Welcome to the Verve Group Q3 presentation. During the questions and answer session, participants are able to ask questions by dialing star five on their telephone keypad. Now, I will hand the conference over to the CEO, Remco Westermann, and CFO, Paul Echt. Please go ahead.
Good morning, everybody. I would like to welcome our investors, our bondholders, and all other stakeholders to the presentation, the financial hearing of our Q3 2023 numbers. The presentation today will be done by Paul Echt, our CFO, and myself, Remco Westermann, both long time in the company. Company is listed on Nasdaq First North Premier and on the Frankfurt Stock Exchange, and some information on the shareholdings here. No big changes there, let's say, looking at the shareholdings. Coming to the next page, a bit of a background of the company. We have been showing a very nice, sustainable growth over the last years, with a good CAGR of 41% and even a higher EBITDA CAGR. This year, 2023 is a special year, where you also see that. How to say it? There is not so much growth in the numbers.
We will get back to that later. The market is under pressure. We are in macroeconomic headwinds at the moment, and that is also, how to say it, resulting in not so good numbers on the advertising side, so advertisers withdrawing their budgets. Quick update on what we do for those that don't know us so well. We are an, yeah, advertising platform, which is strong with data, and what basically happens is, at the moment, a user opens an app or opens a website, and there is an ad on that website. This ad gets auctioned. The auction process must happen very quickly because at the moment the user opens the page, of course, he doesn't want to see an empty ad space. So this whole thing needs to happen within 100 milliseconds. In that time, the ad is being auctioned.
That's the process that we are doing. The ad is being, how to say it, enriched with data, where also our own games play a role, but also third-party data, other data. The more data we can enrich the ad with or the ad request with, the higher the bidding will be on the ad, and that's where the advertisers come in. That's the other side of the equation, where the advertisers bid, and the highest bid is winning the ad, and then the publisher is happy with the highest bid in that sense. So that's what we're doing. It's like a stock market, basically. We do this on all platforms, so for mobile, web, traditional out-of-home, for CTV, and for in-app, where in-app is our strongest part. I would go to the next slide. Yeah, highlights Q3.
That's what's the target today, to talk about Q3. We are still in a market which is depressed. Advertising budgets are depressed. Also, the prices per ad are depressed. CPMs are down 20%-30% versus the same period last year. Our Net Dollar Expansion Rate is still low at 93%, which means that the old customers are still, let's say, spending less than they did a year ago, but we see light at the horizon. We talking to advertisers, to agencies, we really find out that yeah, people are getting a bit more positive, need to do more ads, and that's basically what yeah, how to say it, happening and coming in. Then further increase in market share. We have been working on adding new customers on also increasing our ad volumes.
We'll go into the detail later in the presentation. Extending our market shares in the U.S. and EMEA. We are also working in APAC and LATAM, but that are the smaller shares. Our main market is the U.S., with 70% of the revenues, roughly. Then we were able to show 1% organic revenue growth, foreign exchange adjusted. Overall, we have an 11% reported revenue decline, which is due to foreign exchange and also divestment of our small games, which we did end of last year. Then successful cost management. In times where the revenue isn't growing so fast, where our revenue was always growing faster, but change of how to say it panels. We have to also carefully look at our costs, which we've always been doing, but even do now more.
So with active margin management, cost control, we've reduced the cost end of Q2 by EUR 10 million on annual basis. So that's also starting to yield, and we are able to increase in Q3 our EBITDA from 26% to 29%. So really showing the very nice cost control that we have. Improved cash position. Paul will talk later about the further finances, but we had a very strong free cash flow, despite some earn-out payments that we did, which give also room for further deleverage. These are the headlines. We will go further in detail. So increased market share, further product development, good cost control, and an improved cash position. Going into details in the business update.
Looking at the different elements of the market, yeah, on the left side, you see we are connected to the, how to say, to the general economy, which means that the ad spend is going in parallel with the GDP, and that is, yeah, at the moment, GDP is down, the ad content is down, and the first thing where people or companies that are, let's say, seeing a not so great environment are saving is the marketing budget, because that goes fast. That's what we have been seeing. By reducing the marketing budgets, there is less demand, and on a bidding platform, of course, also the prices go down. That's what you see on the right upper graph. Overall, however, the advertising market is a strongly growing market.
We are over $800 billion as a total market now, close to $900 billion. But let's say it's, yeah, at the moment, a bit hit by the economy. Within the overall advertising market, we see that programmatic, or let's say that digital advertising is increasing. Digital advertising is, yeah, the one. We have so many channels now. We have so many different ways to advertise. It's impossible to do that manually. So in that sense, it's logical that things move to digital, and one of the big ones moving is TV. TV is also going from linear to CTV. We see that much faster in the U.S. than in Europe, but also in Europe, the tendency is very clearly visible. Go to the next slide, where we see the numbers that we are, we did in Q3.
