Latest / Mobile Signal / Can Game Publishers Build the Next AppLovin? Tencent, Voodoo, and more are building.

Transcript

Can Game Publishers Build the Next AppLovin? Tencent, Voodoo, and more are building. · Mobile Signal

• Josh Chandley: Mobile publishers are quietly building their own ad networks, more than most operators realize. Today, we're going to ask whether this trend matters. Welcome to Mobile Signal. I'm your host today, Josh Chandley. Joining me is Vincent - 52E.: Vincent is here, fifty two entertainment. Warren (Upptic): Warren Woodward, co founder and CGO at Uptick. Josh Chandley: Let's get into it. Let's start by taking an inventory. Which publishers, Vincent, are known to be toying with their own ad network? Vincent - 52E.: Yeah, good question. I mean, if you think about it, every large ad network was originally, you know, a publisher, right? And then you have the new kids on the block, you know, the new networks who are coming from being a publisher in the first place. So I mean the one That is quite important, is obviously TikTok. TikTok is one of the biggest audience and app. They launched Pangle very quickly. They actually started serving ads in the US. It was a big announcement earlier this year. And then top of mine, you know, I'm French, so I have to mention, you know, one of the biggest publishers in the world coming from Paris, Voodoo, is playing around and towing and building, you know, their network silently for some time. And so it's absolutely worth mentioning. and finally, from Asia, Bigo Ads is originally a publisher and they managed to grow their network, be certified, integrate their SDK to actually becoming a solid player now. So that's I think the main one. and then yeah, we can talk about smaller initiative, but for me that sets the scene. Josh Chandley: Yeah, I just throwing some more names on the pile. I think the that Felix Brayburg also on LinkedIn was posting about a ten cent in app ad network. so there's there's another one there. and there's even questions about like what an ad network is. I know Vincent, you're probably gonna disagree with me later, but I would count just play, which is owned by Nazara, formerly Blue Tile. And I I would consider that a an ad distribution network, not an N app ad network, but an but an ad network. so that's a lot of best play. You're right, best play. Vincent - 52E.: Best play, best play. Best play, best play. Josh Chandley: Best play. That's a lot of ad networks. what's the prize here publishers are all chasing? Warren (Upptic): Yeah, so when we think about the prize and why a lot of these publishers are trying to build their own ad networks, like if you if you were in this business and in an executive staring at this problem, I think it's inevitable that this is the conclusion you come to that you should you you think you should do. it doesn't mean that that's correct, but to kind of just talk through it from, you know, my perspective on the UA side of things. So like the UA ecosystem today is it's w wildly inefficient. You're having, you know, if if Josh, if you're a developer and I'm trying to buy your inventory, we have but sitting between us, there's a mediation layer, the ad network itself, a DSP taking its own technology fee, and like this is gonna add up to you know forty percent or more kind of arbitrage between me and you trying to buy and sell inventory. So this is like a hat on hat on a hat. And so by the time this This dollar is moving through the ecosystem, you know, 40 to 50 cents of it is eaten up by these intermediaries. So any like C-suite executives like looking at how they can further monetize their businesses, they will inevitably learn enough to just enough to become dangerous with how this works and say, aha, we can capture this margin. This is obvious. We have the inventory, we can sell it directly, we just need A, B, C, and D. And I think it it makes sense on paper. But once you're really in the weeds of the ecosystem and been buying for a decade plus, like I have, and you know, I know you guys have massive experience here as well, it's just not that simple. And this is why we see most of these efforts fail. Like the the two kind of themes that I would identify, the first is just sort of underestimating the technology that you have to invest in. Like you're going head to head and simply owning your own inventory is a relatively small part of the equation of what makes an ad network. work at the end of the day. I am not lacking places to buy ads, you know, and you having another place for me to buy ads is not particularly helpful. It's like that kind of goes to the second point, which is the perspective that these companies are operating from when they're thinking of building their own ad networks to really oversimplify. A lot of it is like, hey, how can we sell our crap? How can we sell our under monetizing users? You know, we want to keep our best users internally, but maybe we can resell the rest. This is like the complete opposite from like a buyer's perspective. And what I what I would say would be a healthier perspective if a company is trying to spin out an ad network from their their own inventory is take a buyer's perspective of how can we create a UA product that is going to get a buyer, a UA buyer, a superior outcome, superior ROAS, then these other players in the space, then the app lovens, then the Metas. I have and w where I see them time and time again fail like. Hey, we have our own inventory. That's why you should buy it from us. and there's so many it it it really oversimplifies what makes UI work, what makes UA campaigns profitable. And I think this is why we've seen these themes of failure in most efforts that rhyme with this. Josh Chandley: Yeah, I mean, I think that the ultimate question about these networks, kind of jumping ahead, is it's a question of is the cost efficiency gain of having fewer people in that chain? is that superior to the targeting loss, to like the loss of efficiency from targeting? I think that historically the answer has been especially recently, no. which is strange because all of a sudden we're seeing so many new ad networks. Vincent - 52E.: Yeah, I absolutely agree on people underestimating the amount of technology you need nowadays and also the amount of supply. So again, we it's not a media plan with brands when you target contextually this and this publisher. You need to go and have your algorithm run on the maximum of supply and just target horizontally. So that's one thing. So even if you have massive in impression in inventory, you call me. I'm not gonna go direct with you, you know, because it's gonna be super restricted. And how are you gonna optimize? And so you need massive scale of inventory. And then obviously, I mean, so I worked a few years at Unity Ads, and I remember, you know, a slide on what was the logic. It was the time where Unity went