Latest / Signal & Noise: Bringing Clarity to Media, Data, Tech and AI / From Publishers to Platforms: Evgeny Popov on CTV, Career Exits, and the Future of Television
Transcript
- 0:03This is signal and noise, bringing clarity to data, tech
- 0:07and AI. Hey everybody, welcome back to
- 0:15Signal and Noise. I'm Brett House, joined by my Co
- 0:17host Rhea Longacre. And today we've got Yevgeny
- 0:21Popov, who has, if you don't know him, he's at Samba TV.
- 0:25He's built one of the most interesting careers in modern
- 0:27media. His path across publishers,
- 0:30agencies, ad tech platforms, data companies is pretty unique,
- 0:34right? And you've had leadership roles
- 0:37of getting it. We'll call you Yev just to
- 0:39simplify the syllabic aspect of your name, multi syllabic aspect
- 0:44of your name. You've been at New York, News
- 0:46Corp Australia, so you were on the media side, Lo to me, which
- 0:50is certainly interesting. Was around for a long time with
- 0:52Andy Manfred hard. Quiet that long ago?
- 0:55Yep. Yeah, Verve, one of the last
- 0:58data management platforms to be acquired, although I know you
- 1:00guys had a lot of pivots and now Saba TV and you've you've, you
- 1:04know, you've participated in a bunch of exits, a bunch of
- 1:07company exits, which some of us have had the.
- 1:10Three times right if not mistaken.
- 1:12Can I just clarify that statistic because I think people
- 1:15will be for it. When do you say exit?
- 1:17You know, people say, well, you would have found us.
- 1:19You actually exited the business as a founder.
- 1:21So have one exit, which I have company which I founded and
- 1:25exited, which is Corefield loyalty app, which was back in
- 1:28Australia. I can tell you more about that.
- 1:30So it's it's a it's a it's a real and then two companies
- 1:34which have been acquired, you know, and me being part of that.
- 1:37So just to like KB out that. Yeah, yeah, yeah, cuz, cuz and I
- 1:42think that all to, you know, when you're the founder, you
- 1:45just make more money out of the whole deal, right.
- 1:47That's what I think it's also. Bragging rights too.
- 1:49Yeah, it's also bragging. Yeah.
- 1:51And I made, I probably make a fool of myself saying it now,
- 1:53but out of two of them I made $0.00.
- 1:56Yeah. I've been through 2 exits in the
- 1:59DMP world with Excel, 8 in new Star in the identity data tech
- 2:03world with TransUnion. And to your point, yeah, it came
- 2:06out OK, but definitely not the founder model to your point.
- 2:11But you know, Rio says you're one of the best connected most
- 2:14network guys in the. End There's a lot of people.
- 2:18Which which I'm interested to hear more so why don't you tell
- 2:20you it's a little bit about yourself for those that don't
- 2:23know you and. Yeah, no thanks, Brad.
- 2:26What a what a wonderful introduction.
- 2:28And Brad and Rio, thank you for having me here.
- 2:31Such an honor. You know, to answer your
- 2:34question about connections, I think I'm a collector.
- 2:37I like to collect good humans. And throughout my career, that's
- 2:41exactly what I've been doing. You know, I've been expat for
- 2:44the past 26 years. I left Russia in 99 and moved to
- 2:50Australia and spent 15 years in Sydney, moved to Singapore,
- 2:54spent five years there and then came to New York in 2018 or 19.
- 2:58And I though it's not one year in Latin America, just just for
- 3:02the fun of it. So like, you know, I've been
- 3:03around, but one thing is, you know, you move countries, you've
- 3:06been with continents, you move companies, but one thing you
- 3:09have is you collect good connections and network.
- 3:13So I think I'll, I'll take Greece compliment.
- 3:15I don't think how true it is, but throughout this, you know,
- 3:1825 years of moving around, you definitely meet a lot of people
- 3:23surprisingly in our industry, Brad, you probably can attest to
- 3:26that, a lot of the same people. So if you talk about measurement
- 3:32incrementality of my network and how fast is growing, it's, you
- 3:35know, I was always surprised and I was still in Sydney back in
- 3:38the base coming to a programmatic IO like you know,
- 3:41when it's sole starter 2013 and and then it's 1° of separation,
- 3:46you'd be the same people in early stages and they're still
- 3:49meeting the same people. So it's interesting to see, in
- 3:52fact, I don't know names, but you see people who account
- 3:55managers when I met them and now the CRO of, of public companies,
- 4:00you know, and it's like, wow, like, you know, you see the
- 4:03evolution and how some people are supercharged in their
- 4:07careers. And I've been so fortunate to,
- 4:08to be part of, of that. So yeah, that's, that's
- 4:12definitely a journey. But I can I can tell you a bit
- 4:16more about I guess how it all started and.
- 4:19Yeah, if you if you want to tell us your.
- 4:21Story. Yeah, yeah, in fact, maybe
- 4:23before you jump in too, I mean, I think that like like that's
- 4:26the reason I I'm sure listeners want to hear about that because
- 4:29I think it's a cool story, right I mean, looking at your resume
- 4:31and like. And this was what I thought
- 4:33would be like really fun discussing is like, it crosses
- 4:36not only all these different layers of the industry, right,
- 4:38publishers, media agencies, programmatic and not TV data,
- 4:42right? So I think that looking at your
- 4:43journey, understanding and like, I love to hear insights,
- 4:46especially your current role. But maybe we, we, we tabled that
- 4:48for a little later, right? About like how you're taking
- 4:50this vast experience, how it's applying to, to CTV, to
- 4:53measurement. It's a rare perspective.
