Josh Chandley: CloudX launched with a promise fair ad infrastructure for the intelligence era. I called it the new neutral. The idea that neutral infrastructure becomes valuable partially because it's not vertically integrated. But on our last podcast, Jim Payne said something that complicates that. There are clear efficiencies to a vertical integration approach, and that's not going away. Today on Mobile Signal, CloudX announces agentic buying, a move on the demand side. So can a platform stay neutral and vertically integrate? Or do you eventually have to choose? I'm your host today, Josh Chandley. Joining me is Jim Payne, co-founder and CEO of CloudX.
Jim: Thank you, Josh. Appreciate you having me back on.
Josh Chandley: last time we talked four months ago. CloudX was just going live. ⁓ What's adoption looked like since then? and what kind of uplift are we seeing?
Jim: It's true. It's it's hard to imagine that it is only four months ago because ⁓ so much has happened in terms of both the product, the customers that are using it and ⁓ Everything we've learned about actually the system, now of it's in practice. So ⁓ you know, traction is is really ⁓ we're pretty happy with where we are. We've got hundreds of apps on the platform now, dozens of publishers. ⁓ it's a a lot of early adopters, people who have wanted to get their hands dirty with the technology and see what the agents could do for them. And you know, they've been h very helpful in helping in in informing our roadmap and and so we try to move as fast as our customers do ⁓ getting great product back up. Yeah.
Josh Chandley: What does an early adopter look like for CloudX?
Jim: Yeah, an early adopter looks like someone who wants the the cutting edge, you know, who is ⁓ probably pretty frustrated with, you know, the amount of work that they have to do to do ⁓ to pull off the a normal monetization routine. So you know the early adopters that we've seen ⁓ are people who have a lot of apps, a lot of line items, a lot of different partners, many different monetization strategies, a a very keen interest in experimentation. I think that's where our early platform has shined. But what's interesting is that Yeah, a large surface area and a complex operation. Yeah, that's really where I think CloudX agentic approach ⁓ probably provides the most value out of the gate.
Josh Chandley: Large surface area? Go ahead.
Jim: ⁓ but you know, we weren't really sure what was going to happen once we launched out of GA, quite frankly. and ⁓ you know, what has really been ⁓ extremely ⁓ great to see for us is that every publisher who's launched is seeing an uplift of 10 to 20 percent somewhere in there. Different categories a little bit less, some some categories a little bit more. ⁓ and that's just by virtue of doing a little bit better job of the monetization ⁓ that the agents can do and ⁓ the demand aggregation that we're bringing to the table. So that's just out of the gate that 10 to 20% lift ⁓ has been quite compelling for our early folks. They're extremely happy with it. You know as well as I do this small market, so you know the word of mouth spreads quickly. ⁓ And if you can get a 10 to 20% lift and do less work and have the amazing cloud X team helping you monetize your app, I think it's a pretty good value proposition. So we're getting a lot of ⁓ of additional interest from people you know down that adoption curve we might have ⁓ you know, wanted to see early returns first before the movie.
Josh Chandley: It's funny that number. In December 2020, I I A-B tested Max against Mopub for the first time. I forget the exact number, but our uplift was somewhere between 15 and 20%.
Jim: Yeah. It's it's I I don't know. Strange coincidence.
Josh Chandley: Funny how that happens. ⁓ yeah. Maybe. ⁓ What does integration look like? Because when a publisher integrates TEE 2.0, that doesn't actually mean they're moving 100% of their supply over to CloudX. So what does integrating actually mean?
Jim: Well so I'm glad you brought up T E two point ⁓ because ⁓ one point ⁓ was the notion ⁓ of a secure auction in which bids and information would not leak out of the ⁓ auction itself and the auction's ⁓ code that we were running ⁓ provably ⁓ in GitHub. So you go to GitHub, you see the code, you see that it's a fair auction, very simple. ⁓ we can actually provably show you that that's the code that we ran. ⁓ we can also make sure that bids ⁓ information ⁓ that might be attached to a bid ⁓ that didn't win the auction ⁓ stripped away ⁓ left in that enclave. ⁓ that was our first approach towards a secure ⁓ infrastructure for basically advertising programmatic auctions where you have a bunch of of and of counterparties who don't necessarily trust each other and want to make sure their signal isn't being misused. including by us, by the way. ⁓ And we did that in partnership with Meta in the earliest days. TE 2.0 takes that one step further ⁓ beyond just our own system and contemplates the idea that, well, publisher might have more than one supply platform for a variety of reasons. You know, some demand is only expressed in certain ways in certain platforms. If that's true, then you don't want to necessarily walk away from that incremental demand, but you also want to be able to control your experience with agents like we offer. ⁓
Josh Chandley: Have an opportunity to create some uplift somewhere.
Jim: Yeah, right, exactly. So how do you actually compare bids from sensitive buyers between two different systems? We realized that the best way to do that would be to s extend our secure enclave to and that's TE two point ⁓ There's a blog post about that we launched ⁓ a couple of weeks ago. And so basically that actual bid can survive through the auction in our system and then actually be compared against another platform bid in a secure fashion. So we don't have to do things like an aggregated bid where the bid is like the median price of all the bids that we've seen from a particular buyer. It's the actual bid itself.
Josh Chandley: Well well that's hard, right? Because some networks don't want you sharing that data.
Jim: That's exactly right. That's why we do in the secure enclave, so we can provably say this is not being emitted out of the transaction, but it's being used for the comparison. ⁓ so that's TE 2.0. So that allows a publisher to incrementally introduce the CloudX infrastructure into their monetization stack. ⁓ so rather than switch ⁓ 100%, they can ⁓ you know do different ⁓ types of hybrid installations. And we have a blog post about that too, if you want to read about that. But I should mention that actually ⁓ interestingly enough, you know, we we do this as just a convenience feature to help people de-risk their installation of our ⁓ a software. But we are seeing a lot of people actually move ⁓ completely full platform with us already because it's working so well.
Josh Chandley: my understanding ⁓ in my mind of TE too is it provides a secure way ⁓ of essentially not leaving your supply platform, whether that's a traditional mediator or a homebrew or something else. ⁓ And When CloudX can provide uplift by whatever you define uplift to be, they can flow in, but otherwise you're essentially guaranteed your baseline performance. Is that a good understanding?
Jim: ⁓ that's a that's a good understanding, yeah. I I think that is probably true.
Josh Chandley: Okay. Okay. Well let's get into why you're here. Before we walk through CloudX's new agency buying in detail, what is it? Is it is it an ad network? Is it a is it a DSP? What are we talking about?
Jim: Yeah, well I you know, it would be much easier to launch something and call it ⁓ a name that people know what it means. No word if we were either of those, I would have gone with those names. But we ⁓ we decided agentic buying because it's most descriptive of what it is actually. And so ⁓ you know, I think last time we talked ⁓ about this ⁓ concept of neutrality and I I know we're gonna get to that, ⁓ because I I do wanna
Josh Chandley: Yes. Yeah, it would be really helpful for me. Yeah.
Jim: you know, answer your question that you brought up at the beginning of the of the pod. But you know, ultimately What we hear from publishers is that, you know, they're on the sell side, but a lot times they're on the buy side too. So you're running a studio, you're running an app, you're ⁓ whatever it is, you're gonna be monetizing the traffic you have and you're gonna be using some of that revenue that you generate to go drive growth. And gro driving growth is buying traffic that works on your app. and in your portfolio of apps. And so to me it's always been you know an opportunity ⁓ writ large to take our vision of agents doing the work and making hard things easier for you and then and unbundling the ⁓ all the different components and giving you transparency, better pricing and more flexibility. That should apply as equally as well to the growth side of the equation as it should the monetization side. And conveniently it's the same it's the same customer. So you know, I look at this ⁓ our agentic buying system as basically a means to provide that growth lever for people in a way that feels very much a first class citizen of the platform that we've built. So you don't have to give it additional data because we already know. You don't have to give us ⁓ your ads and app ⁓ your sorry, your apps and ad units and you know all that information because we already have it. ⁓ you don't have to give us your traffic and geographic breakdowns because already have all that. We have have the transaction information. So So you can very simply create campaigns that leverage all those assets and go out into the CloudEx network and buy that traffic and bring it back to you. And that's what we're launching ⁓ in this agentic buying launch this week.
Josh Chandley: How does it make money? Most demand services on an arbitrage model. ⁓ I think a pessimistic but efficient description of the arbitrage is charge demand as much as you can, pay supply as little as you can, and your profit is the delta. ⁓ pretty normal, is that what you're doing?