from waterfall to bidding, and I have a slide in mind on all the steps and the checks and the servers to end up with the bidding price. So you need to evaluate, you know, the opportunity, how much I want to pay first price auction, right, for this device ID. You have no idea how it's complex and the number of variables you need to compute to put a price on that. So and that's an ad network. So hence, you know, obviously Aplovin sometimes talked about the quality of their engineers building action, right? And I believe that, you know, that is a tremendous complex thing to do if you want to compete. Because you're going to build an A network, then you're going to go plug it left and right. You're going to be competing with the best algorithm in the world. That's something I think people don't really understand when they are willing to build an A network. Josh Chandley: Yeah, but I think maybe maybe the threshold here i is lower than achieving unity or or app lovin' scale. You know, maybe maybe maybe that isn't the the threshold. you know, I think the When you when you think about these big ad networks, especially what's recurrent, what's coming up now, with gaming, what I think is different is that when you think about the Ten Sense, the Voodoos, these massively scaled companies, they're not trying to hold they are, but they're not trying just to hold the users in one game. They're trying to hold it in a whole portfolio. and so if they can efficiently just move players around that portfolio. That's a retention play that is differentiated than the original O builders of ad networks like Mike Meta. and so perhaps perhaps the initial bar for success is is lower. And that's where you can start to see like not just in app ad networks, but also blue tile now Nizara with best play, where I don't I don't view that as a way That is guaranteed to become like an externally sold ad network. They have a new way to acquire users, but a new way to move users around in their ecosystem and keep them. I think that that's a huge value creation. And those are ad networks in my views. so like what is could that be enough? Is that is that enough Warren (Upptic): Josh, I don't know. Josh Chandley: of a go ahead. Warren (Upptic): I don't know if it was a slip of the tongue, but you referred when you you said when you referred to Meta, you said McMeta. And I and I Josh Chandley: Muk meta. Warren (Upptic): think that's perfect in a way. I thought maybe it was intentional because it kind of directly points Josh Chandley: No, go on. Warren (Upptic): to this you know, the the reality of the ecosystem, which is like, you know, Meta is the Mc McDonald's of advertising in the sense of like their core business at the end of the day and their core monetization that's keeping the rest afloat. is this just incomprehensibly robust ad serving product, right? And when you're launching a new network, like that's who you're going head to head with at the end of the day. People don't care where they spend their money. They care on the returns that they get from it. and I think this has been the largest obstacle and and largest contributor to the fail rate of people breaking into the space. We look at companies like even like Chat GPT launching their ad products. Every anecdote I've heard so far is that it's absolute dog shit. They have essentially like infinite brain power, infinite capital, and they still are not able to make a quickly make a compelling entry into the space. When I look at the other side of that, of like who was able to effectively break into the space, this is old news at this point, but when TikTok really emerged in the ad the the UA space is a successful channel, why did it work? A big part of it. was they just directly poached key talent that decide designed the Facebook ads ads platform, the McMeta platform, as we'll call it from here on out. and they've replicated it as close to one-to-one as possible. So they were able to launch with relative parity of the strength of their ads algorithm and a familiar user interface. And that bar of what it takes to actually compete for my next dollar as a buyer, you know, and for if you guys as as developers with with your UA spend, Like this is sort of like how high the point of entry is in my view and like why teams keep failing. Vincent - 52E.: Yeah, something to add there is also brand or performance. So I spent some time at Twitter EMEA and Twitter had a DSP, right? and they had or they tried to build a performance, you know, to capture the app performance business. And it was very difficult. Slash they were struggling because the whole infrastructure, the metrics, the measurement. was all geared toward more brand oriented campaigns. So that's another component that I guess, you know, the Snap, the TikTok probably learned from. Because from the get-go, I believe they were targeting more and building a ad product which was more geared toward, you know, app install, being integrated in MMPs, having the right algorithm and optimizing against the right events and the right data points. Josh Chandley: Yeah, it's insanely hard. here's what I see that's different. Chat GPT has the meta problem, and meta's problem is the hardest problem in the world. Every single like how many, how many billion MAU do they have? All in the same two or three applications? That's like, you know, trying to look look down inside people inside a grocery store. And figure out okay, which one of these people in this grocery store wants to buy a book today? That's that's insanely difficult because everyone goes to a grocery store. I think what's maybe a little different about these game studios now is that the the applications that you're in are more niche and more specific. And so the maybe the bar is lower. You know, maybe if your Tencent or your Voodoo or your one of these networks that are being built inside your O inventory, your job is well, shit. We're in a bookstore. Who who in this bookstore wants to buy books? You know, maybe maybe that's a lower bar. and that creates Vincent - 52E.: Mm-hmm. Josh Chandley: like a an easier beach head to establish. Cause what I'm actually wondering is, you know, is the cost efficiency gain of removing that that one link in the chain in the ad network? is that a big enough of a hole? To to survive the targeting loss of bootstrapping your own ad network. If you're if you're trying to sell to other people, probably not at first, but if you're trying to move people around your own portfolio, I I do think a cross-promotion network is a is a powerful starting point and it does feel achievable at a certain scale, right? Vincent - 52E.: Yeah. for me yeah, it's a little bit different, but I get why you are, you know, putting them together. But yeah, I mean we all know cross promotion, as soon as you have, you know, various studios, you know, the Steelfront, the Embracer, or at fifty