Episode 176: Alex Chatfield on How Influencers are Leveraging Data Technology

0:10 All right, welcome to the architecture podcast. This is Ari Pero. I'm here with Eric Franchie. How you doing, Eric? >> Doing well. How are you, my friend? >> I'm very good. So, Eric and I have very complicated schedules in the weeks leading up to KN. So, we're we're gonna mix it up a little bit. Like I think we're I'm doing some interviews without Eric and then Eric and I are doing the news and maybe uh I think I'll be able to do the news the next couple weeks, but if I can't Eric will do the news by himself. We're all adults. All the listeners are adults. They can handle it. A little change, little mixup.

0:39 They'll be okay. Don't you think? >> I don't know, man. Um [laughter] I think some of the people just love the the back and forth, the banter, the you know the our show, but we'll we'll do our best. Well, today uh I had the pleasure of uh interviewing Alex Chatfield, who many of you probably know, longtime um person at ex at at App Nexus, who I worked with closely with during my time there. Um and he started a company. It was called Advolution, but now it has a better name called Endorseable. And uh it's it's kind of in the middle of influencer, but with like a adte data overlay. Um so it's pretty pretty interesting stuff. I learned a lot from the conversation. Um, and so we'll hear that and then Eric and I will come back with the news of the week where we've got the liftoff IPO with the Walmart announcements uh and some some just other pretty interesting things that are going on in the industry. So, uh, should be a good conversation.

1:39 >> Yes. Um, a geo company was acquired um, and the names have completely escaped Arian. Stay tuned for it. >> It'll be a mystery to us when we find out who got acquired. All right. Stay tuned for Alex Chadfield. Uh, so congrats on getting out of App Nexus Microsoft after a long stint and doing your own thing. You're the co-founder and president of this company we're calling Endorseable, new name. Um, tell me a little bit about it.

2:10 >> Uh, well, yeah, thanks Ari. took uh almost 15 years to finally leave the nest, but I've spread my wings and I'm trying to fly here on my my own thing, which has been really an exciting journey. It's evolved significantly since since the beginning, I think, kind of like any any good startup. But, uh, you know, what we're building is a what we call phantom intelligence engine. >> Okay. What does that mean? Fandm intelligence agent. So, this is like influencer. It's called endorseable. So, that kind of tells us a little bit, but give us the elevator pitch.

2:40 >> Yeah. Really what we do is we we partner and enable talent. So actors, athletes, influencers, creators to translate their social capital into new lines of growth and to empower brands to build both more authentic and more effective endorsement partnerships. >> Right. That makes sense. I mean, there's statistics out there that, you know, uh, influencers are the number one source of of for Gen Z finding new brands and they're trusted more than paid media and all that sort of stuff. Um, so is this an influencer play because there have been other influencer plays. Influential got acquired, bunch of other companies like that. How is this how is this different?

3:20 >> Yeah, it's a it's a great question that the influencer market is very hot as you said, growing, you know, 30 plus% year-over-year depending on what you're what you're looking at. It's already a sizable market. Uh what what's unique about our play is that we're really focusing on fandom and the fans of the spokesperson or of the influencer where traditionally the focus has really just been on what we would call talent data meaning you know the own personal preferences interests of uh of the talent themselves the products and services they personally use.

3:54 >> Okay. So so it's not Taylor Swift saying she likes a certain soda brand. It's Taylor Swift telling her her minions to uh like a certain soda brand. >> It's it's uh the brand partnering with Taylor Swift because they now know that her fans actually are the right consumer set for their product. >> Okay. So, so it's sort of uh you're profiling or uh estimating who who follows Taylor Swift and what why that matters. >> Correct. [clears throat] Yeah. We are we're effectively identifying uh you know celebrity fandoms you know outside of social because that's where we can actually do the identification uh provide real insights to talent for the very first time so they actually know who their fans are. When I got into this I I actually didn't really appreciate how little talent knew about their own fan base. I mean the only data they've ever been given is age and gender.

4:45 >> Yeah. So what like if you're a if you're I guess a musician filling up arenas, do you have a database? Like what using CRM? What's the current what are you doing? >> Generally not. I mean musicians tend to be the furthest ahead because they do sell tickets and merch, but often they're selling tickets through third party platforms like Ticketmaster or StubHub or they're selling merch through thirdparty platforms and so they they don't have the database. Those platforms have the database. Uh and so yeah, generally speaking, talent doesn't have their own direct relationship with their own fans. They've been sort of unconsciously disintermediated from them. Uh because they use all these platforms to interact interact with them.

5:27 >> And that's probably worse for sports, right? Like an average sports player probably has kind of no idea. >> It's even worse for sports because athletes also have such a short career that their moment to capture and build that that database is is much shorter. And I mean, if you're an 18-year-old or a 21-year-old and you're hitting the pros, like you're very focused on excelling at at the sport and you're probably not thinking about how do I capture this moment to make sure I know who my fans are so that I can, you know, use those relationships and that and and you know, that trust uh to my advantage in the future.

6:02 >> Okay, now this is a deep cut, so I'm not sure I remember this properly, but aren't you a dead head? Like, weren't you the guy who created that App Nexus's Grateful Dead sticker? [laughter] >> That was not me. >> That was not you. Was that like Rob Huzzan or someone like that? You know what I'm talking about. >> That that would that would track with Robison. If you said Dave Matthews, I'd say it was it was Jonathan Mars.