Now, what we still see is declined budget. So Net Dollar Expansion Rate, which is the money that the people spent a year ago versus this quarter, or in the quarter a year ago versus this quarter. We had an increase, in the Net Dollar Expansion Rate in Q3 2022, and now we see a decrease, which is really due to the economy. So we have to compensate 7% revenues on that, versus Q3 last year, and Q3 last year was already weaker than the quarters before. Then the over $100,000 retention rate, that is customers, still on board since last year, the third quarter.
We are roughly stable now, so it's a low churn of 4%, where partly even some customers are getting under the EUR 100,000 spend level because they're spending less, so they should come in there again when the economy recovers. Then, in total software clients, so how are we compensating this? So basically, our existing base has done less. How do we compensate this? That's by growing customers, by getting new customers onboarded. We are for the first time here, also now communicating our total software clients, so that's customers that are on the advertising platform, which has nicely increased by 9% if we compare to last year, same quarter, and the ad impressions, where we also see a nice increase in overall ad impressions. It also yields in market share gain, so we are winning against the market.
An example you see on the right side, which is the market share for mobile in-app. Yeah, you will see a really nice gain of it, showing that we are growing our market share also in this difficult market. Go to the next slide. Yeah, just said already, efficiency improvements and cost control. We had the EUR 10 million cost saving program, which was enabled by technical optimization. We acquired several companies. We are looking at merging the stacks, and merging the technology platforms, and yeah, by just becoming more efficient. Also looking at yeah, or planning further efficiency gains, optimizing bidding processes, just using less bandwidth, these kind of things, but also cost reductions are further possible based on further platform optimization. Then we also in parallel, further investing in innovation and platform unification.
So to make it very clear, we're not milking the company, we're really putting a lot of emphasis on growing the company, on innovating, and really gaining market share. And yeah, the company is has a high ability to adapt its cost to protect profitability. We have shown that, and if things would get worse, we still have a lot of potential there, but we would rather really continue to invest in further growth and further market share. Very nice to see that, even though the revenues are basically flat, that we were able to really increase our EBITDA margin from 26% Q3 last year to 29% Q3 this year. Coming to the next slide. Yeah, here are our pillars, our main attention pillars, or our main focus pillars. The advertising market is big.
These are the things that we are really mostly focusing on. The one is in-app advertising, so things in the apps. Then, the CTV market, data and targeting, very important and vertically integrated. Those are the four I will go in more detail in the next slides, but it's a strong base to build from. And just to remind, our mission is, "Let's make media better," and there's still a lot of headroom to go there. Starting with the in-app side, I showed it already quickly before, here in a bit more detail. North America, we are, based on the Pixalate numbers for Apple and for Google, the number one in the market, we were also able to increase our market share there. Also in EMEA, we are developing nicely. EMEA, we're not as strong as U.S.
I said before, 70% of our revenues is in the U.S. Let's say below 20% is in Europe, but also here, we are building on our position and increasing our shares. Yeah, we have done a lot on the in-app side. We have onboarded a lot more new publishers. Some names here, Tripledot Studios, Popcore, Tribune. So it's, let's say, gaming, but also non-gaming that we're going for here, and these are just a few examples. Then also on the demand side, we onboarded partners. So those are demand-side platforms like Integral and Sharethrough that were onboarded. Then on the product side, we have two major updates on our SDKs. SDKs are the software development kits that are integrated in apps, which we are able to display the ads.
So I'm really happy, let's say, relaunch those, then they need to be integrated in the apps, and then the people need to update our apps, so there's always a certain time cycle behind it. But these are major updates that we did, and they are driving eCPMs, and helping, of course, more revenue for publisher partners, more revenue for us, and better conversion also for the DSPs. Then, what we also did on the product side, we launched, expanded video capabilities, rewarded playables. So also on the video side, there's a lot of video advertising going. We increased or improved our capabilities, which should also drive further revenues. Those are a few headlines on the in-app sites.
Yeah, that's what's driving market share, but also helps is that we have these market shares, because also that opens a lot of doors, of course, to potential partners, because having this position, yeah, people like to work with the larger parties in the market. Then going to the next slide, CTV. Yeah, CTV, we have also been working on improving our position. So also here, we onboarded further publishers, for example, France TV , Verse TV, Rakuten, Vizio, and TCL. It's mostly alternative providers that are really strong in CTV, especially if you look at Europe. The incumbents are still a bit slow on that, but are also coming. Europe is anyway a bit behind. U.S. is going much faster. We also onboarded demand partners, for example, Magnite, in this case, also more here.
Also here, adding capabilities, OpenRTB 2.6, that's the new standards, the IAB, so the Internet Advertising Bureau, publisher standards. That's very good because people can work together based on those standards, and this is a larger update, which is really allowing all kinds of ad pod types. Ad podding is that you put certain ads behind each other and then in a combination. Then, yeah, further working on those things, also, the auction process and the ability to apply more granular targeting was also improved, where we have more subcategories. Sounds very technical, but it's super important, of course, for the advertisers to get these kind of capabilities, which just brings a better yield on the advertising. We also enabled Adelaide. We talked about it before.