- 4:54So I think it's going to be a really cool discussion with all
- 4:57that's going on, especially, especially in CTV.
- 5:00So maybe kick us off there. Tell us about how you got
- 5:01started. Sounds like you're kind of like
- 5:03many of us, an accidental media executive, so love to start
- 5:06there. I like that.
- 5:09I like accidental part because it's exactly was that.
- 5:11So to Brett's question earlier about the exits and how it all
- 5:14started. I started as an engineer.
- 5:17You know, I've started my career working for Microsoft joint
- 5:22venture with one of the publishers in Australia in
- 5:24Sydney. But before, before that, I
- 5:27actually came to Sydney to work on one of the first meta travel
- 5:32search enginescalledhotelsclub.com.
- 5:36That was the first exit which got acquired by Expedia, believe
- 5:40it or not. Yeah, Yeah.
- 5:41So by orbit. But then it's now it's in
- 5:43Expedia's hands. And so it was an interesting
- 5:47way. I was head of kind of
- 5:49technology, had a big team. I was like 20 or something, had
- 5:5125 people coding and doing all this cool stuff, and then it's
- 5:56got sold and I didn't know anything about stock options,
- 6:00what the transaction should look like.
- 6:02But I kind of come in with engineering and data pedigree
- 6:06back in the day. So that's always was my DNA.
- 6:10And even now, you know, yeah. It certainly works in the ad
- 6:14tech martech space because of the just the technical
- 6:17knowledge. It just helps a lot because we
- 6:20do end up going down those paths quite frequently when we talk.
- 6:23Yeah. And you know, Rio and I met at
- 6:27just spoke really briefly at Architecture Live and I was
- 6:30telling Rio I was like, Hey, I'm I'm shipping code again, you
- 6:33know, and my team was looking at me like what you can code, you
- 6:37know, it just feels good, you know, like it's kind of 360
- 6:40going back to your roads, going to, you know, obviously at the
- 6:43much faster velocity with with with ability to do.
- 6:47Oh my God, you know. I can go into the woods of that,
- 6:51but there's just like phenomenal pace.
- 6:53It's, it's so good that my wife banned me to work on Sundays.
- 6:59So I'm, I'm not allowed to touch keyboard on Sundays.
- 7:03That's how I deal because that's so addictive.
- 7:05It's, you know, it's really is. Yeah, it's real.
- 7:07I'm getting Openclaw, a bot implemented on my Mac this
- 7:12weekend by a common friend of ours from Snowflake, so I'm
- 7:17going to be diving deep. I use codecs and a little clod
- 7:20as well, but it's it is incredible.
- 7:22Well, just generally speaking, I, I find that like so many
- 7:25things, like in my, that in my job that, you know, I was doing
- 7:281015 years ago, I'm able to do again and I'm able to do it
- 7:31faster, I'm able to do it better.
- 7:33It's, it's a really cool feeling.
- 7:34So yeah, I know where you came from.
- 7:37Real. Like it's so cool that I was at
- 7:39the gym this morning. This is true story.
- 7:41I've purchased the domain on the fly, I created a shell of the
- 7:45project I'm working on which I'll be deploying and I've built
- 7:48a scaffold in as I was doing it between repetitions like this is
- 7:52not the job this. Is that's bad?
- 7:53Yeah. And this is stuff that would
- 7:54take you potentially. Weeks like this is was this just
- 7:58this morning? I have it and I'll better deploy
- 8:00it after this call and I'll send you the link.
- 8:02So like that's all done end to end.
- 8:05You know architecture, deployment, hosting, domain
- 8:08names already. Isn't that incredible?
- 8:10Yeah, and like you and we've we're learning these things sort
- 8:13of on the fly and it's happening.
- 8:15So like and just as an example of my life, I just so the guy
- 8:18that's that's implementing this bot on my computer and getting
- 8:21me open claw educated, took my website, ran a full audit of a
- 8:27site that I spent. I hired a team, you know,
- 8:30outsource that built all this content, spent weeks doing this
- 8:33like back in December, you know, weeks and weeks and weeks
- 8:36getting this site up, you know, and I'm like, I'm still not
- 8:38particularly happy with it. High signals.com.
- 8:41And he over lunch, he's like, I'm making lunch right now and
- 8:44I'm going to rebuild your website.
- 8:46And you know, an hour later he sends me the rebuilt website,
- 8:51completely hosted almost. Every page better.
- 8:53You could have all the assets and it just blew me away.
- 8:55I was like, Oh my God. And it wasn't just an imitation,
- 8:58it was a reconfiguration. There was some better packaging,
- 9:02some better, like sort of how we engage, how we work, type of
- 9:07workflows that I thought about, but I just hadn't implemented.
- 9:11So it's pretty amazing how quickly that stuff can be right
- 9:14You're. Right time to market is, you
- 9:16know, it's a gold like ability and you know, kind of taking it
- 9:20back to this engineering days when you understand how the code
- 9:23works, you can understand what's under the hood, you understand
- 9:26technology and the same product. You're a builder.
- 9:28That's one of my virtues. Like you know, you are that like
- 9:31that takes you to a different route, like different dimension
- 9:34and there's so many different extensions, so many different
- 9:36utilities We can talk about on how to virtualize things.