6:23 >> Dave Matthews. Cuz the reason I [clears throat] brought it up is because like the Grateful Dead were famous for having like the best database in in entertainment, right? um they they knew who their dead heads were and they they would email them and mail physical mail to them and uh I guess fish also. So what's the game plan for a modern um star to create a database to connect with the fans? What do you just cobble it together with like Patreon and Substack and a bunch of other stuff or is there a good way to do it?

6:51 >> Great question. So, you know, if you think about how talent interacts with their fans today, of course, there's social media, but that's really walled off and and not a playable space for us. But there's also biolink companies. So, you know, folks like Linkree or Churnney or Beacons. These these folks have links on the homepages of the social sites that click out of social. Uh >> link and bio. Link and bio. I've heard this before. >> Yep. uh you have you know ticketing platforms, merchandising platforms, you have ACR vendors or companies right which who can tell us these are the television screens that have consumed the specific celebrities content uh and so there's actually significant fan signal it's just highly fragmented across you know a number of different players and so that's that's really where we're focused on on bringing that together.

7:45 >> Oh wow. Okay. So so let's talk about Lincoln Bios. is an ad tech podcast more or less. We try to deny it. We try to be like, you know, the next Joe Rogan, but really it's a bunch of ad tech people listening to this. So, if I if I'm a let's say I'm a influencer or or a star of some kind and I'm I'm posting Instagram and I say, you know, go to my store link in bio, right? Um the user clicks out of the Instagram app and then it redirects through the the vendor that's provided link in bio. they create a first-party cookie and then match it up using live ramp or someone like that and suddenly you have a pretty good idea which which like household according to one of the major data spines. Is that like what we're talking about here?

8:26 >> Yep, that's exactly right. The the link in bios also do a pretty good job of capturing email. So you could subscribe to a LinkedIn bio page. So obviously a more persistent identifier for advertising purposes uh than just a cookie. >> Oh wow. Okay. So, what's the minimum size at which it's worth doing this? Should every like, you know, second string cornerback on the New Orleans Saints have the Lincoln bio and and work with Endorable and be collecting a database?

8:52 >> Uh, it's probably not it's not probably not at that level, but you know, anybody that partners with with brands today or has aspirations to partner with brands today should be doing this because this is effectively the currency. It's really like we like to tell talent your fandom is your most valuable asset and most of you don't own it. >> Okay, let's play let's play this out, right? I'm a brand. I'm like, you know, I'm selling something um you know, some sort of skin cream or whatever. And my my agency identifies some influencers. I mean, I keep saying influencers. I know that's not exactly the way you think about it, so I apologize for that. But um and says, "Hey, we want to collab."

9:30 Now this person who is got a following now has assets that are valuable. They have a list. They have a database. They have capabilities. How's that change the dynamic with the brand versus you know kind of the current way of doing things? >> Uh so really it changes it in two ways. One the the talent can better position themselves with the brand in the first place, right? like, "Oh, yes. This makes a lot of sense. Not only because I love your product, but because my fans are the right customer set for you, they like your product. They're the right age demo, household income, whatever it might be for your product." So, it provides intelligence upfront to get the right partnership in place. And then instead of, you know, relying exclusively on, you know, organic reach of a social post, and we're talking about your really traditional influencer marketing, the brand can now also amplify what they're doing outside of social. So, we'll we'll post the the commercial, the creative on social. Uh, we can now also target the fans of my spokesperson while they're streaming TV, playing mobile games, reading articles, right? is anywhere across digital media.

10:41 >> Oh, I knew this was going to end up in programmatic advertising. I just had a feeling like the gravity well of programmatic advertising was going to be the end of this >> 15 [laughter] years in programmatic art. You can't take it out of me. >> All right. So, uh, so now you have the star Ariana Grande saying like, "Hey, use this cream. It makes your skin better." Then she's shooting off an email to her minions being like, "Hey, use this cream. I love it." And then the whammy is I'm watching uh the latest episode of something on Peacock and Ariana Grande is in commercial and I have no idea why I'm seeing that. Is that kind of the situation?

11:17 >> Y that Hey man, I like it. >> And it's and it's, [laughter] you know, it's worth saying that this is increasingly important for brands because just how how little organic reach talent now gets inside of social. So yeah, like oh yeah, I mean organic reach has gone down 15 20% year-over-year over year for the last five years. >> You know, the new statistics are, you know, something like sub 5% of my followers on Meta will actually see the post that I put out there. And so if a brand actually wants to reach the fans of their spokesperson, this becomes the way to ensure that that happens and to be able to do that across the big screen and you know all screens, not just primarily a small screen. Could we uplevel the conversation a little bit and and give me a sense of like what is what happens in the negotiation process?

