Season 7, Episode 24: Understanding rewarded user acquisition (with Tricia Han and Sampsa Jaatinen)

Let's get real. Half the time, you're not sure if your campaigns work. You just kind of hope they do. Attribution numbers look good, but what if those conversions would have happened anyway? Incremental helps you find out for real. No tracking people, no guessing games. Just clear answers on which campaigns actually move the needle. If you want to stop guessing and start knowing, go to Incremental. Because good marketing isn't about getting clicks, it's about getting results. Check them out at incremental.com. That's i-n-c-r-m-n-t-a-l.com. Or click the link in the show description. The problem is that the distinction needs to be drawn between the components of the economists and the correctness of their analysis. Welcome to the mobile dev memo podcast. This is your host, Eric Souford. And I'm joined today by Trisha Hahn and Samsa Yatin. Trisha, Samsa, welcome. Thanks, Eric. It's great to be here. Hey, Eric. Good to be here. Thank you for the invite. Well, you're very welcome to Samsa. I've known you for eternity. Trisha, we met more recently. I'm very excited to talk about Miss Play today and the rewarded UA space more generally. It's really fascinating space that I've wanted to do a podcast episode on for some time. So before we dive into that, maybe both could introduce yourself. So Trisha, we'll start with you. Absolutely. So I'm Trisha Hahn. I'm the CEO of Miss Play and Eric, probably the reason you and I have not necessarily met before this is I actually don't come from a traditional gaming background, but I've spent a lot of time in mobile apps. So I've been in consumer tech for over 20 years. My very first startup was actually a company that built mobile apps in a mobile platform. But it was pre-iPhone and it was so early that at the time I actually built a mobile ad-serving platform because it didn't exist. That's how long ago this was. But I've always really loved mobile apps. Most recently I was the CEO of a company called My Fitness Pal and it is a large scale mobile app for tracking fitness and nutrition. And the reason I mentioned it is one, it's mobile. It has a lot of the same dynamics as mobile games. What's really interesting about it is that it's a business that's based on data, on user engagement, monetization of that. And then there are a lot of behavioral mechanics that come into play as you can imagine because it's really about getting to good health. So I bring a lot of that over with me and really excited to see how much is transferable actually to the mobile gaming environment. Given the success of Ms. Play, I apparently a lot. A lot of it's transferable. Great, thank you. And Samsung. Hey, yeah, I'm Samsung. And I've been in mobile advertising space for a long time. I joined Unity 2014 through the Applifier acquisition. I started the data science machine learning analytics teams for Unity ads and spend more than decade building the Unity ads network. But now more recently, last fall, I joined Ms. Play. I joined the rewarded space as a chief data and AI officer. And he's still here. So that's a good sign. Yeah. All right. So we just love to start with anybody in mobile gaming is familiar with Ms. Play. It's grown at an incredibly rapid clip over the last couple of years. And it's become like kind of a major share taker of mobile UA budgets. I could imagine that people may be outside of mobile gaming and just in the broader digital advertising space, maybe less familiar. So maybe we could just start with an overview of Ms. Play's business and the rewarded UA model. Sure. That's a great place to start. So for those who may not know Ms. Play, we're often called the OG of the rewarded space. And specifically in terms of rewarded user acquisition. And that's because Ms. Play actually was one of the pioneers in the play in earn space. They started about 10 years ago in out of Montreal, Canada. And the way that rewarded works today is it really is about value exchange. And I think maybe the simplest way to talk about it is by using the analogy to loyalty programs and specifically like points programs and airline programs. And so there's a real value exchange between you got your user, your customer, you got your brand and there needs to be some sort of equal exchange for the two to work together. So the way it works for in rewarded is that you have your advertiser who is your game advertiser. They're of course spending ad dollars. They're giving access to their game and what they're getting in exchange are qualified users. The users, what they're giving, they're giving their time. They're playing the game and they're getting some rewards. We also have another part of the ecosystem here. Now host apps, what they're doing is they're often supplying their audience of players. And what they're getting exchanges, engagement of the audience and monetization. And then all of this activity happens on the ecosystem or the platform of Ms. Play where we are enabling the matching of these players, whether they're coming in directly through the Ms. Play app or through our partner host apps. And then matching those up accordingly with game advertisers. And then we enable along the way, sort of the lifetime of the user, engagement and recommendations. Great. And I would want to anchor, I think the discussion, it was certainly like, you know, just the kind of growth profile, I think, would want to anchor that to ATT, right? So I think ATT was probably a major catalyst for Ms. Play's growth and you know, please disabuse me of that misapprehension if I'm wrong there. But maybe we can just talk