Adelaide is a company that has an attention metrics, which is very important for advertisers, which is showing the attention people pay to an ad. Comes very well with advertisers and is a very nice addition that we did. Going to the next slide. Data. Yeah, data is a key success factor for addressability, targeting. Data is what makes the, let's say, the seemingly impossible equation between a publisher wanting as much money as possible and an advertiser not wanting to pay too much. But with data, we can really show which ads are, or which, let's say, people to target are interesting for an advertiser, and an advertiser is happy to pay more for those. So more data is better targeting and makes more money for everybody. We have a lot of data.
We have data from our 20,000 apps or 20,000 SDKs that we have in apps. It's many more apps, but it's in there. Then we have data from our own games with over 1 billion users, over 5,000 games, and we have a lot of additional data, partly from inside the company, so from Moments.AI, which I'll cover later in the presentation, which is contextual data, but we work also with a lot of external partners in this. What do we do with those data?
Yeah, we can do cross-screen activation, so we can really, let's say, target people that are watching in the mobile, maybe in the morning, then use the PC in the office in the afternoon, and look at the CTV screen in the evening, and also pass on the way there by a digital out-of-home screen. So that's really important that we can target people on different devices. That needs identifiers in that case, so that's why we use our ID Graph. Then it's about precision and quality, not only in environments with IDs, but also in environments without IDs. It's about quality, and advertiser wants to show its ads in an environment which is controlled and not which is, yeah, maybe showing some negative things. And then very important, privacy by default, super important. Regulation becomes more important. Rules become more important.
We have CCPA in the US, for example. There's now six states in the U.S. that also have a kind of similar to GDPR, not as tight, but also there's talking that they will move to opt-in. GDPR in Europe, COPPA, LGPD that's Brazil. There's many more countries now at the moment really going in here. That means that it's challenging the markets to target without identifiers. That's something that we like because it's a disruption, and that's something that we are really very nicely prepared for. We have a very good household reach, so basically, we can really address a lot of people in their households or on a personal basis, which of course, is super interesting for advertisers. Go to the next slide. Yeah, an integrated multi-channel platform.
Normally, we have the picture here where we show that we are between the advertiser and the publisher and trying to cut out as many steps in the middle as possible, or to really do it direct. We chose here to have a picture on the right side, which is showing our direct supply. This is on mobile, where we're one of the leading partners, so more to the right is better. And in that sense, we're doing very well there. We're really putting a lot of attention on that. If there's more partners between the supply and the advertiser, it's getting more inefficient, and that, of course, is something that, yeah, costs money, but it's also giving a way of, yeah, let's say, that less transparency, and a lot of other negatives. People might even bidding against themselves, things like that.
So direct supply, super important. What are we further doing on the verticalization, or what did we do in Q3? Further adding new demand partners, integrating the tech stacks, feature and product development, improving machine learning algorithms. I cover that a bit later in the presentation. Improving our data lake, that's where we collect the data, and also our data capabilities, and adding supply partners, which focus on direct supply, as Jess already said. Brings me to the next page, a bit about our games. We haven't covered games so much in our latest presentations. We said that we're getting questions on that, so very important to state, games is super important for us. It's where we started with, and then we started adding the ad tech capabilities.
But games by itself is a very interesting segment, so a lot of things happen in our games. We're driving the revenues there. As said, we, how to say, it's disconnected or we stop some smaller games, by end of last year. We're focusing on the larger games. We're focusing on as well on the MMOs, as also on the casual games, PC and also mobile. Doing very nicely, Wizard101 had a major update, the Crying Sky Raid. We're partnering with Epic Games on some distribution and also some development. We had a new server launch for, Fiesta Online, one of our oldest games, actually, a game that also is nicely developing. But the other point for our games, why is it so important?
Games is also a means of getting data, a lot of data, and we are close to our gamers, so we have a lot of possibilities to get opt-ins, but also to get, of course, more enriched data by, for example, doing questionnaires and these things. Also, your privacy first, of course, is very important. But, yeah, we did further integration of the gaming data in our ad software platform, so that also allows a lot more targeting possibilities and, close cooperation on testing. That's something, that's super important. If you have new features, if you have new products, on the advertising side, they need to be tested, like our SDK, for example, needs to be tested, and that's what we can do much faster, with our own games than we would be able to do it with external partners.
So a lot of absolute pros here. Also interesting that we are, let's say, having our games on all different platforms, it gives us, again, a lot of different data. Not every game is on every platform, sorry to just make that clear, but we are covering with our games the platforms that you see here. And good gamer base, good base for our further growth of the company. Then I would like to go to the next chapter, which is going a bit in depth. We wanted to show you a few areas where we are further growing, where we're further investing, where we're seeing some good results. I'll cover machine learning. Everybody's talking about AI nowadays, so since we have ChatGPT and many more.
There's a lot of things happening in the market, a lot of positive things coming from there. AI is something that we use already for a long time in this industry, but important, I think, to show and to give a bit of, more in-depth, view on this. Then we have contextual targeting. We are in a market where identifiers are disappearing, so contextual targeting is the future way of targeting. Super important, something that we have been investing in by buying companies in this area, but also by further organic investments, super important for us. And then demand partnerships. We are a company that's super strong on the supply side.