12:10 What what who at the agency is thinking about this? Is are they influencer specialists or are they mainstream? are what platforms are they using to screen which influencers and then what does the negotiation look like and who brings data from either side? >> Yeah. Yeah. Great question. So, uh yeah, in some cases a brand will hire a agency partner to source talent. Uh that individual or the brand themselves will go out and and try to identify you talent that they should partner with. Some brands are using like chat GPT queries to identify influencers that they should partner with. Some brands or or agencies are licensing software that uh you know really only looks at the what we again call talent data. So the the personal preferences of of uh talent themselves. Uh and so they they you know effectively come up with a list of okay here are authentic talent partners for for me and then they start the outreach process which is a very manual process and you know very timely >> not to interrupt but uh haven't meta and Google been investing in this part of the value chain I I think I reported on it like a month ago. Yeah, the new fronts. There was some announcements from YouTube specifically and I didn't see the one about about Meta, but uh >> I think it was part of the Manis the Manis AI integration, which I don't know if that's still happening because of the Chinese government, but like they're they were showing examples of using Manis to find influencers on Instagram and stuff like that.

13:44 >> Yeah. Yeah. And and I have I have looked at the sort of native capabilities of the social platforms. It's again all focused on the like what the talent has been talking about within that environment. >> Not who. Not who. It's what >> not who they're reaching. It's not the who, it's the it's it's about them themselves. >> Okay. So Okay. So that process happens. It's messy. I'm finding people. I'm out. I'm sending blind emails or maybe I'm reaching out to their talent agencies.

14:09 Do do people use managers and talent agencies for a lot of this? >> Yeah. Often times you end up with two agencies as counterparties in the negotiation. You've got the brands agency and and the talents talent agency or talent manager negotiating a deal that's you know a talent fee uh where they're going to you know if we're talking about again traditional influencer marketing they're going to create content uh you know generally in their own voice the brand will have some influence on on approval. Uh that's a fairly lengthy negotiation, but you you've come to terms and the creator uh or influencer generates the content and then posts it across their across their socials.

14:53 >> So So I think you skipped a part that I'm interested in and is relevant to your business, I think, which is the conversation about the audience and the data and the and which techniques to use like and so the buyer does the buyer have any data or they just have like follower count and some very surface level data? Exactly. They don't have any if the buyer doesn't have any data today uh you know unless they are licensing a a sort of uh a platform you know an influencer marketing platform themselves where they're doing the discovery in that platform but again that platform is just going to be looking at social metrics. So you're going to be seeing uh you know age and gender of the audience uh geography. Okay. Uh and then you you're you get very rich engagement metrics. So you know when somebody posts a video on meta, how many views, likes, shares, comments, whatever. Uh but you have no idea who is actually seeing that video.

15:49 >> And right and you so you're you're it's almost like contextual. We know we know this person talks about X and X is relevant to us. So, beauty, beauty, and they don't curse a lot, so they're safe. Let's just give them some money. Right. And and what you're saying is absent your platform, this isn't supposed to be an advertisement for your platform, but absent your platform, also the seller doesn't know very much. >> Uh, that's correct. Yeah. The seller, the talent themselves have no idea. I mean, the talents, they've seen their fan base in a crowded room. They've met them on the street, >> but they've [laughter] never been given a consolidated data sheet of, >> are your fans educated? Are they parents? What kind of cars do they drive? Like none of that information's ever been given to them.

16:29 >> Yeah. Okay. Got it. All right. So then they do a deal and it has it has do these deals have performance aspects or are the deals basically like you know just a line items of here's what the talent's going to do. >> Uh it's mostly line items of what the talent's going to do. So it's yeah we're going to create x number of pieces of creative and we're going to post x number of times and that's that's that's kind of it. And then what how do you know it worked?

16:57 >> Uh well, measurement's been a major problem in in [laughter] traditional influencer influencer marketing. Yeah. I mean, yeah, the the uh there's there's definitely a lot of companies out there now working to to solve that. Uh you know, the the marketplace has grown, you know, like like wildfire for, you know, a number of years. Uh and quite frankly the technology and the data like the infrastructure is you know now starting to catch up.

17:27 >> Yeah. Got it. So one of the theories I've had and I mentioned this on the pod and I think I did in my newsletter as well is that I do think the platforms could uh come in here and and take part of this value chain. measurement in particular, like if you know that a given, you know, piece of content is part of a campaign, it's really not hard to make that a touch point in your measurement models. Um, yeah. Are you seeing anything there? Do you agree or disagree with my prognostication?

17:54 >> I I completely agree that that social platforms are an amazing position to provide a lot of value here. Whether they're going to or not is is sort of the other question, right? Uh you measurement. Yeah, for sure. They could they could help with measurement. they could, you know, the the big measurement challenge that the brand has is that the organic post is occurring in an account that they don't own, right? Like that's not hard to to solve. Like let's just stream that specific data into the brand's account and now it's a unified measurement, you know, within the walls of of that brand's meta account. Great.

18:27 Yeah, you don't even really have to charge for it. I mean, you probably will, but you could just do that. I mean, if if you just offered that um it would make everyone's lives better, right? Um but that that's usually not a good appeal for the big companies. Um so um so uh to close it out I do want to just do a little trip down memory lane. Uh so what app Nexus is you know once in a lifetime experience with the people who work there. Um I know I had to do 10 years of therapy afterwards. So um what do seriously though you know what what do you learn from that experience that's like relevant to what you're doing now?