two we you know we have various developer accounts, you're trying to understand how you can circulate your users from one app to another and essentially pay one CPI and have multiple L T Vs. That for me that's my dream. And that's and that's I believe what know Blue Tile and Best Play app, the rewarded app, managed to do. Some other, you know, folks did as well. and from what I I understand, and it's an assumption, right? But the Voodoo network also, you know, started by hey, let's try to have some cross promotion. And then they realize, you know, actually we are actually building a network there. So those two topics are very much linked, and both are relevant. And I agree, before you open up to third party demand, just serve your own needs and your own studios and your own apps by trying to find a way to recirculate your users potentially within a network and with an ad, right? Josh Chandley: Yeah. And so what once once you accomplish that, I think selectively inviting ideologically aligned publishers or or publishers with similar users and games into this cross promotion network is potentially step one of admittedly many steps, right, of building towards a more generalized ad network. And that's a very different play than we've seen before. and so if you're thinking about a from a publisher point of view, building this SDK network, let's say. Is that a viable path or is the ceiling, you know, really fancy cross promotion tool? Like what do you two think is actually the the theoretical maximum accomplishment here? Warren (Upptic): I would love to maybe start this by turning the question around on you guys, because I have the one company here that's not, you know, shipping games. like at what point would you guys bother integrating with a studio, kind of like a small network studio like this to be part of that? Like at what point does that actually become compelling? Because it sounds like a a pain in the ass from both a business logistics, engineering standpoint to build something like that? Like what would the upside have to look like for for a developer or publisher to bother with integrating into a a a closed loop network like that. Josh Chandley: I think the big question is is is it or is it not a closed loop network? You know, is it this special thing off by itself, or is it part of like the max or the level play or the ad mob mediation ecosystem? Because from my point of view, if if they can get max certified, well then then you're just making a monetization decision. Can they create enough uplifts to be at least the worst, like the lowest performing ad network? worth keeping in your stack. and at that scale, it doesn't have to be APP or Unity. It just needs to, you know, achieve some meaningful share and not have a big A and R issue. Vincent, what do you think? Vincent - 52E.: Well I think, I mean, the problem with an network is you need supply, you need demand. So as soon as you're certified max, and good luck with that, what I'm what I understand is the goalpost is moving, you know, or the threshold is going higher and higher. Like anybody, you know, talking to Applovin, I think, saying, Hey, can can I be certified? You need to be able right away to spend X million per day. Right. So that's the threshold. So So you need that. And once you integrate it, you need the flow of demand, you know, for to serve, you know, and and and you need the campaigns, right? So once you have the supply. So it's always a chicken and chicken and egg game where one it's never balanced. You're gonna have more supply, not enough demand, or the opposite. So it's a very, very difficult balancing act. But for sure, once you are in the stack, it is a monetization player. You need to deliver at least, I don't know, 0.5% share of wallet, and for that. You need to really ramp up your demand and your own UA is surely not enough. Josh Chandley: Un unless you're unless you're Tencent, right? Vincent - 52E.: Unless you're a big boy, yeah. Josh Chandley: yeah. I mean Tencent, I I checked this morning. they o obviously they own MiniClip and they have their own publisher as well. According to App Magic, in 2026, which is far from over, they've already driven over one point two billion downloads. That's a lot. I I'd say that they have the demand and the supply from their games. Vincent - 52E.: Yep, I would agree. Josh Chandley: I mean, is that but is that so large? That's potentially a top ten or twenty network even with just cross promotion? Vincent - 52E.: Warren. Warren (Upptic): Yeah, I mean short answer. Maybe. I mean, what is the upside though if if they if if the if the end game is is cross promotion? To me it's like not it's not a very sexy like end state. that being said, like despite, you know, all the prior skepticism that I've expressed here about moo people moving into the space, like the kind of the parent entity of Tencent and all the sub brands are likely like tip top of the food chain, of folks that could make a play in this space. The missing piece and the secret sauce, not to keep beating the same horse, is the ability to not just having the traffic. Again, having the traffic is is only one leg of the table. Why are you giving a better result to a buyer to where they're gonna spend their next dollar with you? And have they, you know, we're talking about going to head to head head to head with you know multiple of the Mag Seven companies whose core products are serving ads as far as like how they build build the revenue. So to some degree, Josh, you know, you touched on a good point later, which is like, well, if you if you have better product market fit with your buyers and seller buyers and sellers, with your bookstore analogy, maybe you don't have to be as good with the level of what your algorithm is. And I think that's true to a degree. for any of us that have been doing UA for some period of time, if you're buying from a network like Apple1, Unity, et cetera, if you look at the publishers, inevitably There's a spew few specific publishers that are delivering the bulk of the traffic. And that's sort of like your ideal product market fit within their thousands and thousands of apps. so there is to some degree the bar is lowered if you are making a dedicated gaming network and you have worse technology by a factor of you know, factor of two to ten. But at at some point, All of those same users are still present in the Google system, in the meta ecosystem. And I can't stress again like how good these algorithms are by these Mag 7 companies at uniting buyer to seller. And so I think that's the unknown is like how much worse can you be if you have kind of more ideal product market fit with your buyers and sellers of you know, of what a Tencent or a Voodoo could ship as a cross cross promotion building out into a outward facing network. Josh Chandley: The the classic answer is always how how do you compete with with