about how Ms. Play's business has evolved over time and like how it has evolved since ATT and how that just changed the dynamics of UA generally for mobile games. Yeah, it's obviously a really important topic. We might see it slightly differently. So ATT was not the reason necessarily that propelled Ms. Play forward. That being said, it certainly was a factor in where Ms. Play started. So as you know, Eric, ATT, what ended up happening is that Android and companies or Android First were really beneficiaries of that of ATT going into place just because they were able to still attract players. But the way that they did it or at least the really good ones who were able to scale was with a really a strong opt in incentive. And so it was easier to grow and scale on Android First. That being said, the landscape has changed a little bit. ATT was not the reason that Ms. Play was not on iOS. It actually was because for a long time, the Apple ecosystem had guidelines that restricted what they called incentivized installs. They actually changed that last year and that actually opened up the market for a lot of rewarded apps to now be able to be live on Apple full compliance and really start to find those players for rewarded advertisers. Right. I mean, historically, we had like the tap joys of the world that kind of could only operate on Android under the sort of restrictions of the app store developer policies. So maybe we start there. So Ms. Play was historically very strong on Android. So you've now rolled out Ms. Play for iPhone and you acquired my chips. What are the unique hurdles that you're seeing scaling the rewarded ecosystem in iOS under this kind of new set of rules, this different set of rules on Apple? Yeah. I mean, Apple is just a different ecosystem than Google Play with different rules. Not bad, just different. And so I think it's really important to understand what is Apple care about? They care a lot about privacy and data and they care of course a lot about opt-in consent. Luckily for Ms. Play, that's always been part of our model and our mechanism. It's really interesting. Again, going back to this idea of value exchange, consumers, as we know, they're willing to trade off a lot of times information for either less friction or for some value on the other side. And I think that's what rewarded really taps into. And so we've always asked for consent for users to be able to understand what game they might like to play and how they play it. And, you know, actually, this might be an interesting thing just to talk about in general. This is actually where rewarded is really different from let's say your traditional install campaigns. What rewarded and what Ms. Play has done in particular and where they've been a pioneer is that after getting the consent from the user, we are getting a lot of post-installed behavioral data from users, what they are downloading, but also what they're playing, how long they're playing, what they might be spending an IP or IAA when they're churning, for example. And also what are the things that might motivate them as well as the rewards that are interesting to them. And that's actually a really big difference between, again, rewarded in what I might call a traditional ad network type of model. And so, you know, back to the iOS thing, what's been really important is that we are in full compliance and that, again, we're getting that user to actively say, "Yeah, I want this and I give my consent to opt into this platform." Because I, again, I know very clearly what's on the other side with the reward is for me, it feels fair and it feels like a good exchange. Right, and to your point, you know, that trade-off can be a lot more abstract or nebulous just with a standard ad network. It's like, "Well, okay, I'll get maybe better ads." It's maybe less immediate of the benefit than these rewards that you get through rewarded UA. So, I'm sorry. Actually, I would love to get your thoughts on the operating environment just because, you know, coming from Unity, you know, it's sort of a more traditional gaming UA ad network ad network. And moving over to Miss Play, you join Miss Play before it started operating on iOS, right? I mean, you maybe like six months or a year before. I would just love to hear like what your experience has been like seeing that come to life and just the different operating norms, the different operating realities for like a rewarded UA. network. Yeah, so that's that is definitely one of the things that I find very interesting. It was a very interesting change for me from coming from the traditional ad network, which is two-sided marketplace to something that feels like three-sided marketplace, with a consumer to the actual user and the person who plays the games is also taken into account as a party who has needs or motivations and things like that. So it's a meaningful change. It's something that's very interesting. It looks to me like this because of this relationship that we have with the consumers and the ability to ask for their consent, ability to offer something really meaningful back so that people understand what they are giving and what they are getting against that consent, that data that they allow us to collect. The difference between the two environments seems smaller than for the traditional ad network, because this is something that's unique for example and I believe this is true for all the all the ad networks where the constant mechanism is completely different and we do know the difference in value for machine learning when people either can be followed across the applications versus not. So being able to have that consent makes