Jim: ⁓ well that is descriptive of an arbitrage for sure. ⁓ and I think that would be ⁓ the ad network model is ⁓ you know you're buying low and you're selling high. ⁓ and and there's ⁓ it it ⁓ to to be fair, it's not a pure market making spread where you're buying and selling ⁓ the same exact thing. Ad networks take risk in doing it, right? So they're selling a CPI and they're buying on a CPM. So there's some risk inherent in that. There's also a lot of learning that you have to do as an ad network. So the ad network will figure out what works, they spend a lot of money, they're losing. Losing money for a while, there's a J curve that happens, and then all of a sudden they're making a lot of money. So the arbitrage is a sort of a way to pay back the experimentation period. ⁓ so it's not necessarily nefarious, it's just one way of doing things. ⁓ Yeah, ⁓ right, exactly. The problem is like the arbitrage can last very, very long time. You know, so if you're if you're a publisher and you're out there for years and years and years, like that profitable arbitrage is actually quite a lot.
Josh Chandley: No no. It is providing a service.
Jim: of money writ large. and so to be to answer your your prior question, we're not an ad network because we're not doing an arbitrage model. We're not taking any actual risk in the transaction and you know transforming CPM into CPI. We're helping the publisher do that on their own. Similarly, we're not a DSP, although we're probably more like a DSP than not. Why are we not a DSP? And that the answer to that is very simply we're not in programmatic. ⁓ so you know it in a way we're a demand-side platform, I guess you could say, right? Because we're helping a ⁓ publisher express their demand into some type of ecosystem. But I think DSP has a connotation that it's programmatic, it's RTB, and it's QPS. And it's bidded. And that's not exactly what we're doing. ⁓ so it's closer to a DSP though. and I I think importantly, what we're doing is our take rate is a flat take rate. And so that margin is a fixed margin and it's based on your connectivity into our ecosystem. And and so that's an important consideration of what ⁓ of how to think about our agentic buying tool.
Josh Chandley: So flat take rate, no arbitrage.
Jim: That's right. And it's an unbundling too. So we're not ⁓ expecting you to necessarily use us for everything. We're not expecting you to you can bring your own creative. ⁓ you know, it's designed to be a little bit more do-it-yourself than a other another platform. ⁓ And the reason do-it-yourself is because the agents are sitting there doing a lot of the work and optimization for you. And so our view is if the agents are doing the work, you should accrue most of the benefit of that. So ⁓ it we're not real we're not in it to make a large margin like you know ⁓ many ⁓ folks on that side are. And that's because you were trying to build this whole platform play out in a lower cost provider overall.
Josh Chandley: my my brain ⁓ wired for RTB, like real-time bidding. How is how is ⁓ I can't even imagine the idea that there's not a real time bid in an auction. how is targeting different in this agency world?
Jim: Yeah. It it's a great point because we do still run an auction for every impression and there's a bid. There is a bid for every auction, but it's not QPS, right? So it's you don't have an agent that's sitting there listening to every auction and submitting a bid on that. ⁓ it's ⁓ it's a bit more of a vertically integrated ⁓ model so that that we don't have that same supply path issue. ⁓ so it's not and it and we're not bidding out into the outside world either. So it's not RTB in that fashion.
Josh Chandley: Okay. Okay. Okay. I see.
Jim: It's it's a little bit more akin to ⁓ the what the like the the ⁓ ad CP, you know the the context protocol efforts are, which is more like how can I more conveniently or compactly express my demand at the at the other side of the transaction, such that I'm I'm not collecting all this QPS and putting a lot of server hardware against responding with a price and a certain amount of latency. That's kind of a brute force approach. So the agents allow you to then express that demand more conveniently at the edge. And so that's that's why it's more like a peer-to-peer model ⁓ than it is a ⁓ strictly a DSP.
Josh Chandley: That's new. What kind of campaigns are gonna be supported?
Jim: ⁓ everything you'd expect us to support ⁓ we'll eventually do. the it's a private beta th for Q three. So we're we've got a lot of work to do. you know, we're ⁓ no strangers to this. We know it's a lot of work. but ⁓ ultimately, you know, the the campaign types that that people want to ⁓ want to run we're gonna offer.
Josh Chandley: So you're thinking CPI to launch ROAS campaigns, ⁓ CPE campaigns to come and at some point in the future.
Jim: Yeah, exactly. The the more complicated ⁓ campaign types are predicated on data, so we just gotta get the flywheel spinning. So we'll do the simplest things first. Scale.
Josh Chandley: Roger. What's higher priority, ad ROS or IAP ROAS
Jim: Good question. you know AdROS is probably most p ⁓ differentiated, I'd argue, and then also most ⁓ inherent to our platform because we're working on the sell side too.
Josh Chandley: It it's also on pub and admon people hate to hear this, but it's easier. There's a lot more signal and it's a lot more up front. So yeah, it it is what it is. ⁓ yeah. Yeah, yeah, yeah. ⁓ and I I I think I may have already heard this, but ⁓ when a publisher buys through CloudX in this peer-to-peer fashion, whose supply ⁓
Jim: Yeah, yeah, that's right. Yeah. I is extremely sparse, but everyone sees that's right.
Josh Chandley: Are they buying?
Jim: So to ⁓ in order to launch this, we're it's all inside the CloudX ecosystem. ⁓ but you may have seen, you may not have seen, but we we donated a bunch of our mobile context protocol work to the IEB. And that is basically the notion of expressing a campaign on a foreign system. So internally we use a form of this, ⁓ more convenient form. But you can imagine someone with another SSB. implementing this protocol and our agentic buying tool could actually go load campaigns into that ⁓ third party system and keep them up to date and make sure they're all working and you know handle all the different payments and credit issues that come along with that. But there's nothing technically standing in the way of that. In fact we'd like it to happen ⁓ that way. but ⁓ you know we felt like we needed to get the ball rolling on our own ⁓ network first.
Josh Chandley: So a GenSec buying is a fee-based way of expressing your demand. There's no arbitrage involved. And right now it's through CloudX, but in theory, it could be buying from any supply that integrates that.
Jim: Absolutely, yeah. ⁓ I don't think you know, we wouldn't ⁓ artificially stop ourselves ⁓ from serving the publisher.
Josh Chandley: Okay. I guess my last question is, why are you offering this product?
Jim: Well, it's the number one thing that people wanna want to see from us. So they love the agen Yeah, absolutely. They want the they love the agentic selling and they are
Josh Chandley: ⁓ really? Yeah.
Jim: Their first request is, can you help me with growth too? Because I want to do this exact thing. I want to unbundle all the functions that I get from an ad network and kind of take control of each of them in a way that is matches my business and my, you know, taste and sensibility and everything. So that's all the way from creative generation to targeting and modeling to the what the ad experience looks like. And then actually, you know, you reap the rewards ideally if you do it well, because you understand your traffic best. And you can bring your own first party data to the table, integrate that with our cloud agent and it all works to make a better model and a better set of creatives for you. And you don't have to trust a third party really to do that. If that makes sense. ⁓ so
Josh Chandley: Who who owns that who who owns that data? Do do you o own that go ahead.
Jim: Yeah, so good. We absolutely do not. So we don't own that data. That's basically data that you introduce into the agentic flow. And then it creates something that's a model that can then be used by our agentic buyer to express that demand into the network. Similarly, the creative ⁓ generation stuff that's a token path ⁓ opportunity for us. So we basically we we charge you for token usage, but at the end of that, you own the creatives that come out of it. And that can be image, video, and playable. And in fact, ⁓ that creative generation tool suite. Is what we're launching for anybody to take a look at and use, ⁓ even while the network piece is in private beta. So that's a anyone can can come and use that. That's a that's a more public beta offering. And so you can come in and create all these playables for all the apps that you already have in our system. You then generate a thousand different variations. You can go use that anywhere you'd like.
Josh Chandley: Fascinating. This is even for gaming, because I haven't seen anybody really succeed at like a Gensec playables for gaming yet.
Jim: Yeah, well, ⁓ you know, ⁓ hopefully it succeeds. I think it'll be look it's not I don't think it's a business, but it's a it's a piece of functionality that you need in order to achieve the growth goals that you're looking for. And the fact that it's vertically integrated with the rest of our system just it it's a it's ultimately a convenience, but it's also a tighter optimization loop. Right. So the agent can wake up in the morning and say, I'm gonna try a thousand variations of this playable. Puts it out there, sees it works and What doesn't work very quickly and then comes back and makes iterations. And that all just happens automatically while you're not really paying attention to it. But at the end of the day, those creatives, all that data, what worked and what didn't work, and importantly the economic benefits of understanding your own buying and finding those arbitrages, those come back to you. So remember we talked about the ad network J curve. You know, you may be spending a little bit more upfront to get the creative exactly right. But then you reap the rewards of that efficiency over the long run as you spend against that.
Josh Chandley: I mean, man, you're still in the dream, a Gensic creative production, right? So the AI makes the ad, it measures the result and it learns from that and then makes a better ad. ⁓ how f are you gonna walk the whole way down that path? Go ahead.