19:01 Um is it was it helpful uh from an entrepreneur perspective? um give us the uh give us the wisdom. >> Yeah, I mean incredibly helpful. If you look at the network of folks that joined App Nexus and later went on to start their own companies, you being one of them, that network is impressive. It sort of spawned entrepreneurs. Something about the culture, you know, really encouraged or like attracted entrepreneurs and then also encouraged them to like go out and do it themselves. Uh so I I wanted to do this a very long time ago. For me personally, like the rocket ship was too appealing to to get off. And so I I rode it for, you know, all the way through AT&T acquisition and then into into Microsoft and, you know, finally decided that it was it was it was time to to do it do it for myself.

19:52 My business is in programmatic advertising. I'm using so much of my experience and understanding of the open web and you know data signal and digital media to you know bring this new vision to life. What's what is very new and foreign to me is Hollywood and talent and so that's been that's been a fun learning curve. Uh it's going a lot of time out in out in LA. Uh but it's you know it's it it's really like creating this this blend between Hollywood or this bridge between Hollywood and New York that that we're sort of playing right now.

20:30 >> Do you go to Hollywood and start making you know boxes and arrows on the whiteboard or is that not play? >> Not yet. No [laughter] >> I think they do visit a little bit differently. >> Yes they do. They do. Yeah. Uh we are spending a lot of time educating on just the fact that you know your talent data is being monetized by all these platforms that you use to communicate with your fans. You know you you can legitimately step up and demand access to that data. I think you I don't know if you saw the the DOJ uh settlement recently with with Ticket Master Live Nation where you know the the ticketing data is going to start coming back to musicians. That's a that's a watershed moment for musicians. They are going to start to have all the contact information for all the people that you know bought bought tickets to to their shows.

21:21 >> That's a really valuable asset for them going forward. >> Yeah, absolutely. I didn't see that. It's interesting. >> Yeah. >> Well, Alex, congratulations on the success. Thanks for joining us. Uh I think probably people are pretty interested in the subject. You could be reached at Endorseable is the company name. Um, and uh, it was a very interesting perspective on this kind of emerging area. >> Thanks for having me. [music] >> Hello, Ari. How are you?

21:46 >> I'm great. How you doing, Eric? >> I'm good. I'm excited to be here for the news segment. I missed a good interview. >> Yeah, you know, sometimes without you, the interviews are lonely and sad, but you know, >> are productive and [laughter] and fast moving. Anyway, we got news this week. So, uh, let's start with this Walmart news. So, straight off of their website, Walmart Connect expanding access to Walmart first party data across leading platforms. So, uh, Walmart is opening access to Vizio inventory start through Yahoo DSP using Magnite's supply side tech um with additional partnerships uh, planned over time. There's one company that is noticeably absent from this announcement and that is TTD. Not on the launch list after a longtime partnership.

22:37 >> What you make of that? >> Well, I I don't know the specific on this TTD question. Uh is TTD not able to buy Vizio inventory? I doubt it. I I would assume that they had that as a baseline, but I don't really know. Um the the new news here is that someone else besides TDD has access to this data and inventory. Um, it's Vizio inventory with Vizio data only on Yahoo DSP and Magnite is sort of the intermediary because Magnite I think had the relationship with Vizio already as their SSP. Um, so this is not huge news but it gives you a glimpse into Walmart strategy. I think when they acquired Vizio everyone was very concerned it was going to be super locked down. It was be walled garden and uh, etc. etc. and definitely doesn't appear that way. It definitely appears like they're uh Walmart Connect is going to try to maximize revenue and try to expand and uh meet buyers where they are. Um and so I think it's pretty exciting for Walmart, very exciting for Yahoo DSP to be sort of anointed the number two most important player for Walmart. Uh really, right? [laughter] >> Really is.

23:47 >> Yeah. >> Incredible. Yahoo. >> A 30-year-old brand, huh? >> Well, Yahoo DSP is not exactly 30 years old. It's on a roll. >> Yeah. I'm just trying to be dramatic. >> Yeah. Okay. [laughter] I mean, you know what? Who's who's Walmart's enemy number one? >> It's Amazon. Yeah. Right. So, Amazon DSP is not going to be on the list anytime soon. [laughter] >> Of course. Of course. So, okay. So, but TTD loses one of its retail media talking points arguably.

24:18 >> Yep. I mean they uh it's not ex well we don't have Walmart shopper data generically available on the open web. We're still talking about visio inventory and >> yeah there's you know but you it's certainly a crumbling part of their garden >> on their wall >> and and yes the strategy today seems to make a lot of sense of meeting buyers where they are. Um, what would you peg the odds of Walmart somehow maybe building or buying their way into becoming a walled garden and being the, you know, the fourth necessary DSP for a buyer to access like really really really important unique data andor inventory?

25:01 >> What do the odds? >> I think they're strong. Uh, the odds are strong they do something. I I wouldn't necessarily predict a DSP acquisition though, though it is possible. Um the um but they have the Vizio acquisition was a pretty big stepping stone towards a more aggressive stance towards the the whole the whole kit and kaboodleoodle. Um and DSP could be a piece of it. Um although it might be a situation where building their own DSP on the trade desk which is how they talk about it um is good enough and and they don't necessarily need to own that. They just want to meet buyers where they are on other platforms. Um so um I I think there'll be more acquisitions in the future, but I think that um I don't know that a full DSP buyout would be the next thing.