bold of the environment to feel a little bit more similar in our case than for the traditional ad networks. Were there any sort of scale challenges that you didn't expect coming into misplay, just having come from the more traditional ad network space where opportunities, things that surprise you, things that you just couldn't have predicted? Not really surprises because this was known to me but there is a massive difference of course in the owner-operated, inventorated volume of the data when we specifically talk about machine learning. The volumes of the data that we are talking is obviously completely different than what the data volumes that you need to be talking about, multiple orders of magnitude difference. And if you know machine learning well enough that you know more is more, more data is always better for the model accuracy. But we have enough, we have enough to be able to do the most things. We have enough to be able to improve, we have enough of users to be able to really make meaningful machine learning and as said of course today that we have is richer. It's better, we have better behavior, all signals, we have a better understanding of the users in the system. So none of these were surprises, maybe I'll share the funny thing that always when you go to a new place, no matter where you go, people who work there, they think that they have a lot of data. When I was at Unity I thought we have a lot of data and then if someone joins from Meta, they laugh and ask us, they understand that the data again, the difference is, difference is multiple orders of magnitude. AI is changing everything, but it's only as good as the signals behind it. Branch connects customer interactions across paid, organic, offline, email, web and app touchpoints and turns them into the trusted context you need with links and attribution to capture the full user journey. Learn more at branch.io. And while you're there, check out Branch's AI Search and Discovery report covering insights from more than 300 enterprise marketing, growth and digital leaders to understand how the industry is responding to the rise of AI search. That's branch.io. Tristan, what's going on back to you? Just walking through the acquisition of my chips, we'll explain what that was and how it fits into the broader strategy. So we know ultimately when it comes to businesses like ours, it's actually fairly simple. Our customer is the advertiser and there's really just two things that they need. They need Roaz and they need Roaz at scale. And we had as Ms. Playden, our really good job over the course of our history, tens of millions of downloads, hundreds of millions of reward points, many hundreds of millions of data points. But that being said, we knew that we could take all of the learning that we had from having our own first party post-installed data and figure out how to apply that for the benefit of larger scale and for these other applications, you know, to build this larger network that we could take all of that. Again, feed it back to Samsa and his data science team and figure out how do we do this on behalf of our advertisers and within the environment of these host apps. So we met the math team, thought they were fantastic. They really built scale with quality as well as they were really strong in markets where we're also strong where we thought there was a very complimentary. They're quite strong in APEC and in Europe, we're quite strong in North America and Europe. And between what they'd built and what we knew, what we could apply specifically from the AI machine learning perspective that all of a sudden we have something that's much more interesting for on behalf of our advertisers, which is again, network of scale with a lot more data that we can run through the models so that we can do a much, much better job of matching people, of finding people to write intent at the right price, write scale through our models. Yeah, and I guess a good segue from that is the audience network. So so you launched an audience network in May. Maybe just talking me through that, the strategic impetus, how it's going, like what that provides. It's going really well. We are deep in, well, let me just say that Samson team actually have some really intriguing data from models that'll be that are coming out soon. And this is the basis of a lot of what we built by putting all of these data pieces together. We've got a bunch of exciting launches that are actually in progress and globally. What we're able to provide to our advertisers is quality, scale, less friction. And that was the whole impetus of building this misplay audience network. You know, I talk to a lot of customers. They're of course working with a lot of rewarded channels. That being said, every environment, every ecosystem works a little bit different. And of course, it takes some time to learn what works in each of these systems. And one of the signals that that not only do put in, but the can you get out of them. So our job was really like, how do we deliver more scale at the right price, you know, in such a way that our advertisers could really have a unified platform and experience. One button, so to speak, to deploy their spend to get what they needed. I remember we were speaking just a couple of months ago. You mentioned you had like a big announcement in May. And that was like the MAU moment. So that was that. I'm just kind of putting that together now in retrospect. That's right. MAU was really exciting for us. It's where we announced that we had joined forces with the math and the my chips product as well as that we were rolling out the misplay audience network. And so