Meta and Google? You compete by being faster and more agile and more niche, right? And that's how you establish a beach head. That's the default question and default answer of like every VC when you're doing a pre-seed around. Is like, why hasn't Google done this? How are you gonna compete with Google? Or if you're doing a game, like, well, how are you gonna compete with Tencent? And that and that's the answer. And so I wonder like what Does that look like? yeah. Warren (Upptic): I if I can, real quick, Josh, I think there's also a more crass answer of how you compete, which is just you just eat it on margin. You know, when I think about like early in my UA career, when we would be like a top three buyer for a few months on Iron Source or on Vungle, you know, name your network, not not saying anyone not say anything specific. But we would always know, sometimes transparently so, that there was smaller plays in the space. just eating their margin, selling at a loss sometimes because they're trying to claim market share. And the so you know, the the euphemism of being more fast and nimble, a lot of times that's what it turned into practically. They had a worse product. But if your competitor is taking what is the phrase, like your my my my competitor's margin is my opportunity, right? And that's how we saw, you know, early in my career that our spend would move to these smaller players. Often it was just something that that simple and crass. Like Hey, we're just gonna eat all of our margin so you know we can grow our top line ahead of this MA we're trying to do or ahead of going public. Vincent - 52E.: I agree, but if yeah Josh Chandley: Yeah, no, I think go ahead. Vincent - 52E.: No, I just want to throw an analogy there. It's a boxing match. You know, you enter the ring, how many rounds can you sustain? So sure Warren (Upptic): Mm-hmm. Vincent - 52E.: I you can put your margin to zero. How long can you sustain without oxygen? Right. So for me it's it's really a timing and a cash flow game if you play this like, hey, I'm gonna attack this market with very, very low margin. And honestly, it's not very difficult to be lower than than Aplovin, right? But how long? We're gonna survive. Josh Chandley: Well, I I Warren (Upptic): Well set. Josh Chandley: I think the app level analogy is interesting. one company that's a pure ad network, I would bring up, is Moloco. They were only a DSP for a very long time. And one of the keys to their scale is what they called their zero margin SDK. Essentially, their SDK just they just took zero margin with it and all the pricing gains by having an SDK versus buying through indexes. They just pass that directly on to the buyers and that enabled them to scale. And so I think that's the story of how cross promotion can be very beneficial. and I'm wondering if, you know, if you're one of these large game studios. I think have operating a network at breakeven, if it means that you can increase your UA scale by 25% and your UA scale's margin stay the same, I think that's a winning bet. The next question is, well, do you consider supporting the the studios close to you at a low margin, like a low take rate, so that you can start to gain additional penetration? Do you start to try to get your SDK implemented into similar games so that you can buy their inventory? You know, is that like a step that becomes viable? Because I I agree that jumping straight from cross promotion to full scale, like fully scale demand, fully scaled supply, that's a chasm. But but like, can you break that down into bite-sized pieces? I maybe. Vincent - 52E.: Maybe, yeah, it's a good thinking, you know, you're just subsidizing your main business with a zero margin network, just to ease your CPI, help your ROAS so you can spend more. And if you have footprint, maybe you can do some step by step there and then you reach a ceiling, you already improve your business and you don't even go. You don't even make the decision to go into the you know big swimming pool with the big boys and and launching a network full scale with third party demand. So yeah, it can be a play. It can be a play just within your scope to help the business. Josh Chandley: And that's similar to what Pangle did, Warren (Upptic): Yeah, it's it's really Josh Chandley: right? Go ahead. Warren (Upptic): Pangle. Yeah, we I mean, don't get me started on Pangol. The Josh Chandley: I'm gonna get you started on Bangle. Vincent - 52E.: Let's do it! Josh Chandley: I mean that's a pl that's a play, right? They they they had Warren (Upptic): Mm-hmm. Josh Chandley: O and O inventory and now they're finding more ways to to express that inventory. but by accessing more demands. Warren (Upptic): I'm always yeah, I mean I'm quick tangent here and maybe tactical note. I don't know what your guys' experience has been, but I'm I I try to be relatively Swiss and neutral in the UA ecosystem because we have to do business at scale with a lot of these companies for for our clients. but I will say like in my observance, Pangol seemingly wholly exists out of just you know, essentially ignorant users that are not aware they're buying Pangle traffic. And in I'd say just about every case I've seen, the anytime we've drilled down into like what is the performance of Pangle within our TikTok campaigns, it's just been absolute garbage. And the the the it it's largely subsisting just on people not knowing that they need to manually opt out of it. So I think on one hand, has it been a a quote unquote success story of like bolting on an ad network of external apps and monetizing that? Yes, but it is not healthy or sustainable and is just basically living off of people not really paying close attention to what they're opting into by default. So I don't know that we should be taking notes from it as an industry in the current state. Josh Chandley: So so so you think this is an anti this is an anti sell. you're concerned that Pangle is like dragging down TikTok performance. Warren (Upptic): it absolutely is. Like at least you know, w the with Josh Chandley: Okay. Warren (Upptic): the caveat that our bias in our gaming portfolio is more mid core. I would love I'd I'd actually love like anecdotes from you guys if you had similar experience 'cause you guys both I think have more casual gaming audiences. So what has been yeah, Vincent, what has been your experience with with Pangle? Vincent - 52E.: No, I mean necessarily the Pangle network will have different performance than TikTok, right? Obviously. Pangle is an SDK network, you're gonna run into games. TikTok is a social platform with a very unique experience. But again, for me, the blueprint is Meta. Meta did an amazing job to, I think, from from looking from the outside, to manage to keep advertisers happy, but essentially they're all opting to run both in O and O and third party. Again, I was a time Twitter when you know there was