- 9:39You have like I have a 5 agents working with me when I what's
- 9:44called vibe code. 1 is an architect, 1 is a product
- 9:47manager, 1 is a like you know, and they all they all talk to
- 9:52each other. They talk to me.
- 9:52I'm a main input, but they helping me fast track things so
- 9:57much faster. They're contextualized on the
- 9:58project of what I working on, right?
- 10:00So you can like they understand the semantic meaning of what
- 10:03we're building. So they're not just, you know,
- 10:05Claude agent who has 0 context of the project you're working
- 10:10on. And then you can extend on that.
- 10:12You can kind of, you know, train them on the data which you're
- 10:15building or trying to build on. So they become smarter.
- 10:18So it's it's unbelievable like what you can do on your on very
- 10:23this laptop we're just talking on.
- 10:25To breast point before this is all pretty new like the whole
- 10:27like the Gentex workflows like really like got like I'll do it
- 10:31like last November, December, they'd really got to a point you
- 10:34could just deploy them. So it's it's new and it's
- 10:35amazing like how quick how quick it's moving.
- 10:37You know it's. It's it's really.
- 10:39It's really. Remarkable.
- 10:40So yeah, it's cool to hear you're you're digging in as well
- 10:43and like as many of us are. So you have any maybe just get
- 10:46getting back to the to the the questions here, like looking at
- 10:49so you worked across pubs, agencies, data companies,
- 10:52programmatic platforms. Fair to say this was like is, I
- 10:54mean, you know, some really cool jobs, but fair to say this was
- 10:57doesn't seem like it was a deliberate strategy.
- 10:59It's more like you did or was it right?
- 11:01I love it. If you could tell us about that.
- 11:04Yeah, thanks very to keeping me in check.
- 11:05Otherwise I'll be a genetic all the way.
- 11:07So yeah, back to the. We could.
- 11:08We could RIP one out for a while.
- 11:10I know, I know. Back to the story, so so coming
- 11:13out of the you know, you know, that exit and joining 9 MSN,
- 11:18which is JB with Microsoft and working on that, you know, kind
- 11:20of bringing the Microsoft ad Exchange in 22,000 and nine and
- 11:24working in Atlas and exhibitor and the names probably only
- 11:28Brian O'kelly and you guys remember, right.
- 11:31So yeah, I remember that. Yeah.
- 11:33So, so working with those, the building the behavior profiles
- 11:38and the retargeting for the first time Bridge Media anyway,
- 11:41so that was kind of invitation to advertising on the publisher
- 11:44side. But then one of the recruiters
- 11:46came up to me and said, yeah, this is just one startup
- 11:49launching in Australia called Brand screening.
- 11:53It's one of the first Asian DSPS.
- 11:55Would you like to join that team to build a product and
- 11:57technology and being part of that team to effectually look
- 12:00after partnerships? And I said, what is open RTB
- 12:03like, you know, and that was that was the 2000 I think 10
- 12:09when you know, yeah, when I just started and by accident it's
- 12:14almost complete pivot. I took a leap of faith.
- 12:16I was very comfortable where I was looking.
- 12:18I had a big team, was a treasure, FBI and physical stuff
- 12:21and, you know, innovation in rich media.
- 12:24But then pivoting to absolute unknown and taking like a leap
- 12:27of faith on technology, which I did not understand, but that
- 12:33propelled me just, you know, to this universe of first of all,
- 12:37you're not the commodity. Like no one knew what
- 12:39programmatic was back in the days.
- 12:40You know, one of the first people who building something,
- 12:43you're talking to Edmunds of the world.
- 12:45They're talking to companies that don't exist or being sold
- 12:47like wires, like exolates and possibly.
- 12:50Excellent. That's right, that's right.
- 12:52I was and so it's it's funny, we all have that origin story.
- 12:54You certain? You're just like.
- 12:56I remember my eyes being wired up and going I can't believe all
- 12:58this is possible. And Godhead, we've come a lot
- 13:02further since. This so you have any a little
- 13:04little background here, like Brett and I met at a small media
- 13:06agency where we were actually, you know, it's pre programmatic
- 13:09or we're just we're just doing like buying kind of.
- 13:11Arbitrage. Yeah, a lot of it was arbitrage,
- 13:14right? Yeah, it's performance marketing
- 13:15stuff. Yeah, a lot of it was CPA or
- 13:17they would pay you CPA and be arbitrage your CPA for different
- 13:20like friend response brands. You know, that's how we, that's
- 13:23how we met. Yeah, that's awesome.
- 13:25That's awesome. Yeah.
- 13:26And then eventually that type of tech orientation to your point,
- 13:31get you connected to other places and then you're like, and
- 13:34remember when big data was a term like big data, that's what
- 13:39kind of led me to exlate. I was like, I want to get into
- 13:42this space before they were even a data management platform.
- 13:44They were a data marketplace that just had, they were like a
- 13:46cookie farm that was founded in Israel.
- 13:50And so I joined that and then we sort of evolved that.
- 13:54But that was a yeah, getting early in in that journey and and
- 13:57seeing how it's evolved has been.
- 13:58Fascinating. Oh, it's so fascinating, Brad,
- 13:59like you know, blue guy just entered the Australian market,
- 14:03you know, and and and old March apology was like, wow, we know
- 14:07D&P, right. So like data man, like big data,
- 14:10right. So to your point, like how do
- 14:11you manage customers data assets and what that means in marketing
- 14:14activation side of things? It was phenomenal.