far, the reception has been really strong. We've been really excited by not only what we've been able to do in the short term, but also the future kind of strategic opportunity here. Yeah. And may you seem to be having like a renaissance. The last time I think I went three years ago. And you know, it just it because it was this independent show. And then it got acquired and there was the thought. And I think, you know, this was defensible that it was acquired by a public company. So they needed to boost the revenue and there was kind of this noticeable increase in vendors. But everyone who went this year said it was fantastic. So I'm looking forward to hopefully going next year. It was really, really great. And you know, one thing I'll add and what's interesting is I will say that rewarded as a category. I think is also helping to fuel some of the energy around MAU. And the only reason I say that is is you know, we were talking about it internally because misplay was one of the very early entrances here, you know, we as a company used to go to MAU. And you would see just a handful of rewarded kind of boosts on the floor. And if you went this year, it was really remarkable to see that how much of the floor was actually devoted to rewarded. And I think that's just a sign that there's a lot of excitement around the category. And I think it's because it's it's really delivering something new and different. And but more importantly, it's also more proven now for a lot of advertisers. So you really see that. Yeah. Well, I mean, you know, when we spoke the first time, I made the case that I thought rewarded was if, you know, if not the principal factor, certainly a major contributing factor to mobile gaming's rebound back to growth. And you know, like if you just look at the share of spend for a lot of advertisers, I mean, that must be true. There's just this big hole in UA budgets that got filled, right? There are, you know, absolutely right. Yes, we definitely have clients where 30% of their mobile spend is now in rewarded. And I saw that just the other day, some, I think there was another industry stat that there are some some advertisers who are spending as much as 60 to 70% of their spend on rewarded, which is really remarkable. If you think where it was again, five years ago, it was probably like one to two percent of their total spend. And so I think you are seeing this acceleration in the category. A lot of that has to do with how rewarded is different. And it goes back to that post install data and the ability to influence behavior for deeper gameplay for longer engagement, which is quite unique to the to this specific category of UA. Yeah. Well, yeah. I mean, that's, that's just connecting back to Samsung's multiple orders of magnitude comment. You just get so much richer. You just get a richer view of the consumer and you can do a lot more optimization as a result. Actually, I want to talk about that about that because legacy offer walls were kind of seen to be to deliver low quality traffic and high levels of churn, right? I mean, that's just to, you know, I think that was like the kind of legacy view on those channels. How does your AI engine, so you know, you've talked a lot about the AI engine publicly. How does that change the quality and retention of players compared to, you know, let's go like call it legacy rewarded out units like that. - Yeah, so how does that multiple orders of magnitude really contribute to like, you know, better cohorts and, you know, better retention? - Yeah, so before I start answering that question, I might get excited about this 'cause machine learning and this kind of things on my jam. So try to keep me contained if it looks like I'm going overboard. But I have a very simple, but I like simplifying things as well. So there's a threefold simplification that I wanna use for this particle case. One is intention matters, second, the game's matter, and then the timing matters. So when contrasting against the legacy of the walls, I think there's a little bit of a problematic setup or challenging setup that is influencing or impacting how well the offer walls work for something like this. 'Cause if you have a offer wall offering in your game, you give rewards that expect the person to return to your game. It's almost like you're offering like a side quest to go to play some other game for a little, but the reward that you get is really currency in that first game. That reward is completely meaningless unless you really return to that game and keep on playing. And there's limited amount of time people can put into games so that it's always competition of that attention. So that's what I mean by the intention. We don't have that 'cause we don't expect people to come and spend their time and money in misplaced, they just swing by while they are going somewhere else to have fun. But that's how we get to the machine. And of course, the AI engine that we are building is based on those two other principles, where one is we know that the game itself matters. There's, you can sometimes see simplifications that people talk about game categories, but people inside the gaming industry are very proud about their products. We know that games are highly specialized creations. So it's very important to be able to make recommendations for people specifically on things that they like. This is where the behavior, all signals, people's actions, both in the mislay application and the other games that they play meet the actual machine learning. This is where proper meets the road and we can make those recommendations that are great for people. We want to incentivize something that people