Jim: I mean this is a ⁓ yeah, and so this this is not theoretical, it's not hypothetical, it's this is something we can offer you now. And ⁓ and I I understand ⁓ what you're saying. I mean we've I've been in this business a long time and this kind of unbundling has just never been practical. It's never really been possible ⁓ to break up the pieces like this. And so we're very excited to be able to do it now. ⁓ and that's purely facilitated by ⁓ intelligence at the edge. ⁓
Josh Chandley: How many customers have used this tool already?
Jim: ⁓ it's a small number. I mean we we're we're just taking the you know we're just ⁓ opening the curtain on this ⁓ ultimately right now. So you know it's definitely a work in progress, but I I think you'll see that as we get the the private beta, the full buying pieces going in Q three, ⁓ you know, but we're gonna have a lot more to talk about there.
Josh Chandley: When the launch pod went live, I coined the term the new neutral. The idea that like neutral infrastructure becomes valuable partially because it's not vertically integrated. and now here you are vertically integrating. Help me understand, is is neutral a first principle of CloudX? Was it just a starting position or there a way that they can coexist? Cause from the outside, publishers aren't used to the idea of vertical integration and neutral perfectly coexisting.
Jim: Yeah, a lot of times vertically integrated it means that you have to take everything.
Josh Chandley: Yeah.
Jim: Right, and you you can't move outside that ecosystem. If you think about, you know, the Microsoft Windows ecosystem back in the days when it was like the, you know, evil empire, let's say. You had to use everything Microsoft, even if it wasn't the best product for the job you wanted. And if you tried to use something that wasn't Microsoft, you would basically lose access to all the rest of the Microsoft. So they had a ⁓ a way of sh putting the good software with the not so good software and forcing you to take everything. And that's the that was the thing.
Josh Chandley: Lock in. Yeah.
Jim: That's the lock-in factor. And that's what drove everybody nuts. And ultimately people did, you know, the internet allowed them to essentially move away en masse. And then, you know, Microsoft had to respond to that by breaking things up. And so saying you can take this part, you take this part, and you take this part, whatever works for you. So I think that's like a ⁓ that's where the negative connotation around vertical integration comes from. But we like to take more of like a Mac ecosystem approach where the things work really well together. ⁓ Maybe a poor example, actually, but ⁓ things really well together, but you don't have to use everything from Apple. But if you do, it works a lot better. And so that's what we're actually ⁓ proposing So i in our case, if you use the cloud X creative generation. ⁓ model making, targeting, and then you're buying from your peers on CloudX, all that's gonna be extremely well optimized by us and by the agents that are around it. And so you're gonna have a great experience. Are we forcing you to do that? No. You can take the creatives and go elsewhere. You can take the targeting model and go elsewhere. You can use just the CloudX supply side pieces. You can put us alongside some other platforms. We don't mind. You know, we we want you to use the software from us and the platform that we offer in the way that makes sense to And we've created, I think, a pricing model that makes that unbundling possible. So the tooling, like the creative generation stuff, is ultimately a token path monetization, right? But then if you want to connect to the network through us, there's a small toll. And that toll changes depending on how much of our software you're using versus how much of your software you're using. But that's ⁓ the fundamental idea. And we're not here to take a position in any particular piece of inventory. That's the other piece. Is that if A vertically integrated locked-in system is arbitraging you, that's where it gets very uncomfortable because the economic benefits of the vertical integration don't actually go back to you, right? They go back to the person who's putting that ecosystem together. So in this case, I feel like neutral is essentially saying the economic benefits of this vertical integration should go back to buyer and seller. Now I think agentic buying is going be great for buyers because they get out of this world where they're paying a a long-lived arbitrage fee. And I think it's going to be great for sellers because they can work directly with the buyers they want to work with and keep a lot more of that ⁓ economic value that's created by the transactions that we're facilitating. So I look at neutrality in terms of where are our are incentives aligned or misaligned. And I try to put our company and our software and our platform in a alignment with the participants on both sides of the network as much as I can. And the best way I I know how to do that is to facilitate transactions that will that want to happen in the marketplace and do it in a way where the benefits and the surplus of those transactions go back to those participants, not to us. And so that's why we keep the ⁓ the mark you know the margin we're taking on the revenue connected connections low ⁓ but also fixed. And we're not taking a position on either side of the trade.
Josh Chandley: I see, was a was a demand product always part of the plan?
Jim: Well, it was always something that people asked us for and we weren't sure if we would be able to pull it off. ⁓ because there's a lot to it, right? So if a publisher a publisher, you know well, I mean a publisher c publishers come to and they and they say, look, you know, selling ads is one piece of our business, but buying ads is another piece, you know, when can we use your software for that? Or, you know, I wanna buy from my buddy, you know, can I do that on the platform?
Josh Chandley: Honest. Yeah.
Jim: And at some at some point, you know, we decided that we want to do more for our customers, and that's our general philosophy here, is we want to help publishers monetize in a way that is, you know, that they feel proud of. But ⁓ but part of monetization is also growing your business too. So what if you could grow your business in a way you're proud of in the same way? That's kind of the question we asked. ⁓ but i it wasn't always part of the plan because again, ⁓ we weren't really sure if we could pull it off. There's a lot to it, as you know. ⁓
Josh Chandley: Yeah, what does that mean? Why weren't you sure you could pull it off?
Jim: Well think about like I mean you just said like playable creative generation hasn't been really done. You know, hasn't no one's really done that. And I think there's a lot of reasons for that. I I think a year ago you couldn't have done it. There's no way. The tools weren't there, the intelligence wasn't there, the cost profile wasn't there, it didn't make sense. ⁓ so we're now looking around the corner, we say, well, you know, these models, ⁓ and especially like video and image generation models and then code generation models are now cheap enough to create all these variations such that it actually makes some economic sense to pay a little bit of money to get those creators to then run. ⁓ there's no way I'd like again, w if if if things hadn't played out the way that they have, worth of sitting here in June 2025, I would say, you know, we're not not even close to having that be something that would make sense for people to do.
Josh Chandley: ⁓ say a agentic buying and the infra ⁓ they gain mass adoption. What stops you doing what every vertically integrated player eventually does? You know, once you own both the supply and the demand paths, ⁓ the incentive to favor your own shareholders structural, not optional. What prevents it? let's a agentic buying and the supply ⁓ they gain mass adoption. What stops you from what every vertically integrated player eventually does? once you own both the supply and the demand paths, ⁓ the incentive to favor your own shareholders is structural, not
Jim: Mm-hmm. I I I don't think that, you know, we don't look at ourselves as owners of of supply or demand. We look at ourselves as caretakers and facilitators of that. I think, you know, you may just have to take my word for it, but I I think you can also look you can also look at our track record and ⁓ you know, what we've done in the past and ⁓ understand that that is very philosophically you know, burned into our ethos that this is the right way to build a pla ⁓ to build a business is to create a platform where most of the value accrues to the ecosystem. And I think we can build a a bigger business that way. So I'm not saying I'm doing this in some altruistic fashion. I think we can actually grow the market substantially if we do this way. Because the more money that the participants in the ecosystem have, the more content that's going to be created, the more they can reinvest in their business, the more that they can grow. ⁓ and so, you know, a as a result, I think I want to make the ecosystem as big as possible. And if we are doing that and doing that faithfully and people choose us every day, then I think that's going to be the biggest ⁓ economic outcome ⁓ for us too, as well as the people involved.
Josh Chandley: I I I see where you're coming from. there's only so many companies where their first blog post is Hello World, then their second one is the trust problem in ad tech and how we solved it. I think y'all were y'all were straight to the point, right?
Jim: Yeah. Well, right. Exactly. And and look, we we know that we know that ⁓ that's a a big important piece is that people are it's a design win, right? So ⁓ it's a design question. I'm making an app and I'm baking this monetization partner into the core of that app in a way that's very, very difficult to remove. So trust is huge. and you know we've known that since the Mopub days. ⁓ so ⁓ that's why we wanted to talk about it. And ⁓ and you know the thing that we can now do with the secure enclaves and some of the modern technology that's there is we can actually do this attestation around the auction too. So we can make it such that demand partners don't have to just trust what's there. They can say, I want to look at the source code and know that you're running that source code. Just like when we did mopub SCK, the source code was out there, it was over. open source and people could look at it and they integrated the source code directly. So they knew that we weren't doing any nefarious things because we put it out there. You know, you don't have to trust me. You can look at it.
Josh Chandley: yeah, which is I think meaningfully different. Yeah, yeah. It
Jim: Well it wasn't possible. Yeah, you couldn't do it on the server side before. Now you can't, so we did.
Josh Chandley: it's interesting because what it creates is a scenario where nobody has any leverage over anybody. There's just performance. ⁓ and I can see that grow in the ecosystem. especially for ad monetized games, if the if the take rate starts to go down and and men become more neutral, that's going to be expressed directly as increased monetization, increased LTV. And if that's the case, UA scales.
Jim: I I view our is to create those components, create a platform that people can trust and believe in adopt easily in way that makes sense for them, and then have those connections happen ⁓ organically.