The demand side is the smaller side, but that's one of the targets for the coming quarters to really also get closer to the demand side, but I'll cover that in a few slides coming up now. So the first slide, yeah, machine learning, basically, to very simplify it, we take all we have on data, on classifications, et cetera, and we put it in a data lake, and then we have AI, which sorts it out. That's, of course, super simplified, but what are we getting out of it? With all those data in there, we get an identity graph. An identity graph is basically mapping for a user all the identifiers that are attached to that user. So that can be a mobile ID, that can be a cookie, it can be an IP address.
But basically, if this user shows up in an ad, so if there's an ad request from this user with an ID of this user, we can map it or bring it back by the ID Graph to a certain profile. With this profile, we can see that, for example, the user is interested in gaming or is interested in cars or interested in skiing. Even we can see things like has a good mood, bad mood, these kind of things. There's a lot of things that you can interpret in there. That's important then, yeah, individual persons belong to a household graph. All these things are done by AI. It's just too many data to do this manually, and that's what you see on the right bottom side to just give you an idea what our platforms are covering.
So we have 450 billion bid requests per day. So that is request for ads. We do 5 PB of data, which are going out of the system every day, and 25 PB of data that we store every month. So that's really showing it's big data and machine learning on the right upper side. So we have a lot of processes that work on, AI or machine learning, algorithms. Yeah, super important for us. Going to the next slide, where we go a bit more in detail. We split here the AI in, the front office, product and services, core capabilities, and back office. Yeah, to start maybe with the back office. In the back office, we use AI for more precise targeting, to really make sure that we target people, that we follow them.
The development of automating routine tasks is also very important. Our developers work with a ChatGPT comparable tool, which really, it does code review. Normally, in the second program, I needed to do that. Now we use, automation for that. Graphics for gaming is a very interesting field that we are looking into. The streamlined ad buying and placements, resource optimization. So there's a lot going on in the back office. On the front office, yeah, it's about, really towards also the users. The better our ads fit the user, the less, they are, seen as a hindrance, the less they are seen as annoying. So also there on the front end, towards the users, it's super important to give that, as well as to our partners, the advertisers, to give them better insights and, analytics.
Then on the product side, Moments.AI, our contextual targeting tool, Visual Intent, also contextual targeting, ATOM, coming to that in a minute. Household targeting on CTV and ID-less graph is sort of also an ID graph, but basically without an ID. Very interesting product that we are rolling out. And then if you look at the core capabilities, it's really making sure that our bidding routines are let's say further developing, that we get better there. And we're also looking into new technologies. Neural networks is another way of using, or let's say it's a way with let's say neurals or different ways of using AI, which is let's say more, a bit more expensive, but much more accurate. So we're looking into that. And also quantum computing. That's not ready yet.
It will happen in five or 10 years, but also these kind of things we need to look into, because that is probably gonna change quite a bit in the future. Then go to the next slide. Yeah, what's happening in the market? I said before already, identifiers are disappearing. IDFA of Apple is the first one that came in the market. It's not actually, not even the first one. There were also some browsers already taking the cookie out. Now, the big one is lined up by Google, who is gonna deprecate its cookie for web in the first half year next year. People are saying it might be delayed. We don't know it, but it will come at a certain date. And also, we see that, let's say regulation is changing, GDPR, et cetera.
So basically, we are moving from an opt-out environment, where identifiers were widely available, to an opt-in environment with deprecation of identifiers, where there's less or hardly any identifiers available or even no identifiers, partly. That's a big change of the for the market. That's a disruption, and that also, of course, gives a lot of opportunities. Bring me to the next slide. Moments.AI is our contextual targeting solution, which really allows us to target without identifiers. We have 700 off-the-shelf standard segments. We can also make special segments, which we do for advertisers if they want to have them. And the segment is, for example, car lovers, female, certain age category, but also interested in certain, yeah, other interest areas. So those kind of segments we can do, shopper, whatever.
Then we also have a cooperation with Getty Images. That's super important or super nice. Getty is one of the largest photo libraries in the world. And for example, if there is Olympics or if there's football games, a lot of their images are used, and we have a cooperation with Getty, where we can place next to the image, we can place the ad, which fits to it. So also this is contextual. Also, this is a very nice example of how we can use this. And of course, brand safety accounts for itself, is super important here as well. So that's about Moments.AI. Then go to the next slide. A bit more in detail. Yeah, what can we do with contextual?
Contextual is much more than just, it's a football website, and we look for people. We target people that are interested in football. It's really going much deeper. It's of course interest, but we also talk about confidence. We talk about attention, sentiment, and also how old is the content? 75% of the pages that people go to are less than 24 hours old. So that's very important, and viewability, of course, ads are sold that nobody has seen because they were not even loaded. So these kind of things, or they were lower on the page. Very important. Also here, again, we use AI processes and algorithms, natural language processing, content classification models, interest learning models, and brand safety and attention metrics. So these are examples on Moments.AI. On the next page, we have a case study.