are inclined to do anyway. That's usually where it works the best. Trying to incentivize or reward people doing something that's completely unnatural for them playing games that they hate. That would be very expensive and it doesn't help anyone. So the recommendations, just knowing the games, knowing the people and being able to make that match. Then I'll mention a little bit about the rewarding as well, because that's what I mean by the timing matters as well. Because this is important for the client performance. We don't want to take people out of the games that they enjoy. If someone is really having time of their life with a game that they recently installed, they are engaged, they are spending money, it's not really in our interest to try to get that person to install some other game. We would only try to incentivize people moving around when the time is right. So that's why I say that it's this kind of a trio of things when it comes to the AI engine, knowing what to recommend, knowing when to try to make people to do something. And that is very well aligned with the longer term performance of the user acquisition as well. When we see that someone is engaged, that's the behavior that we want to reinforce. When we know that they are about to find something else or need something else, then it's our time to take an opposite stance and act. Yeah, and I think structurally with games, one thing that is underappreciated is the opacity for the user when they are engaged in the game, because the game that a person's playing and maybe is monetizing in is unlikely to show them as, right? That's a lot of game developers do that. They'll shut ads off for someone if they even seem likely to make a purchase, well, you wouldn't want to divert them elsewhere. And so having that touch point with people while they're engaged in another game is unique, because like, you know, I was talking about like, you think about like the visibility you have a new user when they click and install a game, oftentimes that visibility drops to zero for some very extended amount of time because they're playing the game. And people usually play one game at a time. And so if they're engaged in that game, they're probably not seeing any ads. And so they just don't pop up in the bid stream, even if there is a maid available, right? And so I guess that's kind of the unique value prop here is like, well, you've got the, you know, like you mentioned, you've got the OO inventory, so they're coming back and you're seeing them, even if they are engaged in another game. And so you've got that sort of that through line of their behaviors. That is correct. We see Tim, everyone know one thing, because they want to come back and pick up their rewards, but also because we built the economies for the games. We see how people reach these milestones that we set in the games that we reward them against. So we have that window on really what's happening with them with the users when they are playing the games. - You would know this. I mean, just having worked in it in the space for some time. But like that was one of the impediments to doing anything other than contextual ads, you know, going back, you know, years, is that like if your only visibility is in games, if users playing a game, they're engaged, they're likely to monetize, they're not seeing any ads. So you have no visibility. You see the click in the install and then maybe they pop up again in a year when they are back in the market for another game. And so that really only left contextual advertising. So game to game, like building the associations across the game, because you didn't really have a rich behavioral data set on the person. Even if you could have that, you just didn't because they would disappear for long stretches of time while they were engaged in another game that didn't want to show ads to them. Am I characterizing that correctly? - You are, of course, that goes back even more strongly with 80 to the two men's and because when people pop back up somewhere, you wouldn't even know if they are the same, so no different person. But that's true. We have a system where people have reason to stay in our ecosystem. We have that visibility without being in competition with our clients, because we don't, we are not game. We don't want the same thing from the users to go through our system then what our clients want. - Okay. So your loyalty play initiative aims to bring the play and earn philosophy to non-gaming apps, which industries outside of mobile gaming are showing the highest demand for rewarded user acquisition? - We see really good demand in a bunch of industries. I'll say, I think what's really interesting about loyalty play and in particular, we also acquired a company called Connected Rewards. And Connected Rewards had done a really, really good job specifically going into categories like fuel and convenience and QSR and of course, things like Fintech and Commerce are really good opportunities for this kind of rewarded UA. In particular, I'll just call out, like for example, we work with companies as big as Chevron. And I don't think that's one that you would naturally think of for a gaming reward. But I think what the folks at Chevron realized was like, you know, back to what Sam's said before, how do you introduce an experience that is already something that people are actively doing that they like to do? And how do you then, is there some opportunity for a positive brand association and engagement with the brand, even if it's not exactly what the brand is known for? And what I can tell you is like for our fuel category and our convenience or our category, we are