Josh Chandley: All right, let's assume neutrality holds. none of that matters ⁓ publishers, you know, don't make more money when CloudX takes a portion of their inventory. ⁓ to 20% uplifts, ⁓ sounds good. That's what moved me off of Mopub to max. but it depends entirely on the denominator. You know, anyone can show 20% uplifts. ⁓
Jim: Yeah, totally.
Josh Chandley: Against a single ad network. I'm assuming that uplift is a comparison to monetization with traditional mediators.
Jim: That that is a I was making X dollars before out of this particular app. And now I'm making one point two X. That's very much it's very much an apples to apples number. We're not trying to I'm not trying to give you a marketing number. ⁓ because that doesn't serve anybody, right? We want people to come over and
Josh Chandley: Yeah. Course, of course.
Jim: And so they are making more. So that that ⁓ you know, you never know if that's gonna be true at scale across every app, across every publisher who comes along. You know, the only way to know to learn that is to find out and to improve the system. And so, you know, we've launched a lot of publishers now. We've finding out across many different types of ⁓ apps and categories and approaches to monetization, that that does seem to be ⁓ pretty consistently true. And then at the same time, you know, we're also ⁓ just putting many, many more ⁓ demand buyers to increase competition ⁓ in the auction and drive more revenue through our pipes ⁓ to the publisher. And we just hired dear friend of mine, Janae Redmond, who ran the same function at Mopub and in fact took over the Mopub business writ large after it left left Twitter, ran it for many, many years, so she's you know, probably spent more time running this kind of business than me. ⁓ but we just hired her a month ago or so. And so she's you know, her job one is to basically increase liquidity to the publisher.
Josh Chandley: I get it. It does, of course depend on the publisher. But I I assume the majority of apps now on CloudX, either with T E two or all the way over, you know, were at some point using traditional mediation before.
Jim: ⁓ sure. I mean it'd be you know, the you have to keep it the state of the art. and you know, that's just the the way that it is. And ⁓ we've if we're gonna offer you something that we think is better, it better actually be better. Right. So you know, the onus is on us, we know that. ⁓ and so it's a it's a multi pronged set of activities that we do to deliver those results for the publisher ends up on their bottom line. But it doesn't just end up on their bottom line, like here's more revenue. It's also better user experience with the ads, better LTV retention. You know, you know, arguably you can you can pick the partners you want to work with. You know, a lot of people value that ⁓ quite a bit, actually. ⁓ so you know, and so there's a lot of different qualitative factors that are also driving people's enthusiastic embrace of the platform.
Josh Chandley: I think it's a fair point. I I think it it's easy to A B test and compare and see your your ARC DAO or gross revenue. Understanding how retention plays into actual LTV ⁓ is a challenging thing to A B test. ⁓ and so what you're saying is when when these customers are A B testing or just testing traditional mediation against a a CloudX like integrated approach approach, they're seeing 10, 20% uplift to their target metric.
Jim: Right, right. And and and and don't forget qual quality of life too. I mean, ⁓ let's the the the agent agenda stuff is very real. I mean you go to the tool and you say, Hey, I want to change this across all my apps and a couple of minutes later, ⁓ you know, Claude has done that. And you know, it's it's and then we're just doing more and more in that vein. You know, it you it worked pretty well when we launched it, now it works very well, I'd say. And the stuff that you know I just saw in a demo ⁓ this morning is ⁓ even further beyond that. Right. So the the you know you can you can kind of track it alongside the way that that Claude code has has progressed where it you know mostly worked but it would stop a lot and you have to keep turning the crank to eventually you just sort of give it something to do and then at the end of the day it has done that. You know, we're very much ⁓ you know just behind that same adoption curve in terms of the actual quality of life the publisher experience that they have.
Josh Chandley: So to create that 10, 20% uplift, you're charging ⁓ 5%, believe, for some, not all, but some impressions ⁓ that traditional mediators would otherwise charge for. That's revenue and auction data, right? Leaving their side of the table and landing on yours. So here's here's my question. Do you think of traditional mediators as competition, or do you just not see the relationship that way?
Jim: ⁓ we're well, we're we're charging the same five percent ⁓ to the buyer of the impression, the same as any other ⁓ similar platform would do. So that's ⁓ sort of the industry standard five percent fee goes to the SSP media or whatever you want to call it. of course you could do something that's more open source than you know ⁓ changing SDKs and that sort of thing. That might be a free tool. And so you know, I I it's hard for me to say if it's worth it or not. I I I think it's absolutely worth it, ⁓ would be my argument. ⁓ you know, certainly some people use more of a ⁓ traditional approach and that's that's certainly fine. But I I think if you wanna have ⁓ the advantages of actually ⁓ actual bidding, which I think is quite powerful. and being able to use ⁓ an auction mechanism, even if it's a multi-round auction mechanism, to force competitive pressure, market pricing and you know, actual data from the largest buyers and demand aggregators, it's well worth it to have a five percent fee coming out of that.
Josh Chandley: I agree, but I I still don't quite get it. ⁓ do you traditional mediators as competition, or is that just not how you see the relationship?
Jim: I I don't really ⁓ I don't I'm not really thinking about, you know, competition. I think in ad tech like everyone's kind of got a similar a lot of different similar components. So, you know, we work closely with with companies that have products that you could argue fill the same kind of gap, you know, to help monetize an ad unit. ⁓ but those are partners of ours on the demand side. So I it's it's we don't really I'm not really looking at ⁓ you know, competitors in the traditional sense like that.
Josh Chandley: Yeah, I I don't I don't view like a DSP or an SSP or a traditional SDK SSP ⁓ a as a competitor because ⁓ not really competing for that five percent fee. ⁓ what about traditional mediators? ⁓
Jim: Yeah, it's it's it it's just an alternative approach. so, you know, I I I don't I'm gonna think of it as competition.
Josh Chandley: Mm. Okay. Fair enough. ⁓ so if if supply keeps moving over and the demand side fills in, where does this go in 18 months? for CloudX and for traditional mediators? ⁓
Jim: Well, I think I already told you, you know, we're trying to make the ecosystem as large as possible. ⁓ we're trying to do ⁓ you know, we're trying to offer everything that a publisher asks us for that makes sense for us to offer that can take advantage of the network effect that's inherent in what we're trying to do and the data that you already have on the platform for convenience sake. And then this the intelligence that the agents start to accrue about your at your traffic and your business. You know, those are extremely valuable assets, I think. and so those are just gonna get more and more powerful as more data flows through, as more scale comes, ⁓ you know, to your point we got more demand coming through. The larger you are, the more the buy-side aggregators ⁓ spend making ⁓ more time they spend and effort they spend making sure things work properly. You know, so there's a lot of efficiencies to scale, ⁓ which we intend to go get on behalf of our publishers in that eighteen month period. But ⁓ you know, to me what we're trying to do is really just ⁓ build a little bit better of a mobile advertising ecosystem. And ⁓ you know, we're harnessing this power of the tools that we have today in our hands in order to do that.
Josh Chandley: So here's my last question. It's honestly astounding to me how many publishers are watching CloudX very closely, but aren't willing to be the first mover. I'm one of them. convince me that one day I'll make more money running my inventory through TEE 2.0 than gating it all behind MAX ⁓
Jim: Well, look, I I you know, you can talk to some of the people who are who have moved over the platform and, you know, take it from them. ⁓ I I think I've made the case what we're trying to do. But it doesn't really astound me to be quite honest with you. You know, it's ⁓ any adoption curve, anything new, is gonna have, you know, some portion of people who are so excited about the various things we're bringing to the table early and then don't have a lot of necessarily restrictions around what they can do or what they can't do or timing or things like that. You know, there's plenty of people who are, you know, have told us like we're right behind, you know, once we get X feature or once we get, you know, Y platform support, once you get this particular buyer, or, you know, once we get this quarter out of the way or this launch out of the way. You know, whatever you need to do to de-risk that. ⁓ you know, so I I I view it as like a a sort of inevitability that a lot of people that we talk to, this is a good fit for. And so eventually they will be on the platform. And I you know, the the way I I think about that is the same way we thought about it when I was working back at Google, you know, in my early ⁓ salad days, you know, would say, well, you know, somebody hasn't tried Google yet. That's fine. They eventually will. And when they do, it'll be so much better then. So if you haven't adopted it yet, that's fine because you know, tomorrow it's gonna be a better product. It's the the the velocity of the product is is that good. But it's ⁓ you know I i it it it's similar to me also too like like my Tesla self driving. You I'm at the point now where I literally I forgetting how to drive. I can't drive anymore. I've been doing FST for a year. And I love you know, but my my wife still twisting the wheel you know and mashing the pedals. And I'm so it's it you know they're just everyone has a different rate of adoption of ⁓ of the edge and it's it's sort of like your own ⁓ you know this special combination of constraints, risk and excitement about what the technology can do.