25:54 >> Got it. That makes sense. I mean, Walmart can I mean they could buy anything, right? They could they could buy half the Lumiscape, threequarters of the Lumiscape. All the [snorts] Lumiscape. So, uh it's a it's a fun one uh to uh to speculate on for sure. And it's all speculation, folks. >> It is all speculation. >> All right, let's talk about another one. Pubmatic. Pubmatic launches AI containerization. So, another announcement somewhat like index cloud.

26:25 >> Yeah, somewhat. And also again there the AI inside this containerization is being shoved in with like a hammer. Uh the containerization is the story and the AI is like hey you know put some AI in your container right it's like >> right so and maybe that that's maybe that's the difference between the index cloud announcement and this that's a very good point. So it's automatic they're letting thirdparty algorithms run inside their auction right not upstream of it. So buyers get decisions made against unfiltered inventory in real time. That's the the similarity.

26:56 Maybe the the emphasis on custom algos or algo companies is the difference here. >> Yeah. Well, it's the same actually. I think the I mean starts with the IB who has ATRF where the A stands for AI for no reason. And then index announces their containers and it's all like you know they're a little more honest about the press release and they're like it's about speed. Great. And now this one is AI but really when they say AI what they mean is speed. It's all about speed. You can't do AI unless you have speed because AI takes more time to calculate things. Um, and so I really I just discount the AI part of all of this is containerization is the story. Um, and that's pretty interesting um, on its own. The AI stuff is maybe a little bit of the uh, tailwind that's making people interested in containerization.

27:46 >> Who would have thought containerization would be the word of the year for 2026 in the beginning of the year? Besides Adam and Pilots, I >> I think I I it's not a it's a good idea. It's been around for a while. It's been tried and something is in the air that's making it better now than it was a couple years back. >> Shift to the cell side. >> Yeah, it's a shift to the cell side.

28:09 It's a little bit like technology innovations and things that were maybe not as standard in the tech crowd a couple years back. >> Yeah. >> Or maybe it's AI. I don't know. [laughter] >> Yeah. Do you think that, you know, we're going to the second half of the year. Do you think this reuff reshuffles the deck at all in terms of winners, losers, strong companies, weak companies, you know, when we do our end of year recap?

28:35 >> Um, I think it's I think it continues the trend towards uh um um I forgot the word curation. Curation's the word side. Yeah, this is sort of a continuation of curation which is uh more data to the cell side. Um you have companies like swim bringing performance to the cell side and attain and chalice and they're all moving to the cell side and containerization helps all of that. Um and makes it makes it faster and easier and cheaper to deploy those business models. Um so I think that's all that's all just continuing a trend that previously existed. the the gauntlet I threw down after the index announcement or the question I raised publicly was will Bedrock be able to prove that you can be a DSP in a container. Uh that's that's the that's the thing that they're betting on which has not ever been done at scale. Uh and that's unproven and they haven't proven it either.

29:36 >> Yeah. So watch this space. >> Um all right. So, by the way, how good was the hightouch uh interview last week? Did you get uh like anybody give any feedback? I thought it was great. >> Yeah. Yeah, I thought it was really good, too. He was great. People, it's a company people should know more about and don't. I didn't know very much about them. Um so, yeah, I thought it was pretty entertaining. >> Yeah. Yeah. So, they came out with a uh relevant announcement this week. So, they launched exposure log matching for TTD. So basically let brands pipe TTDS raw impression level exposure logs directly into their warehouse and resolve them against their own firstparty customer IDs with no third party ID dependency. This I think is the real maybe one of the first real answers to you know are you looking to step into the live rev shoes that you asked um the the CEO last week. Yeah, this is uh Jeff Green's dream of UID2, right? So UID2 allows anyone to make a request to get an ID and and then save it themselves without paying a fee. And so if you have a match table already of UID2 sitting around, then you could take trade desk logs and use them natively. And um this is exactly the kind of scenario that Jeff was hoping would happen. Um and it should happen more. I I think UID2 has been stunted by competitive uh worries and folks have not adopted it because either they were in the business of having their own ID like a live ramp or they were in the business of buying ads in some form so they saw trade desk as competitive and didn't want to adopt it.

31:10 Um so this is great and uh you know you would expect this to have h already happened more but it hasn't. >> That's so interesting. So cuz there was, you know, debate around how Publix is acquiring live ramp could possibly benefit TTD when I forget which which conversation I had, you know, uh the the line of thinking was well, it could benefit TTD's data marketplace, but could actually benefit TTD TTD in terms of UID adoption in cases like this.

31:40 >> Yeah, it could. Right. Um, so whether an ecosystem of UID2 users is going to create, I'm still pretty skeptical because you have no other DSPs adopting it, right? But but if TDD it remains like the number one source of open web demand, then you know you would you would think more peripheral players might adopt it. >> Yeah. Okay. Well, this is one to watch. Um, this hightouch company all of a sudden get a market architecture.