Traditionally, and this is for Viaplay Group, which is a Scandinavian company. We did a test, for Norway, about subscriptions, they were targeting. They were using traditional, methods with third-party cookies, audiences, and retargeting. We have put, Moments.AI, our tool, next to it, without cookies, without IDs, with zero privacy risk, and we were able to show an 8.32x return on advertising spend, so doing better than their traditional way of advertising. It's very nice to show that, let's say, without the use of identifiers, we can really be even better than party people with identifiers. This has to do that a lot of identifiers are sometimes conflicting, and there's a lot of different, challenges also on the identifier side. Go to the next page. Yeah, ATOM.
In the past, we have been talking about ATOM. It was maybe a bit early at the time, but that's because we're also very proud about this. ATOM stands for Anonymous Targeting on Mobile. Also, this is a product that is there for a world without identifiers. That's now being rolled out on iOS for Apple, which means this, we have a lot of signals on the mobile phone. Everybody who has mobile phone, there's a lot of information on this mobile phone. We put a piece of software on the phone and collect information on the mobile phone, and by that, are able to build segments, but this individual information is not leaving the phone, only the segment information is leaving the phone, and that allows us to better target.
This is fully privacy-compliant, and it's a very nice way of doing it. We have been in open beta, and we are now, let's say, rolling out, scaling the ATOM solution. In Q1, it will actually be bundled to our advertising SDK, so then we expect to really gain a lot of reach. And we are on top also onboarding demand partners, so advertisers. People are really starting to wake up, that there is gonna be a world without identifiers. Let's say, timing-wise, we're really very nicely here now and, yeah, getting a lot of interest in this. On the top side, you see a bit on the product roadmap. There's ML on device for gestures, for example, there's also more insights on it.
There's a lot of stuff that we still can do, with this, ad and assets. Development goes on, but also rollout has started, which will, again, we're not charging actually, especially for the information or for the things. It's just giving us a way to target, so in the end, we will see higher than the expectation, higher CPMs, and make, yeah, just get more bits done on it. Then the next page, demand partnerships to increase our demand side. Roughly, if you look at the advertising side, rightly, 90% of us is supply, 10% is demand. So we're strong in supply. We're directly or connected to the publishers, which gives us a lot of data, which is in the world of, yeah, no IDs is very, very important. But it's, of course, important to also cover the demand side.
That's where there's now more focus on. Just looking behind it, the top five worldwide advertising agencies, WPP, et cetera, tend to control more than 50% of all the ad spend in the world. So out of those over $800 billion ad spend, over 50% is controlled by those agencies. Behind that is, of course, are the advertisers. But it's important for us to partner with those agencies, also with those advertisers, and to really make sure that, yeah, we continue and build on that. The-- Go to the next page. We see the-- Oh, sorry, not to the next page, to the right side, I wanted to say, just to confuse the moderator here.
Yeah, we have a strong position on supply side, I said, but we are working on the demand side, and the main targets in the next quarters are improving our relations with those agents and advertisers, but also part of it, having the right people that speak the right media language, so we're also recruiting on that side. We have people that are good in gaming, we have people that are really good in ad tech, but the language of those agencies partly is a different one again. We have already quite a few people there, but we are further extending that. That brings me to the end of my part, and I hand over to Paul. Paul-
Thank you, Remco. Yes. So, yeah, starting now with the financials of the third quarter and going to the next slide. So basically starting with the commercial development to just re-emphasize on certain things which Remco already touched base on. So we saw 8% year-over-year growth in the ad impressions. With every ad impression, we generate revenues. And as Remco said, CPMs are down, which means the 8% ad impression growth we currently don't see directly in the organic growth of the company. But we take a lot of market share, which really builds a very strong base for future organic growth. And that's also what we see in the growth of large software clients, as well as total software clients.
The large software clients growing with 2% because some of them also due to reduced budgets dropped out of the more than 100K per year revenue basket. But what we especially see is on the total software clients, that we're further onboarding a lot of new customers, and they are also building a very strong base, so we're also further diversifying our customer base. And in addition, these customers, when the overall economy becomes stronger again, will also scale much stronger, which means then we really have an extra accelerator for our organic growth. Coming now to the next slide, and making a deep dive into the third quarter financials.
So here we see a -11% revenue growth, which is, as already communicated on the capital market day, coming from the divestment as well as FX, which makes EUR 10.2 million in revenues. When we take these effects out, we see actually 1% FX-adjusted organic revenue growth. What I think is especially strong in this quarter is that we were able to protect our profit, strong profitability, and by that, also increasing our EBITDA margins from 26%-29%, and that we already saw the first very strong results also from the cost-saving plan, which then also led to another increase in our cash flows, where we now bring the EUR 28 million operating cash flow and also showed a very strong cash conversion of 112%.