seeing up to like a 44% faster return trip rate for customers who downloaded games as part of this loyalty program. And in terms of, for example, even restaurants, another interesting one, it's sort of like quick serve. We work with companies like Checkers and rallies. And what they told us is, and what we've been able to measure is that we saw a 54% engagement from last users once we introduced this program and kind of let them know that this was an opportunity. And so, you know, again, back to the analogy of like, what's different between us and other types of ad networks. Ad networks can deliver somebody to the restaurant. But what we can do is not only do we deliver you, but we can also help incentivize the ordering of the appetizer and the main course. Because we might give you a free dessert. Like it's very interesting. It's happening in real time and it's sort of blending sort of this online and offline experience, which is the way that most people operate anyway. - Right, and that totally makes sense. Have you seen uptake for just other non-gaming app categories? - Yep, yep. Commerce, of course, always strong, fintech strong, but these are all evolving. And I think what's really interesting about them, especially for game companies is that we are now able to give them access to new inventories of players that they didn't necessarily have before. It's just slightly different than what you might get through traditional channels. - Right, my sense is like if you think about the kind of tailwinds from mobile gaming, that's a big one. Just injecting more varied demand into, you know, the IAA economy, which should drive CPMs up. And, you know, that's something that just, you could turn on. And then the other is, this is a little bit in flux and we're seeing this evolve kind of in real time and it seems like there's, you know, some new announcement that's hard to parse every week, but it's to fall out from the Epic B Google, an Epic V apple lawsuits where, you know, link out is permissible and, you know, you're seeing like a surge in D to C revenue. So it does feel like those are, I think those kind of get, both of those get kind of get dismissed as not being meaningful, but I see them both is very meaningful and I see them in concert to be extraordinarily meaningful. Like I mean, that seems like just across both of those, you can imagine, I don't know, like 30 to 50% revenue opportunity with like very little fundamental work needed. That's really just stuff that's like plug and play. - I agree, I agree with you. Again, the industry is so innovative and it's gonna keep finding the next level of experience for users for the overall ecosystem to continue to deliver interesting experiences, monetize better yield overall. And so, you know, this is what gets me super excited about the space in the industry, is that just as you said, we're seeing new announcements every week, every month. It certainly keeps us on our toes, keeps us continuing to innovate and grow and really think about how do we best serve the advertiser? And again, it might be through integrations with the, with web shops, for example, or thinking about our own models and how do we evolve those to really make sure that we're on the cutting edge. - So I called out rewarded, so every year I do predictions for mobile gaming and predictions more broadly for like mobile advertising. I called out rewarded in my 2025 prediction, so those are published on January 1st, 2025. So this is going back, you know, 18 months. But I called out rewarded. I said that's one of the reasons that we've seen mobile gaming return to growth. And I gave a presentation around that same time at PGC in London and I said, look, I think mobile gaming grows 5% to 10% this year and it came in around like seven. And that was one of the big, you know, sort of components of the thesis. Like reward is just driving more growth than people recognizing. When you pair that with the ability to then, you have incremental ad spend on channel that, on a channel that delivers Dimon Shwell-Roe as plus, then you're able to, I think, reducing the commission on the D to C stuff is interesting and certainly there's a benefit to that my sense is like the real value. And you know, Somps, this is kind of aligned with what we're talking about is the personalization. So you know, just a third party web, a web based, an actual web shop is far easier to personalize than anything that you can offer in the AppStore, just because of all the constraints around registering the skews and that kind of thing. So my sense is like those two things in combination present like a real tailwind from mobile gaming and I think there's real reason to be excited about the category. Mobile game developers no longer need to pay up to 30% in major AppStore fees. With XOla Web Shop, you can create a direct storefront, cut fees down to as low as 5% and keep players engaged with bundles, rewards, and analytics. Start today at xola.com. That's xsolla.com or use the link in the episode show notes. How do you apply the personalization with the rewards? How are you utilizing machine learning in the back end to ensure that players are seeing the most relevant things? Yeah, it is a relatively critical part of the AI engine that we are building. There's the recommendations that we touched already and the rewarding just to give a little bit of the basics. There's two different type of rewards that people can accumulate in this play and Wyoming. It's the units and the gems. And units are the basic currency that you can accumulate just by playing the games. It's tied to the time spent or the milestone that you can create