32:08 >> All of a sudden [laughter] they got a market architecture. Hey, we should claim that they raised like their billion dollar round after market to see if people can fact check us on that. [laughter] >> Something tells me they might have another raise or something in the future. Oh god, that momentum. Um, you know, I have I I'm constantly looking for signals and signs of the infrastructure layer changing, right? Like what what happens postprogrammatic? It was an announcement this morning. you know, it's it it's it's it's smallish, but I think it's kind of interesting.

32:43 So, I don't know if you saw this. Um, InMobi scope 3 announcement. >> Yeah, I didn't see this. Tell me about this. >> Okay, so InMobi Inmobi has uh I think via uh acquisition they have something called Glance. >> Glance is a unique ad format. It's like uh I don't think it's in in the US. It's um and it's probably not uh iOS users e either. Um, it's when users just like have their lock screen. Um, Glance serves ads on it and maybe there's some sort of rewards for for the user. I don't know exactly how it works, but it's like the lock screen thing. Kind kind of cool. Um, it's like 300 million users. So, something relatively scalable. In moi is a global company and they're using cellside agents to sell it um in partnership with scope 3.

33:26 >> Wow. Okay. Um, yeah, this is like a vin diagram I wasn't expecting. like Android users in India and agentic AI through scope 3 is a combo. >> Yeah, you got to start somewhere, right? Um and this idea of you of just some some percentage of the space moving to aentic selling, agentic buying, however you're going to call it. Um we need to start seeing some of these signals. And this is this is one of the first that I've seen. the Immoi CBO before you comment um had a wild statement um in the article he is calling for 80% of digital spend to be routed through agents within two years which I guess is possible but I mean that is like sea change >> yeah the the areas that are the hottest around agentic or adp are the areas that aren't very good for programmatic things like uh I think we we may have talked about this on the show I'm losing track of my conversations but like audio Um, people are really excited about using AI to to bid on audio. Um, because audio is not very good at programmatic.

34:30 I know we'll get a lot of hate mail on that one. Uh, but um, the other another TV, right, with >> TV RPA. >> Absolutely. Yep. Another area. Um, another area that this obviously says is very low CPM inventory where uh, you know, India is known as a market that's hard for adtech companies because the the going rate CPMs are so low. They're in the 10 cent range CPM, right? And, you know, running a whole data center listening to a million QPS isn't worth it. Um, so it's another example where they may be there may be like a little innovator's dilemma here where ATCP is not as good as programmatic, but it is able to pick off little areas where it is as good or better.

35:15 >> But that's not the vision for ATCP. The vision for ATCP is for ATCP to be like the the layer. >> Well, let's keep in mind is is not owned by any individual. So there is not vision for it necessarily. It's a it's a collaborative effort of many parties to operate their own businesses in the most efficient way possible. >> Um anyway, I think this is neat stuff. >> Yeah, absolutely. It's very neat. Uh InMob is such a neat company. There's always there's always talk about them going public and I think they did they file an S1 at some point? I don't remember. Uh I'd love to see >> they've yeah long rumored to be like a billion dollar company, something like that. I have no knowledge of this.

35:54 >> Yeah, they have a lot of other businesses too. like they do have consumer businesses in India and uh and AP pack in general. Um so it will be very interesting to see if they go out. >> Agreed. But there is an IPO coming. Lift off. >> We need a bell. Do we have a bell? We need a sound effects. Not enough sound effects on architecture podcast. >> We need sound effects. That's right. Uh producer Mike throw that in there. Um Liftoff is back on. liftoff. Uh if you recall, they uh filed an S1 stopped.

36:26 Now, now now they're coming back. Prices cut um from 26 20 to 30 down to 20 to 22, right? So the the the the price of the IPO was cut. Um but we got to look at some of the numbers and the numbers are pretty interesting. Um it's a $5 billion uh you've got uh enterprise value business. Um and the multiples if this goes out in the 2022 range are you know basically 7x revenue 12x IBIDA so so fairly rational for context apploven is trading at 40xida um so this is not apploven um but it's a interesting business and at 5 billion it's actually larger than a lot of the publicly traded tech pure plays >> do you remember what the gross revenue was like a bill >> don't probably >> uh yeah these mobile companies have just um h I mean the honest truth that no one wants to hear among the programmatic people is that they have a more sustainable business than uh than the programmatic world does. They're not dependent on search traffic nearly as much. Um people are willing to pay for the games uh and customer acquisition works at scale. Um so um so if they could call this a mini apploving and uh and have and can execute um and thereby grow well that that sounds like a pretty exciting stock to own. Um I saw some slides that they were showing um in this uh road show and it was interesting.

37:53 They they said that over half of the revenue comes from outside of gaming. Um so it's about half let's just call it a half gaming company. I don't know if Apploving gives out that stat. Um but >> I don't think they do. >> I don't think they talk about the percentage of e-commerce. >> Yeah. >> The growth percentage of e-commerce and it's [laughter] relative to their core business is tidy. Yeah. >> Yeah. Exactly. Um so this is I'd say half outside gaming is a pretty good number. Um and then uh they say that they're the number three SDK other than Google and Facebook. So that means Facebook and Google number one and two.

38:25 Then there's two more presumably app 11 and Unity. And then they're the third X app Google and Facebook. Um SDK penetration still matters. Um and it's a very sticky and um defensible even though customers do multihome. So you're not it doesn't mean that they have some portion of the publisher universe uses them exclusively or preferred. It just means that they have presence in them. Um so it's a it's a factor in the business that makes it more investable I think and more sustainable than than you know tag on the page in the web world.