Coming now to the next slide, where we see the long-term financial developments. Similar to the previous years, we see a very solid last twelve months, yeah, development, where we now see EUR 360 million revenues, EUR 95 million EBITDA, which brings us now to a 30% EBITDA margin, where we already see, I said earlier, the first effects also come off cost-saving plan. Coming now to the next slide, and here we give a brief outlook what we can expect for the fourth quarter. So due to seasonality in the games, but also in the media business, the fourth quarter is always the strongest quarter, due to Thanksgiving, Black Friday, the holiday season, et cetera.
And there, we also expect a very nice, additional, revenue growth compared to previous quarters, and in addition, also, to increase further our EBITDA, which then also means an additional free cash flow, which also will help us do further delever until end of the year. That brings us to the next, page, 30. And here we see the operating cash flow and CapEx development. On the left side, we see that we now, printed EUR 107 million in operating cash flow, by deducting the interest expenses of EUR 36 million, as well as the maintenance CapEx of EUR 8 million, we come to a EUR 63 million, free cash flow. And on the right side, we also see that due to our strategy to put less-...
Focus on M&A and having not having done any M&A transaction in the last 12 months, we basically also reduced our expansion CapEx tremendously compared to previous years, and therefore have a good cash flow generation, which also also gives us room for deleverage going forward. That brings us to the next page, and here we also see the leverage development. On the left side, we see that we delevered now from 3.6 to 3.2 year-over-year, and that also due to the free cash flow, which we in the fourth quarter expect to be below three times again, and therefore to meet our midterm financial targets. That brings us to the next page, 32.
Here we wanted to show an additional information which had a strong impact on the overall financials, which is the release of EUR 62 million in earnout liabilities from AxesInMotion, where more details can also be found in the report. Overall, we reduced our earnouts compared to end of Q2, from EUR 109 million now to EUR 40 million. And out of this EUR 40 million, we basically have EUR 25 million, where we as a company can decide if we want to pay them in cash or in shares. The majority, actually, by mid of 2025, and EUR 15 million are still on the balance sheet from the AxesInMotion acquisition, depending on the performance of the company.
In 2024, where there's also new products and things coming, so there's also positive outlook as well as, and I think that's maybe also is one of the reasons 90% of the revenues of AxesInMotion is coming from in-game advertising, where, as we saw earlier, CPMs are being down by 30% roughly on average year-over-year. So that's directly impacting, basically, also AxesInMotion, where once the market recovers, there's also a very positive outlook for this business again. So it's less company specific, it's rather the overall market, which puts effects here.
That said, I think we have done also a very good job in regards to how we structured the deal, and therefore took the risk out for the company and which we now then also recognized on the balance sheet. Overall, yeah, we also have a strong cash position, which we also improved quarter-on-quarter, despite some earnout payments, and also the equity ratio increased to 37% following the revaluation of the liability from the earnouts. That brings us to slide 33. Here we wanted to reiterate our guidance. So we still expect to be at normalized 2022 levels in terms of revenues as well as in regards to EBITDA.
We see upside to this, without want to be too aggressive now, so rather taking it conservative. But overall, as Remco also mentioned already, there is some light at the end of the horizon in regards to ad spend. Yeah, Q4 will then be reported by end of February. We come now to the last slide, and then from the financial side, before I hand over to Remco again, our midterm financial targets. Here we also reiterate our midterm financial targets, where we have a revenue CAGR of 25%-30%. Looking at the current organic growth of 1%, that seems quite high, but we're currently taking a lot of investments into our data, AI routines, as well as into the platforms.
And, by doing, step by step, a lot of investments and roll them out, especially also new products, we see very concrete effects from that. And therefore, in addition, when the market becomes stronger again, we are very confident that we can reach that revenue CAGR. Again, where we always say half of it should come from organic, and half of it from M&A. While we in the past also have been able to show 38% organic growth, and therefore, this company can scale much faster in a better market environment. Looking at profitability, we also reiterate on our profitability targets, where we currently being at the higher end of them.
That is also due to our cost-saving plan, where we could really already show the first effects in the third quarter. Looking at leverage, yeah, I say it's two-three. Currently, we are there, slightly above the target, but due to the free cash flow until end of the year, we also expect to be within that range by end of the year again. And now I would like to hand over to Remco for the last commercial outlook.
Thank you, Paul. Yeah, coming to the last slide of the presentation, which is a bit of a summary and, yeah, showing the things. In-app, I mentioned already, that's our core at the moment, that's our core driver. That's where we are strong. Where we have leading positions in the U.S. and the EU, which we further built on. We have built those in the last four years, and we are further building to strengthen those by further increasing our market share. So in-app is very cool. In-app, though, is still a small part of the ad spend overall. And somebody from an agency told me yesterday that there's only 3% of the total advertising budgets go into in-app, while there's 40% of potential people go in there.
So that's also something which we're working on, is really getting more attention on the demand side, on in-app, and the ways you can target that and what you can do there. Then CTV, the fast-growing part. Yeah, TV traditionally was very big. That's now all moving step by step to CTV. So that's our second focus area. Yeah, building on our in-app strengths and our data. So we leverage a lot of what we have on the in-app side and with the data. We are adding demand partners here, adding supply partners, investing in features, and see also a very nice position building up in CTV at the moment. Data and AI, yeah, I went into this much deeper today. Data is the goal of this industry. It's really extremely important and become more important in a world without identifiers.
