So, Felix, you've spent years working with biggest game and app developers to help them increase their ad revenue by designing ad placement, negotiating with ad networks and improving mediation ad stacks. You worked on mobile DSPs and ad networks, so you know exactly how the demand side works. And you've probably gone through hundreds of mistakes, failures and complicated situations as a highlight. And that has made you into the beautiful man that you are today, which is not like

me. So, Felix, welcome to the podcast. Thank you, man. You know, compliments get you anywhere. So, thank you for that warm introduction.

So, I wanted to get you on on the podcast because I want to understand in-app ads and where we are today. And I think I had a couple of stats in my presentation that I was I've been presenting to to a few people. And I think these are from, I believe these stats are from Unity. So, there's two charts. One is why in-app ads matters. And it shows all the genres in the game, in mobile games, from action, adventure to trivia, word, puzzle, racing, role-playing simulation, you name it. And it shows the distribution of in-app ad revenue percentage and in-app a purchase revenue percentage. And the lowest percentage is in RPGs. And even RPG games make 20% of all their revenue through ads. And then you go, of course, to the 95% from hyper casual, the 80% for word, 70% for board games, et cetera, et cetera. So, this kind of shows that no matter what kind of a genre, no matter how core it is, it still makes a fifth of its revenue through in-app ads. And the second chart that I've had in my presentation was the sort of an anti-notion to the thinking that ads may create a worse experience or inferior user experience for gamers. In this chart, it shows the reward video ad completion rate by genre again. And there you can see that the typical genres that we consider the casual ones, the hyper casuals and so forth have a very high completion rate of 80%, it's percent and up. But then we go to the traditional core genres like RPGs, like strategy, et cetera. And there, the ad completion rate is actually over 90%. So, clearly, users are enjoying the ads or at least enjoying what the ads give them. And then the final chart that I had was when you ask core gamers, console gamers, PC gamers, mobile gamers, do they want to get rid of product of ad product placements? And majority of them, far larger side than those who say that they want to keep the ads there. And it's a far bigger percentage than those who say that they want to get rid of ads. So, this is kind of my background of why I think in-app ad revenue is valid. So, did I get this right, Felix?

Yeah, I mean, you still got to take it with a bit of salt, right? Because I always say that in-app ads are still kind of the last bestial of the Wild West. I've described it earlier, is if you've ever played Red Dead Redemption 2, you know, it's set right before the end of the Cowboy Age. That's kind of the age we're in right now within app advertising. So, these stats that you read out, they're fine and dandy, but they're still made by Unity, which has an interest in getting people to use more ads. So, if we just look at the charts, first of all, completion rate, it's a bit deceptive for them to publish this, I think, like that because to get out of rewarded ads, you actually have to close the app because there's no X button until the end of it. So, you know, that requires a lot. You really have to be quite angry. So, it's quite hard. But it's still good indications of like the overall sentiment. Generally, people like rewarded ads and it's kind of touted as this silver bullet for the fact that if you have the best game in the world, you can might get 3 to 5% of people that make an IAP. The rest of the people that you have or the gamers that you have, it's just costs, right? Server costs. So, making them see rewarded ads is kind of on paper the silver bullet. So, in practice, usually the biggest issues I find with rewarded ads is kind of the content of rewarded ads because it's kind of a race to the bottom right now to get low CPI's with shocking content. Everyone's seen like the Trotius CP, like the content that you sometimes get. And then also the end cards have been getting worse, right? So, what a lot of people don't know is you need to kind of manage everything, right? And you can, but it just takes a lot of effort. And if you do all these things and we can get into other parts and best practices later as well, you're able to get quite a good ad experience and reduce complaints and then everyone's happy. And I've actually worked with some clients where your remover rewarded ads and people get pissed off. That's a good sign. That's a good list, right? But also going back to what you said on RPG. So, I worked on Legendary when I was at Network and I helped increase the ad revenue overall from 12% to 18%. And that was one of the things where like every all the developers were really not happy about it. But like in the end, we managed to do it incrementally without reducing any of the IAPs. So, I know it can be done for role-playing games. And if it can be done for role-playing games, it can certainly be done, I think, for pretty much every genre. It's just about taking the time and kind of understanding it and kind of understanding the regular pitfalls. Yeah,

yeah. That's valid points. Yeah, and thanks for not debunking but adding a little bit of a color to these ad platform-based surveys. But I'm curious to understand what happened in 23, like during this year, what has been and more particularly like what has been the privacy change on in-app ads? So, the privacy

change on in-app ads, it's I guess you're referring to the Google CMP and the loss of the IDFA, which was a bit earlier. Is it that kind of what you're getting at?