39:02 >> Yeah. I did a quick search by the way um revenue. So trailing 12 months ending September 30th was 630 633 uh million revenue based on their growth rate. It's probably going to be 750 to 800. So s slightly less than uh than a billion. But but >> that's gross. That's gross. That's Yeah. Okay. >> Yeah, that's correct. Um you're right, by the way, in terms of mobile. Um I made I made this post on on on LinkedIn that um you know, I've been I've been digging into mobile over the course of the past couple of months. Um and um we we've made um our first in like I mean eight years of doing this our first pure play kind of mobile app infrastructure investment at Emperium. Um yeah we we we did it last month and we're making another one. Um, so I agree with you um, in the characteristics of the market and it's also one of these markets that is uh, I don't want to say AI resistant, but it's not subject to, you know, some of the issues that other categories like the open web have with respect to AI.

40:08 People are actually playing more games. They're not playing less games. Um, >> it's not the consumer AI play isn't it isn't in jeopardy here. Consumers aren't going to switch off games to do AI. Although they may have new apps that are like the companion apps effect effectively are AI uh but they act like games basically. >> Yeah. And by the way these ads that apploven and others run they are like completely portable to some of those environments. So it's like kind of uniquely AI resistant.

40:37 >> Well let's talk about the AI tailwinds though. So first of all you have creative right? So a ads in app are much more AI friendly than ads in other places. You create hundreds of variations. You make them playable. You have AI avatars, things like that. Um that's already happening. And if you think about um not market share shifting but efficiency shifting where if if you have let's say the markets controlled by six or seven mobile ad networks all of whom have basically the same technology and the same data um the AI makes them all better and they all get more profitable. Um and they don't even have to shift share and they can be growing.

41:18 Uh right. Um there there is a limit to the amount of consumers are spending ultimately but um but the um the AI tailwind on the a on their ability to optimize is very clearly there. >> Yeah. Yeah. Well said. Um final kind of kind of anecdote on this. I made that post on LinkedIn that we're you know we're digging I'm digging into AI. Uh Jeremy Bloom CEO of Marchitecture I know listens to this. Um, he he said something to the effect of, "I just came back from MAU and I was shook." He said that in a positive way, like, "Oh my god, this market is nuts." MAU, I guess, is the the mobile version of Possible or CES or something like that.

41:58 >> Yeah. It's a Vegas conference that's very focused on uh on mobile app. >> Yes. Okay. So, continue to watch this space. We got some M&A um this week as well. Actually, today as well. Today is Wednesday, June 3rd. So first was Pier 39 acquired Adlooks from Scope 3. So Adlooks uh you know contextual data company Scope 3 acquired a couple years ago. Um, Pier 39, um, who, you know, I I guess is doing quite well, acquired it.

42:27 And this I, uh, as I understand, specifically gives Pier 39, uh, ability to start to get into wall gardens. Um, which I think was, you know, the one area that, um, that they they had a a gap in, if I understand correctly. >> Yeah, private to private spin out. It's pretty probably a small deal. Uh, what what was the rationale for scope 3 getting into this and buying ad loops? I I think there was I think Brian talked about it when they did their big pivot, but like I don't really remember exactly what the thought was.

42:58 >> Yeah. No, my my my mind I I I thought you were gonna ask me this and my mind is uh is is kind of lagging on this one. Um it was it was it was either right right at the pivot or before the pivot, you know, when it was like uh very focused on measurement. I think it was, you know, some unique measurement capabilities, but don't quote me on that one. Um Brian Brian can uh can add us on this.

43:21 >> I I think when when Scope 3 did its pivot out of uh out of green measurement and towards Agentic, you know, the first demo they had that was really good was uh like a Mobian/Doubleverify competitor. It was like analyzing web pages and doing context and then putting that in the engine. Um, so that probably was from the AdLook's acquisition and maybe I don't know if they still do that or if they're 100% focused on transactions. Um, but probably as a standalone business it wasn't important to them anymore.

43:53 >> Yeah, that makes sense. We should get Brian on the pod to uh give us the update on scope by the way. >> Sure. >> Um, all right. Uh, this one both of us heard about and we were like, wait, >> what is this? [laughter] So, so, so a banker sent this to to Ari and I, our personal emails or something like that. Um, you know, we we just like really like, huh? >> It wasn't just a banker. It was Phil Fresen, friend of the pod, the pres early subscriber to market, constant attendee at our conferences. We salute you, Phil.

44:28 >> Is he a banker? >> Yeah, he's a banker. >> Okay, so I said a banker. I [laughter] I did. Stop hazing me. Psychor acquired a geo startup called Scrunch. I don't know about Psychore. I don't know about Scrunch, but the headline number was healthy. The headline number was 225 million. >> Scrunch, which >> it sounds made up. [laughter] [clears throat] >> I like scrunch. It's a decent word. >> Although I defer to you as the the king of >> No, no, it's pretty good. It's pretty good, Dave. Scrunch. It just sounds funny.