Josh Chandley: Well, Jim Payne, thank you for joining me today. congrats on the new launch. I I can't wait to see it live.
Jim: Thank you, Josh. We're very excited about ⁓ agentic buying and ⁓ I can't wait to see what the ⁓ publisher ecosystem does with it. So thank you for having me on.
Josh Chandley: Thanks, Jim.
Jim: Thank you, Josh. Appreciate you having me back on.
Josh Chandley: last time we talked four months ago. CloudX was just going live. ⁓ What's adoption looked like since then? and what kind of uplift are we seeing?
Jim: It's true. It's it's hard to imagine that it is only four months ago because ⁓ so much has happened in terms of both the product, the customers that are using it and ⁓ Everything we've learned about actually the system, now of it's in practice. So ⁓ you know, traction is is really ⁓ we're pretty happy with where we are. We've got hundreds of apps on the platform now, dozens of publishers. ⁓ it's a a lot of early adopters, people who have wanted to get their hands dirty with the technology and see what the agents could do for them. And you know, they've been h very helpful in helping in in informing our roadmap and and so we try to move as fast as our customers do ⁓ getting great product back up. Yeah.
Josh Chandley: What does an early adopter look like for CloudX?
Jim: Yeah, an early adopter looks like someone who wants the the cutting edge, you know, who is ⁓ probably pretty frustrated with, you know, the amount of work that they have to do to do ⁓ to pull off the a normal monetization routine. So you know the early adopters that we've seen ⁓ are people who have a lot of apps, a lot of line items, a lot of different partners, many different monetization strategies, a a very keen interest in experimentation. I think that's where our early platform has shined. But what's interesting is that Yeah, a large surface area and a complex operation. Yeah, that's really where I think CloudX agentic approach ⁓ probably provides the most value out of the gate.
Josh Chandley: Large surface area? Go ahead.
Jim: ⁓ but you know, we weren't really sure what was going to happen once we launched out of GA, quite frankly. and ⁓ you know, what has really been ⁓ extremely ⁓ great to see for us is that every publisher who's launched is seeing an uplift of 10 to 20 percent somewhere in there. Different categories a little bit less, some some categories a little bit more. ⁓ and that's just by virtue of doing a little bit better job of the monetization ⁓ that the agents can do and ⁓ the demand aggregation that we're bringing to the table. So that's just out of the gate that 10 to 20% lift ⁓ has been quite compelling for our early folks. They're extremely happy with it. You know as well as I do this small market, so you know the word of mouth spreads quickly. ⁓ And if you can get a 10 to 20% lift and do less work and have the amazing cloud X team helping you monetize your app, I think it's a pretty good value proposition. So we're getting a lot of ⁓ of additional interest from people you know down that adoption curve we might have ⁓ you know, wanted to see early returns first before the movie.
Josh Chandley: It's funny that number. In December 2020, I I A-B tested Max against Mopub for the first time. I forget the exact number, but our uplift was somewhere between 15 and 20%.
Jim: Yeah. It's it's I I don't know. Strange coincidence.
Josh Chandley: Funny how that happens. ⁓ yeah. Maybe. ⁓ What does integration look like? Because when a publisher integrates TEE 2.0, that doesn't actually mean they're moving 100% of their supply over to CloudX. So what does integrating actually mean?
Jim: Well so I'm glad you brought up T E two point ⁓ because ⁓ one point ⁓ was the notion ⁓ of a secure auction in which bids and information would not leak out of the ⁓ auction itself and the auction's ⁓ code that we were running ⁓ provably ⁓ in GitHub. So you go to GitHub, you see the code, you see that it's a fair auction, very simple. ⁓ we can actually provably show you that that's the code that we ran. ⁓ we can also make sure that bids ⁓ information ⁓ that might be attached to a bid ⁓ that didn't win the auction ⁓ stripped away ⁓ left in that enclave. ⁓ that was our first approach towards a secure ⁓ infrastructure for basically advertising programmatic auctions where you have a bunch of of and of counterparties who don't necessarily trust each other and want to make sure their signal isn't being misused. including by us, by the way. ⁓ And we did that in partnership with Meta in the earliest days. TE 2.0 takes that one step further ⁓ beyond just our own system and contemplates the idea that, well, publisher might have more than one supply platform for a variety of reasons. You know, some demand is only expressed in certain ways in certain platforms. If that's true, then you don't want to necessarily walk away from that incremental demand, but you also want to be able to control your experience with agents like we offer. ⁓
Josh Chandley: Have an opportunity to create some uplift somewhere.
Jim: Yeah, right, exactly. So how do you actually compare bids from sensitive buyers between two different systems? We realized that the best way to do that would be to s extend our secure enclave to and that's TE two point ⁓ There's a blog post about that we launched ⁓ a couple of weeks ago. And so basically that actual bid can survive through the auction in our system and then actually be compared against another platform bid in a secure fashion. So we don't have to do things like an aggregated bid where the bid is like the median price of all the bids that we've seen from a particular buyer. It's the actual bid itself.
Josh Chandley: Well well that's hard, right? Because some networks don't want you sharing that data.
Jim: That's exactly right. That's why we do in the secure enclave, so we can provably say this is not being emitted out of the transaction, but it's being used for the comparison. ⁓ so that's TE 2.0. So that allows a publisher to incrementally introduce the CloudX infrastructure into their monetization stack. ⁓ so rather than switch ⁓ 100%, they can ⁓ you know do different ⁓ types of hybrid installations. And we have a blog post about that too, if you want to read about that. But I should mention that actually ⁓ interestingly enough, you know, we we do this as just a convenience feature to help people de-risk their installation of our ⁓ a software. But we are seeing a lot of people actually move ⁓ completely full platform with us already because it's working so well.
Josh Chandley: my understanding ⁓ in my mind of TE too is it provides a secure way ⁓ of essentially not leaving your supply platform, whether that's a traditional mediator or a homebrew or something else. ⁓ And When CloudX can provide uplift by whatever you define uplift to be, they can flow in, but otherwise you're essentially guaranteed your baseline performance. Is that a good understanding?
Jim: ⁓ that's a that's a good understanding, yeah. I I think that is probably true.
Josh Chandley: Okay. Okay. Well let's get into why you're here. Before we walk through CloudX's new agency buying in detail, what is it? Is it is it an ad network? Is it a is it a DSP? What are we talking about?
Jim: Yeah, well I you know, it would be much easier to launch something and call it ⁓ a name that people know what it means. No word if we were either of those, I would have gone with those names. But we ⁓ we decided agentic buying because it's most descriptive of what it is actually. And so ⁓ you know, I think last time we talked ⁓ about this ⁓ concept of neutrality and I I know we're gonna get to that, ⁓ because I I do wanna
Josh Chandley: Yes. Yeah, it would be really helpful for me. Yeah.
Jim: you know, answer your question that you brought up at the beginning of the of the pod. But you know, ultimately What we hear from publishers is that, you know, they're on the sell side, but a lot times they're on the buy side too. So you're running a studio, you're running an app, you're ⁓ whatever it is, you're gonna be monetizing the traffic you have and you're gonna be using some of that revenue that you generate to go drive growth. And gro driving growth is buying traffic that works on your app. and in your portfolio of apps. And so to me it's always been you know an opportunity ⁓ writ large to take our vision of agents doing the work and making hard things easier for you and then and unbundling the ⁓ all the different components and giving you transparency, better pricing and more flexibility. That should apply as equally as well to the growth side of the equation as it should the monetization side. And conveniently it's the same it's the same customer. So you know, I look at this ⁓ our agentic buying system as basically a means to provide that growth lever for people in a way that feels very much a first class citizen of the platform that we've built. So you don't have to give it additional data because we already know. You don't have to give us ⁓ your ads and app ⁓ your sorry, your apps and ad units and you know all that information because we already have it. ⁓ you don't have to give us your traffic and geographic breakdowns because already have all that. We have have the transaction information. So So you can very simply create campaigns that leverage all those assets and go out into the CloudEx network and buy that traffic and bring it back to you. And that's what we're launching ⁓ in this agentic buying launch this week.
Josh Chandley: How does it make money? Most demand services on an arbitrage model. ⁓ I think a pessimistic but efficient description of the arbitrage is charge demand as much as you can, pay supply as little as you can, and your profit is the delta. ⁓ pretty normal, is that what you're doing?
Jim: ⁓ well that is descriptive of an arbitrage for sure. ⁓ and I think that would be ⁓ the ad network model is ⁓ you know you're buying low and you're selling high. ⁓ and and there's ⁓ it it ⁓ to to be fair, it's not a pure market making spread where you're buying and selling ⁓ the same exact thing. Ad networks take risk in doing it, right? So they're selling a CPI and they're buying on a CPM. So there's some risk inherent in that. There's also a lot of learning that you have to do as an ad network. So the ad network will figure out what works, they spend a lot of money, they're losing. Losing money for a while, there's a J curve that happens, and then all of a sudden they're making a lot of money. So the arbitrage is a sort of a way to pay back the experimentation period. ⁓ so it's not necessarily nefarious, it's just one way of doing things. ⁓ Yeah, ⁓ right, exactly. The problem is like the arbitrage can last very, very long time. You know, so if you're if you're a publisher and you're out there for years and years and years, like that profitable arbitrage is actually quite a lot.