Yes, IDFA especially. Yeah, so the IDFA,

everyone's been touting as the ad apocalypse, which it kind of is, right? Because if you lose the ability to target users, all ads essentially become branding ads. And in theory, the ECPM should then fall quite significantly because you can't do highly targeted ads. That hasn't happened yet because iOS is still allowing or kind of like turning a blind eye to something called fingerprinting. What is fingerprinting? Fingerprinting is essentially when you use statistic methods to kind of deduce and do targeting anyway, but it's not individual targeting. Funny story is that we know exactly that all networks are doing fingerprinting. Even Apple is saying it's not allowed, but everyone is doing it. And there's proof on this because essentially when Apple first announced that they were going to get rid of the IDFA, and basically no fingerprinting was allowed, ECPMs on iOS dropped pretty much overnight by about 50%. You could see across the board and all the networks dropped and it was pandemonium. It was like everyone's like, yeah, here we go. Here's what everything what Chris has been talking about. This is a we're all going to die. And then the interesting thing started happening, like networks started like the ECPM started improving and the delta between IDFA users that have consented and that didn't basically went down from about 60% down to 18% about three weeks after. So what happened then? All networks essentially turned on. Yeah, fingerprinting. And that hasn't changed yet. So right now we're in this kind of really big transitionary phase where a lot of things are happening within app advertising. So it's a very big moment of change. A lot of things are happening. So IDFA is one. So no one really knows when Apple is going to get rid and start of actually like enforcing fingerprinting like to get rid of fingerprinting. And that will have a huge effect on the CPMs on Apple on iOS. So no one knows when this is coming, but the second they start enforcing it, it's not going to be great for ECPMs on iOS. Personally, my theory when this will happen, and then this is pure speculation, right? Because it's down to five people who decide this probably inside of Apple. And I don't even know who they are, right? But I would love to take them off for dinner, you know, brother with a good food and good brandy. And then you could find out, right? So if you look at it, I think from like a game theory standpoint. So the app store right now for Apple is responsible for 10% of their overall revenues. And all of that revenue comes from IAPs, right? And Apple have their own way of doing UA, right? In a targeted way, which is Apple Search Ads, right? Problem with Apple Search Ads is three placements. So you can't really scale as a gaming company. I've never heard of a gaming company being able to spend more than four or five K a day. And you know, for like a take two or anything like that, that, you know, if you can't spend at least 100 K a day on a platform, it's pretty much irrelevant and it's not worth your time. So that's an issue, right? So I don't think personally that Apple will remove this fingerprinting until they have sorted that issue out for Apple Search Ads. It's a bit of a going against the controversy, but I don't think they will shoot themselves in the foot, because essentially then no one can do UA on Apple to find these high value users who make all these IAPs. So I don't think that's going to get killed until Apple Search Ads either makes an audience network, which makes sense because they've been advocating for or hiring for roles for a DSP, which is kind of the key component where you need when you're starting an audience network. So that's what I'm thinking to watch out for. So I don't think fingerprinting and the whole IDFA thing is going to get killed until Apple have launched their audience network.

Got it. And then Apple would would be in a pretty dandy place because they have Apple Arcade, where you can get all your games. And then they have their own audience network, their own advertising tech, which is the search ads. If you want to play the free to play game, give us the money and continue

playing. When they're looking at Unity's revenues, Apple Evans' revenues, Liffdolf's revenues saying like, hey, gee, that's a lot of spend. Would a shame if something bad happened to it, right? Yeah, a lot of spend that we don't get a lick of. That's the problem. Exactly. I mean, you know, buy some Apple shares right there. I think they're going to pull it off. It's a close ecosystem. I always buy Apple shares. They're the real gangsters in the course. Don't only go up. By the way,

I wanted to get some basic terms right. What is a DSP?

Demand side platform. So what we can talk about. So I've worked in two DSPs in an ad network. So I know how the sausage is made. So DSPs is a demand side platform that buys everything programmatically. I worked for a two called Reemerch and Addictive. They're retargeting DSPs. So yeah, they bought mostly banner inventory. But then there's also something called networks. And they have their own demand. So when you go to Apple, and you buy Apple, campaigns through Apple, that's through their direct inventory. DSPs, they buy all their inventory from an SSP or a supply side platform. And that's where they buy all the inventory. So it's a DSPs and networks are the two main demand sources that you get when you hook up ads into your mobile game or mobile

app. Understood. Kinda. So when I get when I like when I open up my Unity ads to run through my game, let's say I'll take Unity ads and I'll

take the app

love and ads.