- 14:17So that was kind of the inception of like programmatic
- 14:20going into China, working with Baidus of the world back in
- 14:232012, burning out on cost because its restructure was so
- 14:27expensive back in the days, like so many mistakes, so many, so
- 14:31much to learn. In fact, Cadreon was had one
- 14:35person in Australia and we were building an ADSP just for them
- 14:40right, and that for Viviki as well.
- 14:42And kind of, you know, just this age is a trade that's AT DS.
- 14:45Remember AT DS, AT DS was kind of just just surfacing up and we
- 14:49competing against media math, the brain, the turns data zoom
- 14:54thing a. Lot of rips.
- 14:56Here and and maybe maybe that's a trade that wasn't even born
- 15:04yet you know So like yeah, yeah, yeah, yeah.
- 15:07So that was like the epoch of at corneas and AD networks, right?
- 15:10So that was. I remember that.
- 15:12Yeah, yeah. That was like the height and the
- 15:15but close to the end of the ad network thing, yeah.
- 15:18That's right. So building on that and
- 15:20developing technology, which was amazing.
- 15:23So that's kind of was my induction into the into the
- 15:28programmatic, into real programmatic world.
- 15:30Three people in Australia knew what, you know, what the hell
- 15:32that was going on and I wasn't one of them.
- 15:36So and. This was all before.
- 15:38This is before you joined Verve, right?
- 15:41Well, it's before news, like I saw that it's after that News
- 15:43Corp and like, you know, remember another acronym,
- 15:45another for that acronym, not Agency Trade Desk, but Publisher
- 15:48Trade Desk. So remember those ones.
- 15:50So we kind of built in the programmatic capabilities.
- 15:53Think of it, you know, the the waterfall after, you know, they
- 15:56run that inventory around those conversations that how can we
- 15:59sell it at at the open auction and building a stack behind it.
- 16:02So I was instrumental of, of actually building and you know,
- 16:07bringing partners like Rubicon Project and Adapt TV back in the
- 16:10days and building it's. Kind of like the, the, the like,
- 16:13like what like before we actually yield management
- 16:16solutions even came out right. This is kind of predecessor to
- 16:18that. Real that was the like I'm proud
- 16:21of this moment. You know I'm not sure people
- 16:23know that, but we were the first we ran the first PMP in
- 16:26Australia for a Rubicon project like that was the first PMP on
- 16:30News Corp. That was first July D of a
- 16:32crater in Australia, you know, for that technology and.
- 16:36That's not Mac night. Yeah, so you seem.
- 16:39Like and and exhalates. Mark Zagorski was the one that
- 16:42brought those companies Tremor the what Tremor Video or Tremor
- 16:45Media was originally, which became Talaria that joined
- 16:48Rubicon project and they formed it became Magnet.
- 16:51Yeah. So you've you've been on a lot
- 16:52of these pivots, right? Like, like, you know, Load to me
- 16:55went through a bunch of different pivots and it started
- 16:57out as sort of a data management platform and and they got into
- 17:00data collaboration. They were like the last DMP
- 17:03standing. Everybody was like when, you
- 17:05know, Exhalate sold and aggregate knowledge sold and
- 17:08Blue Kai sold and everybody's like what's going on?
- 17:11Tell us a little bit about that experience.
- 17:13And Load and and Verve similar similar.
- 17:15There are a lot of visits with Verve you lived.
- 17:19Right. Like a lot of me is Sasha and
- 17:20Andy Mott for a big shout out to Andy, you know, a lot of his
- 17:25energy, his vision and kind of, you know, joining his team
- 17:30effectively initially on the DNP side, just selling SAS to
- 17:33publishers side of Australia, moving to Singapore.
- 17:35But then we realized, hey, there's a untapped opportunity
- 17:39to build a data marketplace in Asia, in APAC.
- 17:42So I've started doing that as a solo kind of a lonely wolf.
- 17:45They called me, you know, running eight different markets,
- 17:48you know, going in, trafficking ads, you.
- 17:50Were effectively like a general manager of.
- 17:51That like, yeah, well, I don't know if you could call me
- 17:53because I was in DV for 60 trafficking ads.
- 17:56I was trafficking ads in DVI was going to publish this acquired
- 17:59data. So I was like end to end from
- 18:01acquisition to to productization to go opening emerging markets
- 18:05like Taiwan or Hong Kong. It's like Jeff was Jeff Green
- 18:08was in Hong Kong when we were kind of working with trade as he
- 18:11was living there, you know, so interfacing with him back in the
- 18:14days when he was approachable, you know, in that capacity
- 18:19before obviously, you know, the growth I have had.
- 18:22So being in that stream was amazing.
- 18:26And frankly, you know, just too fortunate to have the
- 18:29opportunity to grow with a lot of me.
- 18:30They moved me from from super ports into New York to always
- 18:34see global business for them. And in fact, you know, I do take
- 18:39some credit for for that exit because the team I did my
- 18:43leadership with, obviously with Andy and everyone else, that's
- 18:46the data signal which Publicis needed.
- 18:49They acquired a lot of me for international traffic and can't
- 18:53take credit for it. But the team would have built
- 18:55together studying in Asia and moving into Latin America and
- 18:59EMEA. That's the signal which was
- 19:01complementary to Epsilon PII data, which they have at scale.