Josh Chandley: No no. It is providing a service.
Jim: of money writ large. and so to be to answer your your prior question, we're not an ad network because we're not doing an arbitrage model. We're not taking any actual risk in the transaction and you know transforming CPM into CPI. We're helping the publisher do that on their own. Similarly, we're not a DSP, although we're probably more like a DSP than not. Why are we not a DSP? And that the answer to that is very simply we're not in programmatic. ⁓ so you know it in a way we're a demand-side platform, I guess you could say, right? Because we're helping a ⁓ publisher express their demand into some type of ecosystem. But I think DSP has a connotation that it's programmatic, it's RTB, and it's QPS. And it's bidded. And that's not exactly what we're doing. ⁓ so it's closer to a DSP though. and I I think importantly, what we're doing is our take rate is a flat take rate. And so that margin is a fixed margin and it's based on your connectivity into our ecosystem. And and so that's an important consideration of what ⁓ of how to think about our agentic buying tool.
Josh Chandley: So flat take rate, no arbitrage.
Jim: That's right. And it's an unbundling too. So we're not ⁓ expecting you to necessarily use us for everything. We're not expecting you to you can bring your own creative. ⁓ you know, it's designed to be a little bit more do-it-yourself than a other another platform. ⁓ And the reason do-it-yourself is because the agents are sitting there doing a lot of the work and optimization for you. And so our view is if the agents are doing the work, you should accrue most of the benefit of that. So ⁓ it we're not real we're not in it to make a large margin like you know ⁓ many ⁓ folks on that side are. And that's because you were trying to build this whole platform play out in a lower cost provider overall.
Josh Chandley: my my brain ⁓ wired for RTB, like real-time bidding. How is how is ⁓ I can't even imagine the idea that there's not a real time bid in an auction. how is targeting different in this agency world?
Jim: Yeah. It it's a great point because we do still run an auction for every impression and there's a bid. There is a bid for every auction, but it's not QPS, right? So it's you don't have an agent that's sitting there listening to every auction and submitting a bid on that. ⁓ it's ⁓ it's a bit more of a vertically integrated ⁓ model so that that we don't have that same supply path issue. ⁓ so it's not and it and we're not bidding out into the outside world either. So it's not RTB in that fashion.
Josh Chandley: Okay. Okay. Okay. I see.
Jim: It's it's a little bit more akin to ⁓ the what the like the the ⁓ ad CP, you know the the context protocol efforts are, which is more like how can I more conveniently or compactly express my demand at the at the other side of the transaction, such that I'm I'm not collecting all this QPS and putting a lot of server hardware against responding with a price and a certain amount of latency. That's kind of a brute force approach. So the agents allow you to then express that demand more conveniently at the edge. And so that's that's why it's more like a peer-to-peer model ⁓ than it is a ⁓ strictly a DSP.
Josh Chandley: That's new. What kind of campaigns are gonna be supported?
Jim: ⁓ everything you'd expect us to support ⁓ we'll eventually do. the it's a private beta th for Q three. So we're we've got a lot of work to do. you know, we're ⁓ no strangers to this. We know it's a lot of work. but ⁓ ultimately, you know, the the campaign types that that people want to ⁓ want to run we're gonna offer.
Josh Chandley: So you're thinking CPI to launch ROAS campaigns, ⁓ CPE campaigns to come and at some point in the future.
Jim: Yeah, exactly. The the more complicated ⁓ campaign types are predicated on data, so we just gotta get the flywheel spinning. So we'll do the simplest things first. Scale.
Josh Chandley: Roger. What's higher priority, ad ROS or IAP ROAS
Jim: Good question. you know AdROS is probably most p ⁓ differentiated, I'd argue, and then also most ⁓ inherent to our platform because we're working on the sell side too.
Josh Chandley: It it's also on pub and admon people hate to hear this, but it's easier. There's a lot more signal and it's a lot more up front. So yeah, it it is what it is. ⁓ yeah. Yeah, yeah, yeah. ⁓ and I I I think I may have already heard this, but ⁓ when a publisher buys through CloudX in this peer-to-peer fashion, whose supply ⁓
Jim: Yeah, yeah, that's right. Yeah. I is extremely sparse, but everyone sees that's right.
Josh Chandley: Are they buying?
Jim: So to ⁓ in order to launch this, we're it's all inside the CloudX ecosystem. ⁓ but you may have seen, you may not have seen, but we we donated a bunch of our mobile context protocol work to the IEB. And that is basically the notion of expressing a campaign on a foreign system. So internally we use a form of this, ⁓ more convenient form. But you can imagine someone with another SSB. implementing this protocol and our agentic buying tool could actually go load campaigns into that ⁓ third party system and keep them up to date and make sure they're all working and you know handle all the different payments and credit issues that come along with that. But there's nothing technically standing in the way of that. In fact we'd like it to happen ⁓ that way. but ⁓ you know we felt like we needed to get the ball rolling on our own ⁓ network first.
Josh Chandley: So a GenSec buying is a fee-based way of expressing your demand. There's no arbitrage involved. And right now it's through CloudX, but in theory, it could be buying from any supply that integrates that.
Jim: Absolutely, yeah. ⁓ I don't think you know, we wouldn't ⁓ artificially stop ourselves ⁓ from serving the publisher.
Josh Chandley: Okay. I guess my last question is, why are you offering this product?
Jim: Well, it's the number one thing that people wanna want to see from us. So they love the agen Yeah, absolutely. They want the they love the agentic selling and they are
Josh Chandley: ⁓ really? Yeah.
Jim: Their first request is, can you help me with growth too? Because I want to do this exact thing. I want to unbundle all the functions that I get from an ad network and kind of take control of each of them in a way that is matches my business and my, you know, taste and sensibility and everything. So that's all the way from creative generation to targeting and modeling to the what the ad experience looks like. And then actually, you know, you reap the rewards ideally if you do it well, because you understand your traffic best. And you can bring your own first party data to the table, integrate that with our cloud agent and it all works to make a better model and a better set of creatives for you. And you don't have to trust a third party really to do that. If that makes sense. ⁓ so
Josh Chandley: Who who owns that who who owns that data? Do do you o own that go ahead.
Jim: Yeah, so good. We absolutely do not. So we don't own that data. That's basically data that you introduce into the agentic flow. And then it creates something that's a model that can then be used by our agentic buyer to express that demand into the network. Similarly, the creative ⁓ generation stuff that's a token path ⁓ opportunity for us. So we basically we we charge you for token usage, but at the end of that, you own the creatives that come out of it. And that can be image, video, and playable. And in fact, ⁓ that creative generation tool suite. Is what we're launching for anybody to take a look at and use, ⁓ even while the network piece is in private beta. So that's a anyone can can come and use that. That's a that's a more public beta offering. And so you can come in and create all these playables for all the apps that you already have in our system. You then generate a thousand different variations. You can go use that anywhere you'd like.
Josh Chandley: Fascinating. This is even for gaming, because I haven't seen anybody really succeed at like a Gensec playables for gaming yet.
Jim: Yeah, well, ⁓ you know, ⁓ hopefully it succeeds. I think it'll be look it's not I don't think it's a business, but it's a it's a piece of functionality that you need in order to achieve the growth goals that you're looking for. And the fact that it's vertically integrated with the rest of our system just it it's a it's ultimately a convenience, but it's also a tighter optimization loop. Right. So the agent can wake up in the morning and say, I'm gonna try a thousand variations of this playable. Puts it out there, sees it works and What doesn't work very quickly and then comes back and makes iterations. And that all just happens automatically while you're not really paying attention to it. But at the end of the day, those creatives, all that data, what worked and what didn't work, and importantly the economic benefits of understanding your own buying and finding those arbitrages, those come back to you. So remember we talked about the ad network J curve. You know, you may be spending a little bit more upfront to get the creative exactly right. But then you reap the rewards of that efficiency over the long run as you spend against that.
Josh Chandley: I mean, man, you're still in the dream, a Gensic creative production, right? So the AI makes the ad, it measures the result and it learns from that and then makes a better ad. ⁓ how f are you gonna walk the whole way down that path? Go ahead.
Jim: I mean this is a ⁓ yeah, and so this this is not theoretical, it's not hypothetical, it's this is something we can offer you now. And ⁓ and I I understand ⁓ what you're saying. I mean we've I've been in this business a long time and this kind of unbundling has just never been practical. It's never really been possible ⁓ to break up the pieces like this. And so we're very excited to be able to do it now. ⁓ and that's purely facilitated by ⁓ intelligence at the edge. ⁓
Josh Chandley: How many customers have used this tool already?