So the contextual, the combination actually, which we have between contextual first party and behavioral, behavioral being the one with the cookies, is very important. Yeah, and we are very well prepared for the ID-less world, and we rather hope that this is coming faster than slower because it will really help us gain market share. Then cost side, flexible cost structure, ability to adapt our cost and to predict profitability, which we have been showing. Yeah, and growth at the moment, we are a bit hindered by the overall micro environment, but as already said before, we see a bit of light. Talking to our advertisers, talk to the agencies, there's more positivism. We have seen in the past a lot of budgets that were postponed.
Yep, we're gonna spend it in Q3." Then we heard, "No, we're gonna spend it in Q4." But yeah, not wanting to run too fast here, but there is a bit of light showing up, which might bring us, let's say, maybe already in Q4, but hopefully in early 2024, also much better numbers. And then the backwind that we have now with low CPMs might help us on the other hand, when they all go up again. Then yeah, as so, outlook increases CPMs and return to our two-digit organic growth, which is in line with what Paul said before. That is the end of the presentation. I hope this gave a bit more insight. We tried this time to go more in-depth without becoming too technical, but I would like to hand over to the moderator and to yeah, ask your question.
Thank you very much.
If you wish to ask a question, please dial star five on your telephone keypad to enter the queue. If you wish to withdraw your question, please dial star five again on your telephone keypad. The next question comes from Fiona Orford-Williams from Edison Group. Please go ahead.
Good morning, Remco and Paul. First of all, can we talk a little bit more about CPMs in Q4? On the graph that you showed, the September showed a narrowing of the rate of decline over August. Has that continued as you've seen it into October and November? And second question would be about the SDK upgrades you've got, you've done in the quarter. What sort of lag would you expect to have before you start seeing the benefits of those? And are there more SDK upgrades in the pipeline? And my third one, because we're always allowed three, is on market share gains. Where do you see them coming from? Who are you gaining from? And is it about reach or is it about price? Thank you.
Thank you, Fiona. Paul, I will start with this and please fill in on the things. First, on the CPMs. CPMs have a lot to do with demand. At the moment, there's more demand. The people start to bid higher, and that means also the CPMs go up. And on the other hand, if there's less demand, which we have seen now in the latest in the last quarters, of course, the CPMs go down. We see a bit of a tendency of them going up, but that's normally always a bit in Q4, because Q4, by definition, there is more demand. We have Thanksgiving in the U.S., we have Black Friday, Cyber Monday, Christmas coming, which is more a topic in Europe. So all those things drive demand and also get higher CPMs.
Nevertheless, as said before, or indicated before, we see a bit of light in the tunnel, and we are hopeful that we will see improvements there during the quarter. Then your question on SDK upgrades. There we normally take until we really see an SDK in the market. Of course, before it's released, there's a lot of testing going on. So I'm not talking about that now, but after release, till it's really in the market, we talk about two months till we really see effects, and it can take up to half a year or more till we really have 80%-90% penetration into the ads, into the apps. It depends a bit also on what kind of upgrade it is.
If it's a bigger one, then, publishers also tend to, upgrade their apps with it faster. So that's an, important thing. So there's a cycle, and we typically, try to... Let's say we do on a regular basis, SDK, so say once a quarter, but these ones are now really substantial, much bigger than the ones that we did before, so also getting a lot more features there. But still also working on next upgrades on it, yeah, further improvements. And then market share. Yeah, it's good to see it, especially on Apple. We have been winning market share in the U.S. That's where it shows that our contextual capabilities with Apple is largely without an IDFA now. So that is really where our contextual capabilities are showing, that customers are interested in it.
They're seeing good results, and that's helping us there. But overall, our focus on in-app and also being, let's say, in a leading position there now opens, of course, a lot of doors. Yeah, we have really good technology there. We have a good demand side on it. We have, yeah, very strong supply. So that really, they all fit together, and that with the data part, contextual, or in the case of Android, still behavioral, making sure it's strong. So I hope that answers your question.
Thank you.
The next question comes from Ellis Acklin from First Berlin. Please go ahead.
Hey, good morning, guys. Thanks for the detailed presentation. Just one question from my end regarding the earnout release. I was wondering if you could break down that release in terms of what was due in cash only, versus what you had the option to pay in stock or cash?
On the AxesInMotion, everything was to be paid in cash, and therefore, the full EUR 62 million is basically coming from this cash position. Yeah, or from the cash payment obligation. Yeah.
Okay, thanks.
Maybe just to add to it very briefly. So the company still adds a lot of new customers, and there's also a product development ongoing, as well as the multiplayer, where we don't want to say too much about right now. But overall, the outlook for the company remains strong. And also given that the overall softness in CPMs, et cetera, is also impacting the company, we also have a good outlook. So it's rather something where we see some delays, I would say, in terms of the plan which was initially set up, but not that the company as such is not in good shape, or is not developing in the right direction, or that the product doesn't have the same market that we expected when we did the acquisition.