45:00 >> Um, so GEO is a hot space. There's companies we've never heard of getting bought for $225 million. That sounds good. >> Exactly. That's the signal. That's Yes. I mean, the investment in in GEO uh like both of us are are happy about this one. So, it's you know um the this space is wild, right? Like you know the companies are raising, you know, $20 million getting valued at a billion dollars um or selling for healthy nine figures. So, it's it's certainly a a a strategically valuable space. um that is going to probably have you know more noise than than less over the course of the year.

45:37 >> Yep. I keep coming back to Nate Elliot on this show saying um saying um number one request from CMOS is how is my brand looking in the AI? >> Yep. Agreed. Um other stuff uh I I want to ask you this thing. So, this this quote you put in the in the notes from Senator Ron Weiden was scary to me. Um, I'll quote it and then you give us the the the backstory. We should start treating the adtech industry as a national security threat.

46:09 >> Indeed. What's going on? >> It's all the blue blazers. They think we're coming after them. Uh, it's like a January 6 situation except it's like dudes with blue blazers and gingham shirts. Uh, no. Uh, [laughter] >> talk about a a AI image. That's way >> Yeah, someone AI. >> Um, no. Uh, what what Senator Widen was talking about is a real concern. Uh, it's location data. It's always location data. So uh basically they came out that you can use commonly brokered location data to find uh US troops where they are overseas because location data doesn't know what it should not show and you don't need personally identifiable information. You just need the location you can infer all kinds of things that are happening um around the world. Um I am on the record saying I think location B data sales should be banned like fully banned like COPA. Um which is not a very popular point of view in this industry but I really think it um and then interestingly I um I think um Alan Chappelle um pointed out on Twitter that um Connecticut just passed a ban. Uh so I I may be getting this I should have this in the notes but I don't. I think the net is that 25% of the US population now lives in states that ban location data sales. Um, and yeah, um, and it's increasing. And it's good that it's increasing. There's no way to make it safe. There is no way. You can't you can't make it less precise because a place like New York City being 100t off is less precise, but if you're in Wyoming, 100 feet doesn't mean anything.

47:44 And you can't bundle it. can't put 10 users and call it a cohort because one of those 10 users hung out at your house all the time and you can figure it out. Um, and and so on and so forth. And the more gymnastics you go into to make it safe, it just doesn't work. It it's unsafe by its nature and it should not be it should be protected like like personal information like social security, credit cards.

48:07 >> Yeah. Um, impossible or very difficult to argue with that one. Although I'm sure somebody's going to come and try to start the argument with you. Well, there a lot of companies are dependent on it. So whether you like it or not, >> I don't love that it was the adtech industry. Um, >> right. >> This is adtech is blamed for things that are like there are there have been situations where search data has gotten leaked and then they say, "Oh, it's the it's the surveillance capitalism. It's the adtech industry or location data or things like that or people listening to the conversations on your phone." Right?

48:41 [laughter] Uh so um you know it's a there's also a pretty big branding problem that the adtech industry has. >> I you know that's so interesting. Um I was listening to Mark Andre on um on Rogan. Um I think they uh it was the the interview was posted last week or this week. Um it's good. It's like wide ranging. You know Mark is is is you know super intelligent and know you might not be your your favorite, but it was it was really good. One of the things he he made the point of was that the AI industry just you know does itself no favors and you know with some of the companies um you know just like anti- PR anti anti you know kind of kind of good branding um it's wild that you know an industry like adtech which you know um is is all about uh there's a subset of marketing and communications um has an issue like this and like maybe we should do something about it.

49:36 >> Yeah. Um, I mean, I think the AI industry putting out ads like billboard ads like stop hiring humans. I mean, what are you thinking? What are you doing, man? >> Yeah, that's what he [laughter] that's that's what Andre was pointing to. You know, some some of the, you know, uh, the narratives put out by some of the CEOs for certain. >> He's the guy who invested in Cluey whose motto was cheat on everything. So, you know, he should maybe look in the mirror a little bit on that one.

50:00 >> I think Andre Harowitz has invested in like over a thousand companies. I don't know if he necessarily like directly, you know, made that investment. I'm no apologist for Mark Andrees, but you know, [laughter] just how these firms work. It's a big firm, man, with a lot of money and a lot of a lot of partners doing a bunch of things. >> To to address your point, I don't I don't know that the nuance of like privacy is is important, but we use your data in a safe way has never worked.

50:24 It's really hard to get across. Um, and I don't know if we're ever going to be able to win that argument. No, that's not the argument to win. But I think, you know, the the the argument around all of these services that um remain free, you know, due to advertising, um all of the people that get access to information and services and tools because of advertising, like that higher level stuff, it's still like like nobody does it well. Um and it's like super frustrating, you know, when you you hear this and it's not, you know, kind of kind of uh understood. Anyway, I don't know. We should do [laughter] We should do something. No, nobody's Nobody's doing anything. Let's get a Let's get a committee together.

51:06 >> Yeah, exactly. A committee to change public opinion about advertising. >> I'm on it. >> Let's get on it. [laughter] >> All right. You want to call it there? >> Yeah, let's call it there. Uh, thanks for listening. This is a great uh great conversation with Alex and also also, you know, some pretty interesting news this week. Um, we'll be back next week. >> See you next week.