- 19:05So that's, you know, to your podcast signal, that's what they
- 19:09bought it for that. Was the signal, Yeah, it was the
- 19:11international signal. That was it.
- 19:12Absolutely. And that's very, very hard to
- 19:15get. I mean, we saw that with Excel
- 19:168. I mean, you know, I was a
- 19:17TransUnion where they have credit data, but you know, which
- 19:20is certainly multinational, but it's very hard to get once you
- 19:23get outside of the US because of governance and privacy
- 19:26constraints. And it requires partnerships.
- 19:29Tons of tons of. Like it was tons of work, if
- 19:31you're right, like, you know, and if a lot of me benefit from
- 19:35many things. They had a GMP business.
- 19:36So a lot of broadcasters, publishers, they had that GMP
- 19:40technology to understand their consumers and to profile and to
- 19:43cluster them, you know, to activate them.
- 19:45So you have that technology already deployed in a lot of
- 19:47networks and pubs. So we use that footprint to
- 19:51monetize some of their data anonymously for the marketplace.
- 19:54You had that leverage of SAS integration across different
- 19:58geographies as a backbone. Effectively build upon to build
- 20:02a marketplace on it. So yeah, so that's why I put
- 20:05that exit. Just a little tick and you
- 20:07probably will shout at me, but you know what?
- 20:09I'm proud of that work and I'm going to stand by it so.
- 20:12Yeah, yeah, you're a fundamental part of the team that's worth
- 20:16counting. Andy's one of the most liked
- 20:18people in the industry, like he's almost universally liked.
- 20:21I mean like Adam Solomon, who was there for a little bit.
- 20:23He's he's, I'm pretty tight with him and he's only good things to
- 20:26say about him. So must have been a good
- 20:28experience working together. Sounds like you had a good
- 20:30outcome, so Congrats on that. Yes, and Adam as well, like
- 20:33worked with Adam while he was there and Auntie.
- 20:36So yeah, that was it was a great team, great team.
- 20:39So then, so then you move to. Then you're at.
- 20:41Verbs verb follow. To me, I'm getting the
- 20:45chronology wrong here. Yeah, I'm so am I now, verve.
- 20:50Like I remember Verve back in the day in New York City, they
- 20:53were like a location data company, correct.
- 20:55You know all the, all the, all the, you know, mobile, mobile.
- 20:58Data kind of Foursquare in some ways, right?
- 21:01Yeah, like. And all the XLA people are
- 21:03complaining, you know, complaining that it was, oh,
- 21:05it's crap data. You know, we had cookies.
- 21:07So I'm not sure how much how much worse off better off they
- 21:10were or worse off they were, but yeah.
- 21:13Yeah, I'll tell you about the worst very interesting story.
- 21:15So parent company Verve is MGI media game invest.
- 21:20So what they really focus on is gaming or not rated on console
- 21:24on your phone and on laptops, right.
- 21:26So you have the gaming mogul with with a lot of games and
- 21:32they said, well, let's just start the ad business because we
- 21:36have a ton of PII data for gaming and we're going to infer
- 21:39behavior for gaming, for purchase, etcetera.
- 21:42So let's just buy businesses on both buy sell side and build a
- 21:47business out of this and they so Remco was behind orchestrating
- 21:51all of that and it's very knowledgeable and knows how to
- 21:54acquire businesses at the right moment for the right price.
- 21:57That's his. That's his.
- 21:59Superpower. And it is, yeah.
- 22:00He's just, he's just, he's just good at that so.
- 22:02Yeah, that's, that's an important CEO superpower.
- 22:05If like you, if you don't have all of the organic growth you
- 22:07need, you need inorganic growth, which can true right can help
- 22:11bring some cash in the door. The ability to bring in money
- 22:13and ability to identify acquisition targets and pull the
- 22:16trigger out, yeah. Yeah.
- 22:17And they, you know, and they did this like they and they're still
- 22:20doing this, you know with Captive 5 acquisition and Junior
- 22:22Group etcetera. So like you have a plethora of
- 22:25businesses on buy and sell side which they trying to orchestrate
- 22:28it and make sense of it. And at the time like they they
- 22:32was need of transformation and consolidation across both sell,
- 22:35buy and sell side. They had three SSPS.
- 22:38You probably know them, Brad Pub native and smarter, and I think
- 22:42yeah. Yeah, yeah, yeah.
- 22:43You know, so like they bought a lot of a lot of good meat and
- 22:46mobile fire first SSPS and then they have they have acquired
- 22:50like 3 GSPS and contextual player and OK, now they have an
- 22:55ecosystem play. And more importantly, how do
- 22:58they would that propagate the gaming data through it as your
- 23:01unique source and differentiation.
- 23:03So back to the like AFC, every conversation we have, we've go
- 23:06back to the signal like that's, yeah.
- 23:08And was it still was it, did it continue to be mobile centric
- 23:12despite the game you know, or did the gaming acquisition?
- 23:15Did they buy, did they buy the games themselves or did they buy
- 23:19the publishers or just the games?
- 23:20Like how did that work? Yeah, they're both the publisher
- 23:23and it produced a lot of like, I think 5000 games or something.
- 23:27You know, like car racing games, you know, like little utilities.
- 23:32Honestly I'm not, I'm not the big gamer but they'll even
- 23:34believe it or not. My dad is 80 too and he plays
- 23:37games anyway. So so they're.
- 23:39Business, I mean, that's all that everyone knocks on app on
- 23:41app lovin all the time and you know, but you think about their
- 23:44whole business, it's just mostly just mobile, mostly games.