So it's more market sentiment and there's good development ongoing, and there's a very strong product roadmap for the future for AxesInMotion.
The next question comes from Jörg Philipp Frey from Warburg Research GmbH. Please go ahead.
Hi, guys. Just could you talk a bit, you said you see light at the end of the tunnel. Which industries are you currently observing and improving advertising trend? Anything you want to point out here? And then regarding, I find this an interesting statistic regarding just in-app advertising's share of 3% of the overall ad market, and what do you see what you can do to really increase this market share, and what are you-- what do you think is holding advertisers back to increase this share?
Remco, do you want to take it? I think you're on mute. Maybe if you-
I'm also ready. I started already answering, but it's on mute indeed. Sorry for that. Good question. Thank you. Yeah, light at the end of the tunnel. We need to be a bit careful with it because early this year, everybody was expecting, let's say, everything going up end of the year, which it really didn't, at least not in Q3. But what we see is, let's say, interest seems to level off. Long-term interest is going down. And we also see companies that started saving on the marketing budgets, that they are losing market shares or, let's say, losing revenues, and they need to, let's say, start doing something there again. But talking about sectors, there's a few sectors that are, let's say, not recovering at the moment.
That's more that we see, let's say so startup, smaller startups that used to spend good marketing money. Crypto was a segment which we, we were not so strong, but in the industry, did quite a few things. It's not so spendable anymore. And also hyper-casual games, which were, let's say, based on buying ad and sell... Let's say, buying users via advertising and selling, ad and make money with that. Those are the sectors where we don't see light at the moment, but we see more light on the, how to say it, especially on the, on the consumer goods sectors, where people are really starting to, further invest again. So e-commerce, these kind of sectors, where we really see, budgets going up at the moment.
So-
Yeah, sorry. Is that answering your question? And then I would go to the next one.
Yeah, yeah. Just, just... It's nicely confirming my hypothesis that, well, this very in or excess inventory-driven market in 2023, well, led to a general reluctance to advertise. And with more fresh products in many industries, advertising is coming back somewhat.
Yeah. Yeah. No, that's what we see. And let's say Christmas, especially, you hear, you see it also in the press, companies really saying they need to start advertising again, and they, even though people don't expect a great Christmas season for, for a lot of industries, but they still expect consumers. Yeah, they need to trigger consumers by advertising. So that's, yeah, I'd say it also what we saw in, the Black Friday and things, especially in U.S., that there is more willingness to, to do a bit. Then on your question, second question about, in-app as, as a segment of advertising. What we still notice is there are still a bit of prejudices there, like, in-app is gaming, and gaming is, young boys, yeah, eating pizzas.
So in that sense, there is a bit of market education to do there, which we are working on, because with in-app, we have a very good household reach, and basically everybody is a gamer nowadays, but also everybody has apps, and it's not only game apps that we talk about. We talk about weather apps, we talk about news apps, we talk about dating apps and many, many more. So in that sense, the reach is there. It has a lot of to do with the education. And yeah, as we all know, mobile phones are now so good that you can watch videos and everything at high quality. So it's something that takes time. Marketing managers tend to be a bit conservative in adopting new things. That's what we have seen here, but it has been seen before.
So we are really happy to be in this segment because it, on the other side, means that there is a lot of growth potentially. So-
Maybe just to add, Philipp, one or two sentences also on the demand side. So we're also seeing with the strategy which we were driving over the last years, where we invested a lot into demand, into our demand side segment, that we're now really getting also bigger advertisers, direct contact, et cetera. All this takes a lot of time, but also there we see some good traction, et cetera. And I think that we can also expect that especially also the demand side segment will further grow, also in the coming quarters, on the basis of new customers on board.
That sounds good. If I may, one housekeeping question: Can you provide the number of cash earn-outs that you paid in the third quarter?
So we basically paid EUR 10.2 million to AxesInMotion in the first nine months. EUR 5.2 million out of that was in the third quarter, and then there was another EUR 1.6 million earn-out for another small acquisition from the past, which we paid. So in total, in the third quarter, EUR 5.2 million + EUR 1.6 million cash earn-out.
Sounds good. Yeah. Thank you very much, and all the best for you in the remainder of the year.
Thank you, Philipp.
Thank you very much, Philipp.
As a reminder, if you wish to ask a question, please dial star five on your telephone keypad. There are no more questions at this time, so I hand the conference back to the speakers for any closing comments.
Yeah. Thank you very much. Then we come to the end of this. I would like to thank all investors, all bondholders and all stakeholders for their interest in our company. And yeah, 2023, as said, is not the easiest year, but I hope that we were able to make clear that we are really gaining market share, and that at the moment, that CPMs and budgets are coming back. This should give us a lot of tailwinds instead of the headwinds that we see now. So thank you all very much, and of course, Paul, myself and our investor relations people are all available to answer further questions. Thank you very much.