something called Twitter Audience Network. You run a Twitter campaign, which is very specific on the Twitter feed, and then you know, opt-in by default, opt out by default, and then you're gonna run on mop up, on some gaming, some non-gaming, the brand and you know, different objectives. So it's a very different inventory, and then you have your product, your ad product needs to be very well tailored if you have a front end, so the client. Click or don't click, you know, to run on your third party. And you have to have a super good algorithm to to ease and average out the output, the performance. Because this is gonna this is gonna be very different inventory and performance level. And your job is to make it smooth, average out the performance, allow the advertisers to opt out if they don't want to run on your third-party inventory that you have no idea what it is. So it's a very fine balance to find and I don't know if TikTok, you know, did as as a good job as Meta on that. Apparently, Warren, you think the answer is no. Warren (Upptic): They they have not. In fact they I think they've Josh Chandley: Ha ha Warren (Upptic): just recently rebranded Pangold to some to something else, I think because it's starting to build a bad reputation. Vincent, you you touch on a I think two points that are relevant to this larger discussion. I mean, w to some degree, you know, we're we're a fairly, I think, niche podcast here. Like we're a subset or ni a a niche within a niche industry and trend go deep in the weeds here. But I think something that maybe people newer to the space of mobile app or mobile game performance marketing, need to note that Vincent is calling out is just about every ad network out there, you are buying a bunch of other crap beyond the thing that you think you are buying. when you are, you know, when you're buying from Twitter ads in you know two 2012. You I think, you know, a naive, like early in their career, person thinks, great, I'm buying inventory from Twitter. No, you're buying all this other stuff as well. When you're buying TikTok, you're buying Pangle. What the hell is Pangle? I've never heard of it before. It's usually the process for teams discovering this. and so it's just like a practical tip of like get a nuanced understanding of like, great, I am buying through your ad network. What are the different streams? and this ties to the broader theme of the the show today, Josh, of like building an owned and operated network from you know, in-house from a publisher, even these leading ad networks have found shit, we cannot float a whole business just off of what we own and operate. We need to plug into, you know, we need to plug into these to to Apple into Unity to these DSPs and be reselling this as well too and kind of bolt it on and have dark patterns in our UI. Do people have to opt out or don't know that they're buying it? And I think this is kind of another layer of why these owned and operated networks built from publishers have had a tough uphill battle. Josh, what are your thoughts? Josh Chandley: Yeah, it's true. I think like even the even these other networks too, they may sell their inventory through and maybe demand flowing into that O and O or even through to that next exchange that you don't know about. I think the the the most a more charitable description of the same thing might be that when you look at TikTok's gonna hate me. when you when you think about TikTok and Pangle, maybe it's that In app ad inventory in general works really, really well with like casual and ad monetized games. And social media, like Meta, Google, TikTok, they work really well with mid core. And so one of the challenges Warren (Upptic): Yes. Josh Chandley: with the audience expansion is okay, cool. So you're successfully now integrated in this new place. Is that actually helpful? Is that actually gonna expand? Yeah, demand that your your your current or inventory that your current demand actually wants. Vincent - 52E.: I think at first you have no idea and just experiment, right? You make the decision to expand your supply and then you try to make it work with your algorithm and have your advertisers, you know, be happy with that. but yeah, true. It really depends on your type of campaigns, type of advertisers, because yeah, it's all about finding the right users, but also in the right context, right? So that is a a key thing, I believe, for anybody making a a call or a decision to invest in building their own network. Warren (Upptic): This this reminds me of something that we used to always talk about on my former team that we just called the My Talking Tom problem, which was like every time we would launch a new ad campaign on a gaming ad network, they would always kick and scream like, No, you need to buy run of network. Like it's really best practice. You gotta buy, run a network first Vincent - 52E.: Ha ha. Warren (Upptic): and optimize down from that. And we would be like, We know for the last hundred times we've launched campaigns, these Josh Chandley: Why do we want to waste our money? Yeah. Warren (Upptic): like same five my talking tom ads. apps like rise to the top, eat our budget, and never do anything for our game because we've got hardcore RPG strategy games. Like no, can you please just trust us and let us like we don't need to learn for the hundred and one time that we need to block these f five my talk and Tom apps. and I think that's kind of what you're speaking to, Vincent, is like when you're going into these expanded networks, to some degree what you're buying at the end state, you've already learned several times over on other networks. but because we don't have the transparency to know. Okay, when I'm expanding into Pingol, I'm getting these apps. When I'm expanding into the DSP inventory of an app level and etc., I'm you know heavily getting these apps. We have to relearn these things and it's eating our budget like with each each incremental new test. Vincent - 52E.: Yeah, I mean on this topic, I think shout out to Google. I mean again, Google with the universal campaign was exactly this kind of conundrum of hey, we have all those properties, you know, search, ad mob, YouTube, Google play sponsor leaks, trust us. Just give us your goal, we're gonna figure that out. What is the right product inventory on our end? You don't even have a proper breakdown of where you add RAND, right? But it works. And you know, you have the Google stamp that okay, it's gonna be Quality inventory, they have you know check and balances in terms of you know shady publishers or fraud. You trust Google. I do trust Google, and then it's just like, hey, this is my ROAS, go ahead. And honestly, Josh, I've listened to your very interesting podcast with DT. You had this kind of conversation between, you know, the app growth, the SDK network, and the preload, very different retention curves, very