution. Come work with us," and so on and so on. And, let's say five years back in time, what we saw is that all of the major companies, they don't came up with innovations anymore. So they were really just what Sam said, monetizing their platforms and squeezing out as much as they can.
But that's a situation for an agency because we have to talk to the clients, and we are also a business, so we also need our margins. It's not so that we are working for free. Yeah. And this led us come to the point that we searched for multiple partners, that that's not meaning that we are just working with MGI, but we have with MGI something special, and this means that we are talking on a different level, on a transparent level, where we can say, "Okay, this is the case that we are seeing right now. How can we solve it together?" I mean, Rami, where are you? I can't see you there.... how many data sets you are pushing to us, so that we can analyze them and also think of solutions.
This is not common, quite common in this business, that somebody's helping you to create your own solutions. Take the, the stuff, the DCO that we invented with Verve. How many platforms are willing to be so nimble and to do actually something? You know how the process of implementing stuff functions with big companies? They don't ask you like, "What will you solve?" The first question that you get: "How much money do you bring to the table?" Is innovation as a You're investors, do you ask your startup really, "How much will I regain in the next months?" And yes, you ask that, but, "How much will you bring me?" I mean, startup will then die immediately because they don't know. A good innovation don't know. WhatsApp, what was the business model of WhatsApp? And look at now.
Yeah. And a closing question for on this topic for all of you. We can start with Jeff. So answer with two sentences: How do you think the future of advertising will look like in the mid to long term in terms of database targeting?
I'll keep it short and bold. I think of the 11,000 companies -12 ,000 companies in the ad tech, martech ecosystem, I think you see a third of them vanish in the next 5 years.
Jay, we just run straight down the line.
You know, I would say that I don't... I don't have any data to support or not support that. There's certainly gonna be a lot of companies that are gonna get wiped out. But again, the pace of innovation is extraordinary. So let us not, you know, and we don't wanna end up or... I would hate to see us end up. I'm a very half full guy, and, you know, one of the exciting things about the industry that we're in, and you guys are investors in this company, is the innovation, the speed of innovation, both for mid-sized companies like this. Don't underestimate, don't underestimate Google and Facebook.
Yeah.
Just don't ever do that. I mean, the, you know, just don't... It's easy to pick on them. But these are the companies, companies like MGI and their size, they still have the speed, they still have the flexibility, they still have the creativity, to really win in the open web. And I think on the innovation side, I just see the amount of money that still goes into startups, and there's some brilliant startups that are really not businesses long term, but they're great technology.
I said two sentences, not two.
I know. Sorry. I took his third.
Oliver?
I will save two sentences and say, I see it the same way.
Okay. Sameer?
Oh, look, scale and simplicity, consolidation will happen, if not one, two-thirds. There will be a lot of companies which are not ready, or they don't even happen to believe that they are not gonna be ready. But I very strongly and firmly believe, support with the partners and external viewers, we are extremely confident that we have a very bright future. We are preparing ourselves to solve for the complex problems across multiple channels, to help businesses all the way.
Saurabh?
Just an extension of what I think Jay said. Speed of innovation is gonna be existential, right? Like, I would hate to be, you know, a company at some point where my only moat is scale, right? I think you have to have a value, you have to solve a real problem, and then you build scale on top of the moat. But if you grow big enough where, you know, the only advantage you have is scale, and you stop innovating as a company, I think you're on a downward trajectory from there. So I'd love to stay nimble and love to, you know, adapt quickly.
Remco?
Yeah, I would also build on that. I think agility or agility is very important for us, and that's partly innovating ourselves from within. It's also buying smart buys, purchases, but it's also partnering, and I think that got not so much. I mean, we are partnering with Apple and with Google. We are one of the early partners that is working with Sandbox, testing together with them. We're also working with SKAN, we do a lot of things. So it's not only we are an island and everybody is against each other, it's also about smart partnering in the market. Rocket, it's a startup that we're working with, but they're pretty good, they're pretty smart. You don't need to buy everything, so we partner with them.
So also those things, I would really say it's playing the ecosystem right, seeing the problem, and then finding multiple solutions from different angles, because nobody here really knows where it exactly goes. We know the direction, but we don't see the exact solution for it. And having a very good—and it's also more than two sentences, sorry. But having a very good opt-in base with our games, we didn't talk a lot about that, but that allows us... First of all, we have a lot of data from it, but it allows us also a lot of testing. We can test efficiencies, we can even use panel out of them, ask two things. So this whole opt-in gamer database is super helpful for also, yeah, rolling things out. So it's a combination of it. There's not one answer to it.
It's, for me, it's playing the right cards and then executing them in the correct way.
Interesting. And any last questions from our audience? And any questions from the web?
No more questions from the web.
We've overwhelmed everybody too much with data, I think.