in the game and then there's an element of personalization that goes to get to it with that. The gems are tied to the secondary special currency that is related to making IAP purchases. But there are different surfaces where that genes how those units are accumulated. There are different live uptasks. There are different type of special tasks that people can engage with. And all of these surfaces are something that we can control in terms of this, of course, the base accumulation of the rewards tied to those achievements and the gameplay progression. But we can always use machine learning to adjust how fast those milestones are achieved or what is the actual outcome of those milestones. We can expose these to people ahead of time so that they see the correct in front of them. And then, of course, all of these live uptasks are something that we fully control so that we can have a system where we adjust what do we reward people for? And this goes back to that question of do we want to make sure that they feel like they are rewarded for meaningful things, they get rewards for doing the things that they feel like they should be rewarded and then the meaningful outcomes for the advertisers as well. So we see this as a two different types of problems. We do predict things. We do predict those things like install likelihoods. We do predict germ likelihood. We want to understand when the engagement in the specific game is declining. What are those culmination points and trying to understand them so that we know how to react? But we also see this as a closed control feedback problem. So we look at the rewarding person itself, why do we have a reinforcement learning problem? Because predictions are something where we just predict what happens if we don't do anything. And then the reinforcement learning helps us to understand what happens if we start taking actions and choosing those actions. What happens if we reward this person? What happens if we don't? What happens if we reward them to do a versus b? So we are building this optimization system, this AI engine that combines all of these predictions and this ability for us to choose different actions and optimizing them, giving them the personal attributes of the person, what they are doing. Again, that's why I said that the timing matters. Where are they in their journey with the game? Where are they in their journey with misdplay? What is meaningful for them right now? So combining all of these things, we are building this engine that's capable of making these personalized assessments, but also taking those personalized actions, which is super important for us and super powerful for making sure that we can guide people to do what we think is best for them as well. Maybe I could just overlay the qualitative aspect of this, Eric. I mean, you're seeing what we're seeing in the industry, which is a growth in the category. And some of that is, I think, consumer led. They're really savvy. And I think a lot of consumers these days, they're used again, credit cards with points, loyalty programs for all of their favorite places, whether that be the movie theater or their favorite restaurant or their favorite shop. And so many of our consumers are like, well, I'm doing something that I enjoy anyway. I'd like to get, it's fun that I also get a reward for. So that's like one interesting thing. But when I think back to even working for, for in the health side, there is a lot to, you know, there's a lot we see in the data. And then there's a lot we try and understand for our players in terms of what does motivate them. Why are they here? And I think it's what gaming companies know really, really well. There are rewards. There's also just the social element. There's competition. And there are a lot of other factors. And so that I think has to go into the personalization as well. And we spend a lot of time thinking about the segmentation, what really is important for each of these players and what are they looking to get out of it? Because it's really different as you can imagine. And all in all, I think we are going through much of a recession of this industry as well. Because all the big ad networks have built their AI and Chains they've seen the benefits the industry has seen, the power of these things. So of course, we believe that this is something that's going to be true for rewarded space as well. And obviously we want to be the old G's for building the AI and Chains for this purpose as well. Yeah, I think the point about consumer expectation of just like royalty programs, reward programs is a really salient one. If you think about a lot of the arguments in favor of web three, it was like, well, people want to have ownership of these assets that they're buying. And I think that proved to not really be that compelling. But maybe I'm a consumer. I'm spending on a consumer product. I expect to get some kind of benefits of participation over time. And that does sort of track with my usage. And that does seem like, yeah, that's pretty pervasive across the consumer economy. Why does it not exist in the app economy? And to your point about there is more opportunity beyond just the major platforms. Another thing that we a lot of times hear from our users and from people in the industry is like, the reason they like rewarded and they like misplay, there is an element of discovery too. It's really hard sometimes in the app stores to find games that you like. We know they can be game. This is kind of a more natural organic way to find and discover games that you might really enjoy. It's hard. There's so much noise out there. Yeah. And there were, I don't know how many