- 23:46So yeah, keep going. Yeah, yeah, it's certainly
- 23:48thematically reminds me of App Lovin and what they did to build
- 23:51their ecosystem. But yeah, they became a
- 23:53mediation for that, right. So they had supply, they had,
- 23:56they had large demand. And you know, the consolidations
- 23:59was, you know, I came in as a head of international and kind
- 24:02of mapped it together. Yeah, and I was interesting.
- 24:05I never managed CEOs, right, although there are small
- 24:09businesses, 50 people companies which have been acquired, I
- 24:11never managed people you know, who had exit strategy, who had
- 24:15the owner strategy. So, like I actually learned a
- 24:17bunch on how to have, you know, follow the incentive on how to
- 24:22imagine right trying to consolidate businesses with
- 24:26potentially, you know, conflicting interests and or
- 24:29self-serving interests, you know, which rightly so if I was
- 24:32a Co founder, the. Tech becomes the the tech of the
- 24:35data is the easy part. It's actually the mechanization
- 24:38differences, the incentive structures right the in
- 24:41alignment across organizations that that becomes the hardest
- 24:44part. You get fiefdoms, but that's the
- 24:47hardest part. You know what?
- 24:48But that's such a challenge. You know, you can look at it.
- 24:51How do you learn from that? How do you, you know?
- 24:53You know, for me that was a great learning curve and and
- 24:56verbose was great for that. Yeah, in between that had a bit
- 25:00of agency experience and you know, and now I'm now I'm at
- 25:04Samba. So kind of I think I work for
- 25:06everybody except except the brand.
- 25:08So brand if if you know, if you know you're looking to.
- 25:13Swing over from the in. Three years time, you know, if
- 25:15this next step is probably, you know, working with the CMO that
- 25:19was that would be banned, but. Yeah, yeah, I've never worked.
- 25:22Brand actually, Brett, have you? Worked.
- 25:24Have you worked for? Brand, I don't think you.
- 25:26Have I've not, I've been on the vendor tech side forever, yeah.
- 25:31Yeah. So, no, but Samba TV, I mean,
- 25:34everybody knows it is CTV. Maybe that's where their
- 25:37knowledge ends. Obviously it's data.
- 25:40There's some measurement. You have ACR technology, which
- 25:42kind of brings me back to Nielsen because ACR was talked
- 25:46about all the time. That's automatic contact,
- 25:48automatic content recognition. Yep, that's right, right, Which
- 25:52is, which is basically the ability to track on the
- 25:55television glass what's being watched in all of the content
- 25:59metadata like Grace note did that, which was required by
- 26:02Nielsen. But tell us because, because I
- 26:05think we kind of walked through this trajectory for a reason.
- 26:07There's a purpose, we there's a purpose.
- 26:09We went through your this be enduring path of, of Yen's
- 26:12background because because it's a really interesting perspective
- 26:15to bring to the television and connected TV world.
- 26:19And then how do you see all that connecting and, and tell people
- 26:21a little bit about Samba TV as well and, and what you what your
- 26:25vision and mission is? So for me it was, I think,
- 26:28natural progression. I was like, you know, what is
- 26:30next, what what areas of unfamiliarity I can enter to
- 26:34like to get better AT and you know, I never I sold HCI data
- 26:38for lot of me like we had that we had Visio data back in the
- 26:40days, you know, escape before the could be for the for the
- 26:45problem. So who doesn't have Visio data?
- 26:48Right. That was like bad prior, but you
- 26:50know, I really want to get into, you know, the measurement side
- 26:53of things. And, you know, I'm sure you want
- 26:54to talk about that, but also understanding the technology
- 26:58behind it because it's it's fairly complex, but might be a
- 27:01bit controversial because Samba TV has rebranded and the turf,
- 27:06you know that Brighton Rio, we we dropped the TV at CES.
- 27:10So we have somebody where somebody and you know, and this
- 27:13is yeah, I'll tell you why. So it's somebody like had a
- 27:16legacy. Edit that out in post production
- 27:18because you might you might face some scrutiny from the PR and
- 27:22comms team. We'll see that, Yeah, yeah.
- 27:24No, I think it's, but I think it's a public knowledge.
- 27:26Yeah, yeah, yeah, but, but, but you know, I've been wrong
- 27:29before. So, so we'll find out.
- 27:32So historically, you're right, Brad.
- 27:34So you know, Samba basically is used to be television data and
- 27:39measurement companies specifically, you know, working
- 27:41on quantum, automatic quantum recognition technology.
- 27:46But since then we have acquired a company called Samazio, which
- 27:50is you're probably familiar with, right.
- 27:52So Casper's business which was based pure contextual.
- 27:56So what we have done is mapped TV, let's say attention in a
- 28:02viewership signal and then intentionality of the web.
- 28:06So you know 2.5 billion URLs that you crawled, you can
- 28:09understand context, content, semantic meaning,
- 28:12classification, pure NLP for the sprinkle of NLU of understanding
- 28:17how you derive that metadata and then you map it together for our
- 28:20ontology graph, which is knowledge.
- 28:22So you can connect down to show level, content level time spent.
- 28:28And are you licensing, so are you licensing this to the
- 28:31manufacturers, the OEMs, are you licensing this to the CTV
- 28:34companies? Because there's, so my
- 28:35understanding is like reviewing the content, taking snapshots
- 28:38and turn under like providing those signals, right, which you
- 28:41turn into that ontological graph.