different product. And hey, do I want to combine that into one big delivery and trust DT to find the right user for the right ROAS? Maybe. But it's a it's an ad product and front end kind of question. But again, Google, I think, to my understanding did it successfully, but it I think it's super hard to do. Josh Chandley: It is. I think like merging them both together is is difficult. And that makes me wonder about, you know, Voodoo. So shout out to to James Eau Claire, his product App Goblin. It does it does show what could be a voodoo ad network slowly spreading. I'm not sure. But one thing that they have that is gonna trigger Warren's like TikTok Pangle vibes is they have B Reel. is that a potential and like additional demand mechanism that could lower the bar for them to be successful? Like if they already have this demand flowing in and then they have their own gaming demand and gaming supply. Is that a is that a good combination to bootstrap a an in-app ad network with? Warren (Upptic): I mean my inclination is no, like I I I just looking, pulling in the background, just kind of B Reels growth levels, like that app has been fairly fairly dead for some time. So I think it it it it's gives some curb appeal, right? We also have this like owned and operated social network, but I think practically it's not super material. yeah, and like I'll echo you a shout out for for people not familiar with app. App Goblin, built by our friend, our friend James, like great free tool for research of like where ads are serving. and yeah, definitely a great free product that you can check out an open source research research engine. Vincent, what are your thoughts? Vincent - 52E.: yeah, I don't think B Real can be instrumental in helping, you know, a network taking off. I think you're being a little bit harsh on on B Real Warren. It's not a zombie, it's not dead, it's still like a a product a lot of users love. I don't use it personally but Warren (Upptic): Are you guys avid B roll users? Am I the only one? Do you know a B roll user? Vincent - 52E.: No, mean I'm not the target audience. Josh Chandley: No. Vincent - 52E.: but I think you know it's still it's still it's obviously an interesting MA from Voodoo behalf. But to answer Josh's question, I don't think it can be instrumental to really boost the network, in my opinion. Josh Chandley: So then the the the real question is can these networks escape cross promotion purgatory? Vincent - 52E.: Is a very nice way to put it. maybe. I don't know, maybe, but again, you're gonna need to throw a bit more resources when you're gonna jump into the big swimming pool with the big boys. meaning the meta, the upload in, the ad mob algorithm, right? so I'm I I will not bet for sure that you know the next step is opening up and go beyond cross promo personally. Warren, what do you think? Warren (Upptic): Yeah, for me the the philosophical issue that keeps a lot of these networks from success, you know, you're you're in that cross promotion purgatory that you referred to, Josh, and the thesis is great, which is like, hey, when we got a user monetizing, as you said, Vincent, I want them to jump between all of the games or apps in my ecosystem. I wanna monetize them four times, five five times. What these teams don't wanna do is then sell their most valuable users to their competitors. So, you know, back to something we're talking about earlier, usually the thesis is, we've got all these non-monetizing users. We can sell them via ads to other people. But eventually people like me that are buying that inventory is like, hey, it looks like no one is spending any money from this ad inventory that I'm buying from this new ad network. and so I think that has been like the philosophical wall that has kept a lot of these teams from success. Like you can't say I'm only going to resell my crap and then be surprised when no one wants to buy your crap non monetizing users. Josh Chandley: No, and if you and if you really want to integrate it has to be through max level player ad mob, which means you have to you have to win that integration. And so Vincent, I have a I didn't ask you this before, but I'm gonna ask you this now. So enjoy. Ten cent voodoo like pick whichever big a whichever big publisher it is, what kind of uplift would you need to see from them? To be excited about integrating their SDK, assuming it authorizes a bitter in Max, let's say, or your mediator of choice. Vincent - 52E.: Yeah, I mean as you know, yeah, I mean it's obviously a very interesting question. But as you know, as soon as you integrate a new network, are you gonna make uplift incremental revenue or is just cannibalization between networks? And then the next question is okay, well, they need to make at least I don't know, one, two percent of my monetization, otherwise just useless code that is, you know, polluting and generating A and Rs. So each time I'm adding a new SDK network, do I see a huge uplift, you know, five percent of my APDA? No. Is it pushing prices up with better bigger bit density? Probably. and then it's just cannibalization between networks or you know, eating the cake of of another. That's my short answer. Sorry? Josh Chandley: how much uplift would you want to see yeah how much uplift would how much uplift would you want to see to your arc doubt like three percent five percent what would what Vincent - 52E.: I want I want to talk double digit, double digit uplift, Warren (Upptic): Yeah. Vincent - 52E.: right? And then I integrate tomorrow. Josh Chandley: shit. You would need you would need a competitor to drive ten percent plus ARP Dow lift for you to be all right. And and why yeah, and and w Vincent - 52E.: Probably. Yeah, yeah. But yeah, obviously it's double digit APA. Yeah, it's it's a dream. It's a dream. Josh Chandley: yeah, but why why is it so much higher? Because I bet you if it was not a competitor, you wouldn't say ten percent. Vincent - 52E.: Yeah, I mean I don't know, but it's just you know, say just a way to say instrumental, you know. obviously one two percent, I take it, right? it's pretty rare. adding just a new source of demand and having right away, you know, your data team saying, Hey, it is sure, it's certain, I stamp it. This is because you added a new source of demand, you have plus two percent of ABDA, pretty rare. I mean, you tell me. Josh Chandley: Yes. Vincent - 52E.: So having double digits is just like you know, unreachable. So yeah, it's just a discussion of you know, what do you need? What does it take for you to accept integrating a new SDK, right? knowing that you probably already have between ten and and fifteen. Warren (Upptic): I have a maybe a a macro question because part of what I'm assuming why Vincent is su suggesting a really high number, my assumption would be it would only get worse over time. But I I don't come from the admon side of things. You guys have both