attempts at TikTok for games discovery. You get an app or Tinder. I've seen a bunch of these where you hop on and you swipe left or right. I don't know which-- I don't-- I never used Tinder. I don't know which was corresponds to liking or disliking. But these attempts that like, incentive, are like of motivating greater distance. discovery through these kind of game-of-fight aspects. And it turns out, well, what you really need to do is give someone something. That just there needs to be incentivized in some way. And just having another surface area for discovering a game is probably not that compelling unless they get something in return. - Yeah, but what we're finding is you can't do this manually and you can't do it with your instincts. And this is why you need the advanced machine learning, the AI, to do this at scale in a personalized way, to be effective. - So I'm cognizant of the time. I do wanna touch on the unit economics of this model. Maybe you could walk me through how the unit economics for rewarded UA differ from traditional in-game app install campaigns. So we talk about the sort of data flow, you know, the ability to do personalization, but it's not to be like, how did the unit economics change when you're introducing the rewarded aspect and the deeper connectivity? - Well, this is actually where the industry is headed. I think already. So the old model was like the UA channel got paid on the install and right and the user experience is totally separate. And these days, I think what we all see is that what matters more is like the overall progression, like day seven matters, day 30 matters, beyond that matters a lot. Because again, in order for the game maker, of course, to make back their ad spend and to grow their business, they need to see that engagement, that longer term engagement from players based on their own model. For us, what Miss Play does is that, you know, we are in effect sharing the ad revenue that we're getting from our game players with our players, right? Some portion of that as well. So they also benefit. So that's where rewarded is a little bit different and become sort of win-win-win for the advertiser, for the player. And then of course, even for the host app as well, they participate in that in the overall unit economics. - Okay, and I think like it took a good place to end is how do you see the reward at UA space evolving in the midterm? And maybe I could also, you know, just ask for your perspective on how you see mobile gaming evolving in the midterm. Like what do you see coming next for rewarded more broadly than mobile gaming ecosystem? - For the mobile gaming ecosystem, we see rewarded continuing to grow and continuing to be a much larger part of the overall marketing stack. Again, you know, we've seen it over the last five years even where it was single digits as a portion of the overall UA spend has now double digits and just continuing to grow. So we're really excited about it. That's what we see for the space overall. And in terms of mobile games, I will say like one thing that I think that the industry is a little bit concerned about actually is AI, right? And is there a lot of slap out there because people are able to create a lot of content and games very, very quickly. You know, it's something that we also think about a lot in terms of how do you make sure that the games have to be quality back to Samson's point in order to find its audience? And so for us, what that means is actually we're often curating the games even on our platform because we want to make sure that they make sense for our players and they're actually going to enjoy them and like them and that there's a really good model behind it. You know, it's something that we think about a lot. We talk to a lot of our advertisers about, and obviously to our players about it. Yeah, my thought about the industry is what I've already mentioned. I think we are going through a very rapid materialization of as posivity today that the technology is behind these AI engines and the whole industry. My non-advertising take here that I would like to share is that I feel like sometimes, like this is from the game, more like a gaming. I'm a huge gamer still. I play on all sorts of platforms. And I feel like there is at least a little bit of an attempt to blur the lines between mobile gaming and console gaming. It's looking at something like Steam Deck. I don't know whether that's PC gaming or mobile gaming. It blurts the line. It's something that you can carry around. So maybe that becomes a little bit of a competition for what we traditionally mean by mobile gaming. Who knows? I don't know. It's hard to say that there's going to be other games like that. Of course, mostly Steam games, I would think are more hardcore than majority of the mobile games. But I think that's an interesting thing to look at. Tricia, Sansa, I appreciate your time today. I appreciate you both sharing your wisdom. How can people learn more about misplay? How can they engage with misplay? Well, thanks for having us on again, Eric. They can find us at misplay.com. They can download the iOS or Android app on Google Play or the iOS store. And or we actually again are working with a lot of other host apps. And they can also interact with misplay there. Probably hard to find, but lots and lots of places. Or they can come say hi to MAU next year. Or they can come say hi at MAU. Yes, we're a very global company. So we're also at China Joy and Tokyo Game Show and a lot of different places around Europe as well. Great. Well, thank you both so much. I appreciate your time. Take care. Thanks, Eric. Thank you, Eric. [MUSIC PLAYING]