- 28:42So like, how does the business model work?
- 28:45Yes, that's it. So the data is coming from 25
- 28:48OEMs, Rio and it's a global database is one of our major
- 28:51major US PS. So for example, in Germany,
- 28:54Philips is still very predominant TV player and we
- 28:58have you know Sony's TC, LS handsome of the world where you
- 29:01know it's fairly representative long tail of OEMs. And we love
- 29:05that, love that for a reason because geographically,
- 29:08psychographically, you know, income level distribution of who
- 29:12buys the HTVS gives us a representiveness across
- 29:15geographies. So that's very important
- 29:18diversification aspect on who which households and what
- 29:23properties of this household own those TV's and at what price
- 29:27point, right. So you have the representatives.
- 29:29So we collect that lead data on the chipset level, right?
- 29:31So that's how ACR technology works.
- 29:33You have your, you know, visual and audio fingerprint, you know,
- 29:36couple couple fingerprints per second.
- 29:38In fact, you collect the data and then you more importantly
- 29:42you map it against a large content library which you have,
- 29:46right? So you then you can get what
- 29:48program or ad was on the screen when it add which household
- 29:52sword for example, and then gives you that level of, you
- 29:55know, ACS is very powerful because it gives you that
- 29:58neutrality, right? And also gives you
- 30:02holisticality, linear cable, streaming, gaming and.
- 30:06You're capturing this at a device ID level, right?
- 30:10You're capturing these signals and then I'm assuming you're
- 30:12not. You probably anonymize it to a
- 30:15certain extent, right? Put it into these models, but
- 30:17then you can drill down to see which content was reviewed at A
- 30:20at a household level. Exactly, yes and then you have
- 30:23you know and then we have identity craft which we have
- 30:25acquired one of the Dutch businesses friend 6:00 which
- 30:28gives us kind of identity resolution gives us map and
- 30:30crosswalks between you know any identifier you can think of and
- 30:34it becomes. Yeah.
- 30:34And that's where you would map across channels once people go
- 30:37off the television glass to mobile to desktop, to laptop,
- 30:40etcetera. That's right and gives you like
- 30:42household, you know graph understanding and who is viewing
- 30:46what at what time and also ability to then the attribution
- 30:50against that. So.
- 30:51So so I got a question, just got a curiosity.
- 30:53ACR. Is that actually what is the
- 30:55actual hardware behind ACR? So Brad, do you think of it as a
- 31:00as a little program running on the chipset level within, you
- 31:06know, which is deployed that are firmware into your TV set?
- 31:10So you basically when you every first time you start a TV you
- 31:13have to consent to it saying you know, I consent on Sony for
- 31:16example, on the TV's you consent to be, you know to send this
- 31:19data for ACR purposes. And it's installed on the
- 31:22hardware itself. It's certain hardware itself, it
- 31:24can be rewired with the updates because, you know, you, you're,
- 31:28yeah, like think of it as an OS for TV, right?
- 31:31Which is and you can kind of reload it and understand how you
- 31:35capture what you capture. But that requires that
- 31:38relationship with the OEMs where you're actually at the
- 31:41manufacturing plant when they're putting shift 138 into like
- 31:46that's actually where it's happening at that.
- 31:48Right. That's why CS is like we've done
- 31:51CS is our not just temple event. This is where we've started the
- 31:54business. That's the event for you, Yeah.
- 31:56The event, this is where the TV's manufacturers were coming
- 31:58in before there was smart TV's with some tickers at the bottom
- 32:02saying that you know, like this is the first generation of apps
- 32:05we know. And that was that when Ashwin
- 32:08and and executive team effectively start doing those
- 32:12deals and start collecting, you know, OEM relationships which
- 32:16became fairly large, you know footprint for for Samba
- 32:21globally. So.
- 32:22Yeah. And is there, is there a
- 32:24modernization fee and then how, how are you paying, what's the
- 32:27compensation for for the the TV manufacturers, the OEMs?
- 32:31So there's different different deals with different
- 32:32manufacturers. The idea is to and all those
- 32:36different restrictions, Brad, like for example, Netflix
- 32:38blocks, you know, not just us, but Samsung, LG, every
- 32:42everybody. So that there's like there's
- 32:43four restrictions of what content can be recognized,
- 32:46what's cooperated, what's not. So there's like, you know, as
- 32:49you can you only imagine the fragmentation of that
- 32:52relationship is is a nightmare. So you have, you know, you have,
- 32:57but there's also little workarounds.
- 32:59For example, if you Brad, I do sometimes a plug in to my HDMI
- 33:03to my TV and use it as a screen. Well, now it's not a TV anymore.
- 33:08So if I watch Netflix on my laptop and screen it.
- 33:13Yep, the source of the canonical devices.
- 33:18Your laptop not correct. So like, you know, you know, the
- 33:20actual fingerprinting of the TV app itself, you just stream it
- 33:24as a source, as a projective actually, right.
- 33:26So you have all different scenarios, fire sticks, etcetera
- 33:29and how content is, you know, can be consumed on the on the
- 33:33screen and then all the, you know, legalities around that.
- 33:37So assuming you pay the OEMs then to for to be installed,
- 33:41right, and our viewers may not know this, but like you know,
- 33:44the prices of hardware just gone down so much the last 10-15
- 33:47years. A big reason for that is the
- 33:49OEMs are able to make to make money because of these types of
- 33:52deals, right, where they're actually sharing data.