Yeah. So I think we'll end on that note, and I'll hand it over to you, Remco, for an outlook and closing remarks.
Yes, perfect.
Thank you.
... Where's the, you have the thing? Yeah, right. Should I put the table in the middle? Yeah, but there's so much stuff on it. I'll do it from here. Sorry, the good thing is I saw not a lot of people using their phones while we were doing the presentations. I first of all would like to thank you all for being here, for listening, and we see as one of our tasks also, and it sounds very strict, educating our investors, our analysts, what are we doing? And we know that it's not that simple, and I hope that we really got a lot further there today with deep dives.
We tried to make it not too technical, and we all know that at the moment you go into details, there's always a detail behind the detail or many, and that makes it even more complex. So we try to find a good middle way between that, and we would also like to get your suggestions. What can we do better? What should we do next year on the Capital Markets Day? I think we showed today that we have come a long way. We started with a bit of history. We have really been able to, from a gaming company, to use the strengths of the gaming part and to come from a B2C company to B2B company, still doing B2C, but really using the strengths of all those things and are now in an...
I think that became clear in the panel into an environment where a lot of things are happening, a lot of opportunities are there, and where we are convinced that we are very well positioned to take it. Contextual took a bit longer to really start being accepted by customers, and it's still not there fully, but identifiers are disappearing, and that's a huge opportunity. And we have the scale, we have the people to really build in on that. I would like to go into some takeaways, 'cause it's a big world we're in and you cannot do everything, and I think that's also something that we try to bring here. We have, in the last years, last 10 years, but especially also in the last 4 years, we're doing AdTech, really been able to build a few very strong positions. The one is in-app.
Pretty proud that we are really the number one in the US on Android and on iOS, and that we have built those positions. Also Europe, strong position, and we are building and extending that into other markets, and it's further increasing market share. We can still do a lot more. What's helping us a lot is that we have this direct supply. Sameer showed it before, and there's so many re-trading of traffic and so many re-trading of volumes, that our SDK base, our direct relations publishers, are super valuable. Where we're not so strong yet is building the last mile towards the advertising to the agency, and that's what we're working on. We have a position there, but we have a lot more possibilities.
CTV, we want to basically copy a bit what we also did for in-app, grow it, become stronger, build on our USPs, 'cause also that's a very, very important segment. Yes, we are a multi-channel platform, so we also serve the other segments, but you cannot build everything at the same time. I think with spearheading in-app, we do very well because that's a device that's everywhere. CTV is the next big growth thing. Yeah, and the glue between that is data and AI. I think I shouldn't talk too much about it anymore because you have heard so much about data today. But it's all about mix. That became also clear. It's not just contextual, it's not just first-party data, it's really finding the right mix and innovating on that and being well prepared for an ID-less world. Then cost.
We have a flexible cost structure. We do not only buy, we also partner. We look at flexibility, we look at execution, and we just did, also not to forget, a EUR 10 million cost-saving round on personnel, which we could do because we're integrating platforms. But I think that's also very important. We are, and for investors, we are watching our profitability. And even in a time where you have a lot of headwinds, where the revenues are not growing as we were used to, we protect our bottom line. Very important. And then growth. Yeah, as said, revenues are flat, but we are gaining market share. We're really growing the company, and that's super, that's good. And if you look at where the market goes, nobody has the crystal ball, but we will see economies going up.
We see market leaders already starting to spend a bit more, so we're very confident that we will also profit again from higher demand and higher CPMs in that. So those are the, let's say, takeaways. Strong position in in-app, strong position in CTV. Upcoming, building, basis is there, but we need to scale on that. Data, AI, very strong focus point, agile on the cost side, and ready for growth. So I hope that you take that with you as an investor and challenge us on it, ask questions, and we hope that with the Capital Markets today, we are really adding also to you understanding what we're doing, and also to see the potential that this company has. Brings me to thanks. I would like to thank our external speakers. This was an experiment to bring people in from the outside, so did you not...
We are a big team, so we could also have had many more people from us on the stage here, but I think it was really good, and I would especially like to thank our American colleagues. Jeff took 20 hours of flights in for getting here. Jay, a little bit less, but also still, they're jet lagged, they did a great job. And also, Ollie, of course. I mean, you're a great partner to work with, and Germany is a bit closer, but also you coming here, telling the story, so we would really especially like to thank the external speakers. Then a big thanks to the team, all the people that were on stage, but also the people behind the curtain. Sören sitting there, and I think you're ready for going to bed. After a lot of long nights.
And Heinrich, many others with it, also Jenny, and, of course, Danesh. And, we're very happy to have somebody on the ground now also in our investor relations team in Stockholm, in Sweden, so that makes things good. Yeah, and the biggest thank to our investors, analysts, people that are here, but people online also. Great that you support us, follow us, help us, and please continue to challenge us, and we'll do our best to build this company to grow it. Thank you very much!