Jim: ⁓ it's a small number. I mean we we're we're just taking the you know we're just ⁓ opening the curtain on this ⁓ ultimately right now. So you know it's definitely a work in progress, but I I think you'll see that as we get the the private beta, the full buying pieces going in Q three, ⁓ you know, but we're gonna have a lot more to talk about there.
Josh Chandley: When the launch pod went live, I coined the term the new neutral. The idea that like neutral infrastructure becomes valuable partially because it's not vertically integrated. and now here you are vertically integrating. Help me understand, is is neutral a first principle of CloudX? Was it just a starting position or there a way that they can coexist? Cause from the outside, publishers aren't used to the idea of vertical integration and neutral perfectly coexisting.
Jim: Yeah, a lot of times vertically integrated it means that you have to take everything.
Josh Chandley: Yeah.
Jim: Right, and you you can't move outside that ecosystem. If you think about, you know, the Microsoft Windows ecosystem back in the days when it was like the, you know, evil empire, let's say. You had to use everything Microsoft, even if it wasn't the best product for the job you wanted. And if you tried to use something that wasn't Microsoft, you would basically lose access to all the rest of the Microsoft. So they had a ⁓ a way of sh putting the good software with the not so good software and forcing you to take everything. And that's the that was the thing.
Josh Chandley: Lock in. Yeah.
Jim: That's the lock-in factor. And that's what drove everybody nuts. And ultimately people did, you know, the internet allowed them to essentially move away en masse. And then, you know, Microsoft had to respond to that by breaking things up. And so saying you can take this part, you take this part, and you take this part, whatever works for you. So I think that's like a ⁓ that's where the negative connotation around vertical integration comes from. But we like to take more of like a Mac ecosystem approach where the things work really well together. ⁓ Maybe a poor example, actually, but ⁓ things really well together, but you don't have to use everything from Apple. But if you do, it works a lot better. And so that's what we're actually ⁓ proposing So i in our case, if you use the cloud X creative generation. ⁓ model making, targeting, and then you're buying from your peers on CloudX, all that's gonna be extremely well optimized by us and by the agents that are around it. And so you're gonna have a great experience. Are we forcing you to do that? No. You can take the creatives and go elsewhere. You can take the targeting model and go elsewhere. You can use just the CloudX supply side pieces. You can put us alongside some other platforms. We don't mind. You know, we we want you to use the software from us and the platform that we offer in the way that makes sense to And we've created, I think, a pricing model that makes that unbundling possible. So the tooling, like the creative generation stuff, is ultimately a token path monetization, right? But then if you want to connect to the network through us, there's a small toll. And that toll changes depending on how much of our software you're using versus how much of your software you're using. But that's ⁓ the fundamental idea. And we're not here to take a position in any particular piece of inventory. That's the other piece. Is that if A vertically integrated locked-in system is arbitraging you, that's where it gets very uncomfortable because the economic benefits of the vertical integration don't actually go back to you, right? They go back to the person who's putting that ecosystem together. So in this case, I feel like neutral is essentially saying the economic benefits of this vertical integration should go back to buyer and seller. Now I think agentic buying is going be great for buyers because they get out of this world where they're paying a a long-lived arbitrage fee. And I think it's going to be great for sellers because they can work directly with the buyers they want to work with and keep a lot more of that ⁓ economic value that's created by the transactions that we're facilitating. So I look at neutrality in terms of where are our are incentives aligned or misaligned. And I try to put our company and our software and our platform in a alignment with the participants on both sides of the network as much as I can. And the best way I I know how to do that is to facilitate transactions that will that want to happen in the marketplace and do it in a way where the benefits and the surplus of those transactions go back to those participants, not to us. And so that's why we keep the ⁓ the mark you know the margin we're taking on the revenue connected connections low ⁓ but also fixed. And we're not taking a position on either side of the trade.
Josh Chandley: I see, was a was a demand product always part of the plan?
Jim: Well, it was always something that people asked us for and we weren't sure if we would be able to pull it off. ⁓ because there's a lot to it, right? So if a publisher a publisher, you know well, I mean a publisher c publishers come to and they and they say, look, you know, selling ads is one piece of our business, but buying ads is another piece, you know, when can we use your software for that? Or, you know, I wanna buy from my buddy, you know, can I do that on the platform?
Josh Chandley: Honest. Yeah.
Jim: And at some at some point, you know, we decided that we want to do more for our customers, and that's our general philosophy here, is we want to help publishers monetize in a way that is, you know, that they feel proud of. But ⁓ but part of monetization is also growing your business too. So what if you could grow your business in a way you're proud of in the same way? That's kind of the question we asked. ⁓ but i it wasn't always part of the plan because again, ⁓ we weren't really sure if we could pull it off. There's a lot to it, as you know. ⁓
Josh Chandley: Yeah, what does that mean? Why weren't you sure you could pull it off?
Jim: Well think about like I mean you just said like playable creative generation hasn't been really done. You know, hasn't no one's really done that. And I think there's a lot of reasons for that. I I think a year ago you couldn't have done it. There's no way. The tools weren't there, the intelligence wasn't there, the cost profile wasn't there, it didn't make sense. ⁓ so we're now looking around the corner, we say, well, you know, these models, ⁓ and especially like video and image generation models and then code generation models are now cheap enough to create all these variations such that it actually makes some economic sense to pay a little bit of money to get those creators to then run. ⁓ there's no way I'd like again, w if if if things hadn't played out the way that they have, worth of sitting here in June 2025, I would say, you know, we're not not even close to having that be something that would make sense for people to do.
Josh Chandley: ⁓ say a agentic buying and the infra ⁓ they gain mass adoption. What stops you doing what every vertically integrated player eventually does? You know, once you own both the supply and the demand paths, ⁓ the incentive to favor your own shareholders structural, not optional. What prevents it? let's a agentic buying and the supply ⁓ they gain mass adoption. What stops you from what every vertically integrated player eventually does? once you own both the supply and the demand paths, ⁓ the incentive to favor your own shareholders is structural, not
Jim: Mm-hmm. I I I don't think that, you know, we don't look at ourselves as owners of of supply or demand. We look at ourselves as caretakers and facilitators of that. I think, you know, you may just have to take my word for it, but I I think you can also look you can also look at our track record and ⁓ you know, what we've done in the past and ⁓ understand that that is very philosophically you know, burned into our ethos that this is the right way to build a pla ⁓ to build a business is to create a platform where most of the value accrues to the ecosystem. And I think we can build a a bigger business that way. So I'm not saying I'm doing this in some altruistic fashion. I think we can actually grow the market substantially if we do this way. Because the more money that the participants in the ecosystem have, the more content that's going to be created, the more they can reinvest in their business, the more that they can grow. ⁓ and so, you know, a as a result, I think I want to make the ecosystem as big as possible. And if we are doing that and doing that faithfully and people choose us every day, then I think that's going to be the biggest ⁓ economic outcome ⁓ for us too, as well as the people involved.
Josh Chandley: I I I see where you're coming from. there's only so many companies where their first blog post is Hello World, then their second one is the trust problem in ad tech and how we solved it. I think y'all were y'all were straight to the point, right?
Jim: Yeah. Well, right. Exactly. And and look, we we know that we know that ⁓ that's a a big important piece is that people are it's a design win, right? So ⁓ it's a design question. I'm making an app and I'm baking this monetization partner into the core of that app in a way that's very, very difficult to remove. So trust is huge. and you know we've known that since the Mopub days. ⁓ so ⁓ that's why we wanted to talk about it. And ⁓ and you know the thing that we can now do with the secure enclaves and some of the modern technology that's there is we can actually do this attestation around the auction too. So we can make it such that demand partners don't have to just trust what's there. They can say, I want to look at the source code and know that you're running that source code. Just like when we did mopub SCK, the source code was out there, it was over. open source and people could look at it and they integrated the source code directly. So they knew that we weren't doing any nefarious things because we put it out there. You know, you don't have to trust me. You can look at it.
Josh Chandley: yeah, which is I think meaningfully different. Yeah, yeah. It
Jim: Well it wasn't possible. Yeah, you couldn't do it on the server side before. Now you can't, so we did.
Josh Chandley: it's interesting because what it creates is a scenario where nobody has any leverage over anybody. There's just performance. ⁓ and I can see that grow in the ecosystem. especially for ad monetized games, if the if the take rate starts to go down and and men become more neutral, that's going to be expressed directly as increased monetization, increased LTV. And if that's the case, UA scales.
Jim: I I view our is to create those components, create a platform that people can trust and believe in adopt easily in way that makes sense for them, and then have those connections happen ⁓ organically.
Josh Chandley: All right, let's assume neutrality holds. none of that matters ⁓ publishers, you know, don't make more money when CloudX takes a portion of their inventory. ⁓ to 20% uplifts, ⁓ sounds good. That's what moved me off of Mopub to max. but it depends entirely on the denominator. You know, anyone can show 20% uplifts. ⁓
Jim: Yeah, totally.