It's more complicated. So like I said, right? So it's in app advertising. It's a bit like reading in the middle ages. You either know it or you don't. And no one really teaches you it because it's not that many people that do it. But it's way more complicated. So for example, if you work with IronSource, let's

say, for example, Unity, yeah,

Unity, but IronSource has the network. So let's say Mishka, you have created another game and you want to buy some users for that game and you have decided let's go to IronSource. Let's buy some. Let's buy some. Of course. IronSource all the way. I was there. Quite

good. So yeah. And then I'm not some let's get Luna Labs for for our creatives. Right. We're good. You

got your Luna Labs creatives. And then you mix IronSource and you get your credit card details and you're buying users. Right. So I then have the app. I do the add monetization. I see the impressions come in in my app or game. And then I'm getting users. It looks like IronSource, right? But to ensure that they get a lot fill, like what you also see is. So typically, if you're seeing $100, let's say a day from IronSource, maybe 40 to 30% of that comes from IronSource directly because they can also resell inventory because they have a direct link with a publisher or publishers where you call an app or game that shows ads. So what they can also do is use their piping to sell that to third parties. And that's what you see in the add-ups TxT file, which you need to list when you start having ad revenue. So it's not only IronSource demand directly that you're getting when you plug in IronSource, it's also their third party demand, which is a list of exchanges, DSPs, like a whole myriad of different places. So everything is very complicated. It gets really complicated quite quickly. But yeah, the fundamental is simple, but the practice is a bit difficult. So for a layman listening to this, don't worry about

DSP. What you need to worry about is you set ECPM. And the way, okay, CPM is cost per 1000 impression cost per mill. And CPM is important because when you divide CPM by IPM, IPM being installs per 1000 impression, you essentially get your CPI, which is the cost per install. And that is one of the most important metrics in your game. And CPM, the way I understood is, is basically how expensive or cheap it is to advertise towards the audience, towards the audience that you're looking for on a specific platform. And then IPM installs per 1000 impression means how cool your creative is, how well it hits with the target audience. And so when you divide that sum of how expensive is to advertise to what is your knockout percentage or how many of 1000 impression actually installs it, you get a lower CPI, the higher the IPM is. I understand that. What is an ECPM? So effective CPM, right?

So effective cost per mill. Effective cost. Let's just run through like the basic terms of modernization because ECPM gets by far the most attention. And it's not the most important metric by a long shot. So usually you can tell if someone understands and knows what modernization is, depending on how much or little they talk about ECPM. So if you're talking to a network and you keep complaining about ECPM, usually the person working at the network will think you don't know anything. And the reason for that is you care about ad revenue. Ad revenue is the most important metric. And ad revenue is a component of two things. That's fill and ECPM. And those things together bring you your ad revenue. So I would gladly take a lower ECPM if I meant I got a higher fill and a higher ad revenue. So the metric that we use are Northstar our load star or poll star, whatever you want to call it, is a poll star. And ad revenue per daily active user because that's the only metric that you can look at in ad and modernization that takes into account the fluctuations of UA. So ECPM is just a mere component of that. And ECPM, like you said, is pretty much what the network is willing to pay for your inventory. So how does a network decide how much they're willing to pay for your inventory? So ECPM is calculated by all networks in different ways. They have different algorithms to do this. But if we're doing it pretty simple roundup of how it works, they take the average of the last seven days on a bunch of metrics. Google uses about 84 metrics, taking into account like time of day, like even the weather, like where are they doing it. But in principle, like the main metrics that they look at is the amount of unique users, the amount of like times the user has seen an ad before in that day, the install rates on the average like user in that app for the last seven days. Also, they look at like the click through rates and a bunch of stuff like that. But that's the main metrics that you look at. And you take the average of that over seven days. Another big metric that they look at is the ad network bring that user to that game or app where they're now seeing it in the bidding. If they did, they usually have that tagged on their server side. So they pay more for it. So let's say you brought a user via app loving to your game, app loving then sees it on the admin side, they will be higher for that because they know more about that user. So that average is calculated on the last seven days taking into account all those components. Like I said, Google uses 84. I'm sure they every network uses as many, but those metrics that I listed are the main ones that matter the most. Got it. That's how eCPM is calculated.