have more knowledge than me there. So I'm I'm curious, like, have you guys generally seen when you make a new integration that that imp performance it's you know, excluding like the primary players, that that import performance generally improves over time? Or is like best foot forward in the beginning and it only kind of incrementally gets worse? Or is it a mixed bag? Vincent - 52E.: I mean I can say two things. Of course, more code, more SDK, ANRs, drop off retention for sure. Then honeymood. You integrate someone new, like you test a new mediation, you have, you know, during the test period, everybody's lowering their margin, you're gonna think it's amazing, and then progressively when you're gonna look elsewhere, back to normal. Josh, what do you think? Josh Chandley: I I'd say generally that's dead on. I'd say when you launch, they're gonna enter a an LP where they're accepting lower margin. The question is. once they see what you have and what you have to sell and what they can put into it, are they very excited or not? Usually for the average, like the median ad network, decent volume, then it kind of slows down to like a new normal and it may slowly grind up from there a little bit. Sometimes it can be off to a big start, at like sh straight after LP, but you usually that's a network that's backed by a trillion dollar company. Vincent - 52E.: Can I give a shout out? Josh Chandley: And so I guess go a question. Vincent - 52E.: Can I give a shout out? Yeah, we didn't mention, but there is a a very interesting French publisher. I mean it's public on GitHub, right? But called YSO. They launched their network and they all certify in Max. so that Josh Chandley: Another one. Okay. Vincent - 52E.: yeah, so that's super interesting. I have no idea how they made it, but yeah, they made the cut. I have no idea if it's scaling or not. I did not integrate them, but it's worth mentioning. Another angle is you know with dynamic floor. so for for people who know, you know, it's one of the trends the past two years, and so a service company called NEFTA started with dynamic floor and right now, you know, they are are working on building their own network with the SDK footprint they have. And I think it's pretty interesting path, you know, how do you go from zero to one to scaling? Going the route of, hey, I am a service provider, I'm I'm delivering a service, integrate my SDK, good. And then I can the SDK is already there. Then I can switch to, hey, I have some demand. Can I serve my demand? And then you know, slowly but surely, you know, you can potentially grow and at some point be certified. So that's the two initiatives I'm seeing in in Europe that I are interesting to follow. Josh Chandley: So if you're an incumbent ad network, if you're an ad network with your o with with a demand team and an SCK implemented in Max or or whatever mediation platform brings you joy, is this publisher owned supply? Is this a is this a threat or a rounding error? Vincent - 52E.: I don't see that as a threat as a publisher on my end. Warren. Warren (Upptic): Yeah, my my intuition is that like I mean some some variant of all roads lead to app leveln. Like if you're building this thing up, it's some in one way or another, you're integrating to app levelin. There you're reselling through Josh Chandley: To Max. Warren (Upptic): their sales team, their algorithm. Yes. exactly. so I think like overall, it's almost like it's just a collective bargaining that you get by building these rather than integrating direct apps. I don't see an environment where One of these truly breaks through as a standalone ad network in the foreseeable future. But I do see there's that doesn't mean that there's not paths to explore these initiatives and increase the revenue for these businesses overall. I think just the end state does not look like this is new ad network X that I'm buying from. But another one of those invisible layers when you're buying from ad network why that it's plugging into. Does that make sense? Josh Chandley: Yeah, it does. I I'm just like stuck thinking about how I feel about it. I think I'm more bullish than you both. I think it could be a threat eventually, especially from the largest from that largest ad networks or from the largest game publishers. You know, I I really do believe a company the scale of ten cents has the capacity to to start to to work through some of these issues. It's just hard to understand how applicable that is. Is that a one off or is that actually a trend? And it sounds like what you would both say is that's a one off, Josh. You're talking about a trillion dollar company. Vincent - 52E.: There's not many, so yeah. I don't see ten of those popping up next day to my mailbox asking to integrate their SDK, yeah. Josh Chandley: Fair enough. All right, so make a prediction. How many publishers launch their own ad networks between like now and the the end of next year? Is it z zero, one, two, three, five? Like what what's the prediction? Warren (Upptic): Yeah, I my prediction I will take the the over on on three. I'm gonna say three or more, but let's define that a little bit. I mean they're going to announce through PR and to their shareholders that they're launching an ad network. What will it look like? I I Again, I'm back to this thesis of it's just gonna be another one of these invisible layers that ties into the leading ad networks as far as the end state. I don't think that's what they will be designing it and imagining it at, but again, that idea that all roads lead to app levin, like you're not gonna beat the leaders in the space as far as the technology level serving the inventory. So I believe that while There's a a small handful of businesses that have the inventory with Tencent being one, to launch this, that they're not gonna be able to replicate and go ahead to head with the the ad tech layer, the algorithm layer. And so eventually they will become subservient to app level and the like that have these superior algorithms. So three plus with a caveat that we are loosely defining, you know, they will self proclaim an ad network. The outside world may not observe them as such. Vincent, what do you think? Vincent - 52E.: Yeah, I mean in terms of publisher, you know, gaming publisher, I don't really see a lot of candidates. So if I think outside the box and we mention you know Chat GPT or it can be another of those you know big AI bots, potentially, because again, they're gonna very quickly reach a ceiling, they're gonna make some mistakes, and at some point in your chat bot, you know, window. You cannot really flood the user with ads and videos and stuff. So at some point you need to deliver those elsewhere. And obviously they raise a lot. They're gonna start, you know, obviously pushing subscription more and more. But at some point everybody needs an ad business. It's pure margin and you need to