Josh Chandley: Against a single ad network. I'm assuming that uplift is a comparison to monetization with traditional mediators.
Jim: That that is a I was making X dollars before out of this particular app. And now I'm making one point two X. That's very much it's very much an apples to apples number. We're not trying to I'm not trying to give you a marketing number. ⁓ because that doesn't serve anybody, right? We want people to come over and
Josh Chandley: Yeah. Course, of course.
Jim: And so they are making more. So that that ⁓ you know, you never know if that's gonna be true at scale across every app, across every publisher who comes along. You know, the only way to know to learn that is to find out and to improve the system. And so, you know, we've launched a lot of publishers now. We've finding out across many different types of ⁓ apps and categories and approaches to monetization, that that does seem to be ⁓ pretty consistently true. And then at the same time, you know, we're also ⁓ just putting many, many more ⁓ demand buyers to increase competition ⁓ in the auction and drive more revenue through our pipes ⁓ to the publisher. And we just hired dear friend of mine, Janae Redmond, who ran the same function at Mopub and in fact took over the Mopub business writ large after it left left Twitter, ran it for many, many years, so she's you know, probably spent more time running this kind of business than me. ⁓ but we just hired her a month ago or so. And so she's you know, her job one is to basically increase liquidity to the publisher.
Josh Chandley: I get it. It does, of course depend on the publisher. But I I assume the majority of apps now on CloudX, either with T E two or all the way over, you know, were at some point using traditional mediation before.
Jim: ⁓ sure. I mean it'd be you know, the you have to keep it the state of the art. and you know, that's just the the way that it is. And ⁓ we've if we're gonna offer you something that we think is better, it better actually be better. Right. So you know, the onus is on us, we know that. ⁓ and so it's a it's a multi pronged set of activities that we do to deliver those results for the publisher ends up on their bottom line. But it doesn't just end up on their bottom line, like here's more revenue. It's also better user experience with the ads, better LTV retention. You know, you know, arguably you can you can pick the partners you want to work with. You know, a lot of people value that ⁓ quite a bit, actually. ⁓ so you know, and so there's a lot of different qualitative factors that are also driving people's enthusiastic embrace of the platform.
Josh Chandley: I think it's a fair point. I I think it it's easy to A B test and compare and see your your ARC DAO or gross revenue. Understanding how retention plays into actual LTV ⁓ is a challenging thing to A B test. ⁓ and so what you're saying is when when these customers are A B testing or just testing traditional mediation against a a CloudX like integrated approach approach, they're seeing 10, 20% uplift to their target metric.
Jim: Right, right. And and and and don't forget qual quality of life too. I mean, ⁓ let's the the the agent agenda stuff is very real. I mean you go to the tool and you say, Hey, I want to change this across all my apps and a couple of minutes later, ⁓ you know, Claude has done that. And you know, it's it's and then we're just doing more and more in that vein. You know, it you it worked pretty well when we launched it, now it works very well, I'd say. And the stuff that you know I just saw in a demo ⁓ this morning is ⁓ even further beyond that. Right. So the the you know you can you can kind of track it alongside the way that that Claude code has has progressed where it you know mostly worked but it would stop a lot and you have to keep turning the crank to eventually you just sort of give it something to do and then at the end of the day it has done that. You know, we're very much ⁓ you know just behind that same adoption curve in terms of the actual quality of life the publisher experience that they have.
Josh Chandley: So to create that 10, 20% uplift, you're charging ⁓ 5%, believe, for some, not all, but some impressions ⁓ that traditional mediators would otherwise charge for. That's revenue and auction data, right? Leaving their side of the table and landing on yours. So here's here's my question. Do you think of traditional mediators as competition, or do you just not see the relationship that way?
Jim: ⁓ we're well, we're we're charging the same five percent ⁓ to the buyer of the impression, the same as any other ⁓ similar platform would do. So that's ⁓ sort of the industry standard five percent fee goes to the SSP media or whatever you want to call it. of course you could do something that's more open source than you know ⁓ changing SDKs and that sort of thing. That might be a free tool. And so you know, I I it's hard for me to say if it's worth it or not. I I I think it's absolutely worth it, ⁓ would be my argument. ⁓ you know, certainly some people use more of a ⁓ traditional approach and that's that's certainly fine. But I I think if you wanna have ⁓ the advantages of actually ⁓ actual bidding, which I think is quite powerful. and being able to use ⁓ an auction mechanism, even if it's a multi-round auction mechanism, to force competitive pressure, market pricing and you know, actual data from the largest buyers and demand aggregators, it's well worth it to have a five percent fee coming out of that.
Josh Chandley: I agree, but I I still don't quite get it. ⁓ do you traditional mediators as competition, or is that just not how you see the relationship?
Jim: I I don't really ⁓ I don't I'm not really thinking about, you know, competition. I think in ad tech like everyone's kind of got a similar a lot of different similar components. So, you know, we work closely with with companies that have products that you could argue fill the same kind of gap, you know, to help monetize an ad unit. ⁓ but those are partners of ours on the demand side. So I it's it's we don't really I'm not really looking at ⁓ you know, competitors in the traditional sense like that.
Josh Chandley: Yeah, I I don't I don't view like a DSP or an SSP or a traditional SDK SSP ⁓ a as a competitor because ⁓ not really competing for that five percent fee. ⁓ what about traditional mediators? ⁓
Jim: Yeah, it's it's it it's just an alternative approach. so, you know, I I I don't I'm gonna think of it as competition.
Josh Chandley: Mm. Okay. Fair enough. ⁓ so if if supply keeps moving over and the demand side fills in, where does this go in 18 months? for CloudX and for traditional mediators? ⁓
Jim: Well, I think I already told you, you know, we're trying to make the ecosystem as large as possible. ⁓ we're trying to do ⁓ you know, we're trying to offer everything that a publisher asks us for that makes sense for us to offer that can take advantage of the network effect that's inherent in what we're trying to do and the data that you already have on the platform for convenience sake. And then this the intelligence that the agents start to accrue about your at your traffic and your business. You know, those are extremely valuable assets, I think. and so those are just gonna get more and more powerful as more data flows through, as more scale comes, ⁓ you know, to your point we got more demand coming through. The larger you are, the more the buy-side aggregators ⁓ spend making ⁓ more time they spend and effort they spend making sure things work properly. You know, so there's a lot of efficiencies to scale, ⁓ which we intend to go get on behalf of our publishers in that eighteen month period. But ⁓ you know, to me what we're trying to do is really just ⁓ build a little bit better of a mobile advertising ecosystem. And ⁓ you know, we're harnessing this power of the tools that we have today in our hands in order to do that.
Josh Chandley: So here's my last question. It's honestly astounding to me how many publishers are watching CloudX very closely, but aren't willing to be the first mover. I'm one of them. convince me that one day I'll make more money running my inventory through TEE 2.0 than gating it all behind MAX ⁓
Jim: Well, look, I I you know, you can talk to some of the people who are who have moved over the platform and, you know, take it from them. ⁓ I I think I've made the case what we're trying to do. But it doesn't really astound me to be quite honest with you. You know, it's ⁓ any adoption curve, anything new, is gonna have, you know, some portion of people who are so excited about the various things we're bringing to the table early and then don't have a lot of necessarily restrictions around what they can do or what they can't do or timing or things like that. You know, there's plenty of people who are, you know, have told us like we're right behind, you know, once we get X feature or once we get, you know, Y platform support, once you get this particular buyer, or, you know, once we get this quarter out of the way or this launch out of the way. You know, whatever you need to do to de-risk that. ⁓ you know, so I I I view it as like a a sort of inevitability that a lot of people that we talk to, this is a good fit for. And so eventually they will be on the platform. And I you know, the the way I I think about that is the same way we thought about it when I was working back at Google, you know, in my early ⁓ salad days, you know, would say, well, you know, somebody hasn't tried Google yet. That's fine. They eventually will. And when they do, it'll be so much better then. So if you haven't adopted it yet, that's fine because you know, tomorrow it's gonna be a better product. It's the the the velocity of the product is is that good. But it's ⁓ you know I i it it it's similar to me also too like like my Tesla self driving. You I'm at the point now where I literally I forgetting how to drive. I can't drive anymore. I've been doing FST for a year. And I love you know, but my my wife still twisting the wheel you know and mashing the pedals. And I'm so it's it you know they're just everyone has a different rate of adoption of ⁓ of the edge and it's it's sort of like your own ⁓ you know this special combination of constraints, risk and excitement about what the technology can do.
Josh Chandley: Well, Jim Payne, thank you for joining me today. congrats on the new launch. I I can't wait to see it live.
Jim: Thank you, Josh. We're very excited about ⁓ agentic buying and ⁓ I can't wait to see what the ⁓ publisher ecosystem does with it. So thank you for having me on.
Josh Chandley: Thanks, Jim.