deliver them outside of your ONO. Is it gonna be a network in gaming, serving ads in gaming through Macs? Maybe, maybe somewhere else. you know, CTV obviously everywhere. But they have the capital, they're hiring the best. You mentioned Warren, the best practices just go and hire the one, the engineers, the product people who build that in another big company, hire them, you're gonna fast track, you know, your initiative. But probably, again, my bet will be a Chat GPT or or one of those big you know, AI company at some point gonna need some ad revenue and they're gonna need inventory to serve outside their ONO. Josh Chandley: I think I agree with I I still disagree with you guys, I think, on the some of the depth. I I agree that like three plus is inevitable. I I I would take the over on five. I wouldn't take the over on five succeeding, but I would take the over on five. I I think the this pull towards vertical integration and owning your own distribution is a trend. I think that's why We're we're seeing like like apps like best play become more interesting because it allows publishers to own their distribution more. I think that's why all sorts of ad networks are popping up. The the question is, it's not to me is I think the the low bar is can we build a new next gen kind of cross promotion system? I think that's very achievable. The grand filter is can you cross that chasm? Can you convince Vincent to accept less than a 10% ARP DAO uplift? Can you convince Vincent to start like and me to start writing checks and routing our demand through you? and there are really big risks to us doing that because now we're not just now financially funding competitors, we're training their machine learning models so that they can better acquire our users for themselves. And so there are risks there. I really do see a path that just requires perfect execution. I I do see this layered approach where, you know, whatever large publisher you're talking about, there's going to be, you know, another three to five percent of inventory out there owned by studios that have very similar users and and could get, you know, five, ten percent uplift from integrating. And then once you absorb that, now there's another layer that makes sense. So I can see this this slow story. I just think it requires perfect execution. It's it's just insanely difficult. And so excluding the publishers we've already named, who's the most likely candidate to move next? Warren (Upptic): Yeah, I've been thinking about this, Josh, since we were you know planning what we want to talk about today. And for me, the you know Tencent we've already called out before. The other entity that I think has a very interesting you know, I think there what are the components we need to to to make a emergent ad network from publisher owned inventory. The first thing you need is a ton of really compelling inventory, and then you need you know, I'll just shorthand to a ton of money. Right. And that ton of money you know buys you the people that build great algorithm, great U UI to have a standalone ads product. So that's all to say the other player that I think is very notable is you savvy games slash the Saudi PIF fund because we have in here within the savvy games group, we have brands like Scopely. That are one of the few houses that just have a very proven long-term expertise in performance marketing. And then you also have the brand giant of an electronic arts. I know earlier this year, I don't know too much about it, but they they launched EA advertising. it looks like a much more brand focused effort, but I think that's part of kind of the balanced diet of launching an ad network. Like it when we think about like you know buying buying YouTube tube traffic, there's two sets of buyers. There's the big brands. Buying impression-based campaigns, trying to get you know top of funnel, and then you have performance, buying through Google Ads. And Savvy Games Group has maybe the the different pieces to sell to that broad swath of buyers, so they actually can sell more of their inventory overall. So I think they're another interesting player. I don't think they have the technology component, but they have the bottomless capital required to pursue that. So that's my kind of dark host. Dark horse candidate for this. Vincent, what would you bet on? Vincent - 52E.: I mean I don't see any other companies, right? So Savvy obviously, big pockets, big big publisher, big names. other than that, honestly, I don't really see anyone having the scale. I mean, we can name, you know, some some of the biggest large some of the largest mobile gaming company in the world, but they are you know a tier below, right? so I would absolutely not bet on on someone else. Josh Chandley: Really? Warren (Upptic): That might be the smartest bet of all. No one. Josh Chandley: Maybe not betting. Yeah, I mean the the names that came to my mind triple dot, but then they don't have the right incentive because they're partially owned by AppLovin. and the r the reason they come to mind is because if you're gonna try to build an in app ad network, y the necessary ingredients in my mind are you need to have demand, well they all have demand, but you also need supply. and so Triple lot would be interesting if it wasn't for the fact that it was owned by AppLovin. Another one would came to my mind was Take Two, but they've already been bitten by this once. And so the the odds of a second one I think are pretty unlikely. That the only thing other than savvy that comes to mind is Dream. They are starting to have their own inventory, but more importantly. They're at mega scale and they're casual enough that buying in app ads makes a lot of sense for them. That's a maybe. I still think it's super unlikely. and so I even though I think that savvy is very much under indexed for for ad monetization, I don't know if they have the people in the in the in the building or the the planet, I guess they're so huge to to build an ad network. They're probably the best the most incentivized to do it, right? Vincent - 52E.: I will be very surprised on my end. Might be wrong, you know. Yeah, I Josh Chandley: I'd be shocked. Yeah. Vincent - 52E.: mean, you know, these days, twenty twenty six, you know, every day there is a news where you're just shocked. But so why not? Why not? Warren (Upptic): Yeah. Vincent - 52E.: I will be very, very surprised. Again, you need some you need a specific DNA, you need to be focused on that. It's it can be, you know, a dispersion, you know, you need you need to stay focused in this business when you build games at the quality of a dream games or or or you know, even triple dot you mentioned. I would be very surprised if those companies launch a network. Josh Chandley: All right. Well with that, let's call it. Vincent Morant, thanks for joining again. Vincent - 52E.: Pleasure. Warren (Upptic): Thanks y'all.