- 33:54I'm assuming this is. That's it.
- 33:56That's a good description of it. In fact, I believe in some
- 33:59instances hard ways, negative margin, negative, you lose,
- 34:03yeah, you lose money on TV's. So they obviously supplementary
- 34:06income is to have this type of. Relationships, so that's how you
- 34:11can buy like a 72 inch flat screen for like 500 bucks at
- 34:15Sam's Club and that's why you can do that because of.
- 34:18You remember like there's businesses, is it telling?
- 34:21I think maybe I mean, I'm misquoting, yeah, who giving us
- 34:24hardware for free. But then there's there's
- 34:26monetization like like so that's not far from reality because to
- 34:29me, you know, is already thin, thin margin, single margin,
- 34:33single digit business. So, so they, the service is a
- 34:37symbiotic and very healthy relationship.
- 34:39And then brought to your question, like, what do you do
- 34:40with the data after that? You know, obviously there's a
- 34:42plethora of use cases on how that data can be used depending
- 34:46on who you're talking to. Obviously, as you understand,
- 34:50any of every ad which been aired, you know, we have we have
- 34:54we have viewership of that data. And so it can use, be used for
- 34:58for conquests, it can be used for business intelligence for
- 35:01market channels. So they're like this utility
- 35:04even outside of targeting. So plenty discovery, you know,
- 35:08is, is just the beginning of the game.
- 35:10We have companies work with us purely for analytical use cases.
- 35:14So for example, companies who look at this data.
- 35:17What was the last show you watched on Netflix before you
- 35:21moved to Amazon Prime? You start watching XY and Z and
- 35:25then so churn analysis, for example, you know, and that
- 35:29becomes a very large business to go back to sports to basically
- 35:33to build a more sophisticated models on what people watching.
- 35:36Why this changing? Why they're shifting
- 35:39fascinating, really fascinating utility of what you can't think
- 35:42of that data. So it's not some people say, oh,
- 35:44ECR just, you know, targeting Desperate Housewives on, you
- 35:47know, on, on desktop. Yeah, Yeah.
- 35:49Maybe this is for customer growth, customer retention,
- 35:52right? New, a new customer comes into a
- 35:54new platform like like Prime, and you want to show them
- 35:57content recommendations that are relevant to them.
- 36:00And you're pulling that from their past behavior on a
- 36:02different platform, possibly on a different device, which is
- 36:05pretty powerful. That's right.
- 36:06That's right. And so do you think you, you
- 36:09mentioned Telly, which they their tagline is great.
- 36:12It's the biggest thing to happen since color TV.
- 36:15That's a pretty bold statement. Do you think Telly has?
- 36:19And I've talked to a few of the folks over there.
- 36:22They have a waiting list for their devices, apparently, like
- 36:24people really want them. So yeah, keep going, Brett.
- 36:26Yeah. Do you think, do you think that
- 36:28there's a future in that? And it's, it's in like the
- 36:30vision that I see it is, is it's basically like your, your
- 36:33iPhone. They want to make it this sort
- 36:35of device that's central to your life.
- 36:38It's a lot smarter than what televisions have really ever
- 36:41been, you know, so you can do. It's kind of like your mobile
- 36:43phone has turned into really a a handheld computer, right?
- 36:48I'm torn, but I'm I'm not sure, Brad, like, you know, I think
- 36:51it's consumer choice, right? You know, some, you know, low
- 36:55income families might find it more attractive because it's a
- 36:58free hardware and, you know, gives you the basics of what you
- 37:01want to watch. But then it comes with, you
- 37:03know, comes with ads, it comes. With yeah, and the AD model I
- 37:06found a little weird because it sounds like there's some almost
- 37:08like some built in AD redundancy where it's like you get ads on
- 37:12the, the actual physical device and ads within the, the, the,
- 37:16you know, the streaming services that you subscribe to if you, if
- 37:19you or if you, if you're doing fast type of channels.
- 37:23So I'm like, I'm like, so you're actually kind of are you
- 37:25doubling your ad load? I mean, as a consumer, it
- 37:27doesn't seem that. Attractive for me personally as
- 37:31a consumer, I think that's not, you know, it's not experience I
- 37:34would like to have in my in my living room, right There's also
- 37:37I'm I'm maybe a speaking completely wrongly here like you
- 37:42can't cover the panel. I think because there's a camera
- 37:44detection or something, there's some sensor, you know, so like,
- 37:47you know, there's this that I don't know how privacy, if
- 37:50privacy, there's just. Apparently you can see if you,
- 37:54well, can see if you're sitting down to watch.
- 37:56I mean, apparently there are some sensors.
- 37:57I don't know how. Privacy.
- 37:58Oh boy, oh boy. Apparently it's pretty.
- 38:00Yeah, sounds it's been. Intrusive I I probably enough
- 38:02with the ring you. Know listen, I was about to say
- 38:05I I have enough problems with my Roomba vacuuming my my my floor.
- 38:08You know, like I would have. The television is watching.
- 38:12Yeah, yeah, I already have a problem with that with with the
- 38:15cleaning devices. But that but that's one thing
- 38:17that Samba TV can't do because it's Speaking of of watching,
- 38:20unlike the the sort of Nielsen viewer.
- 38:22And if people don't know, you know, Nielsen television, you
- 38:26know, it's kind of the old Simpsons episode, right?
- 38:28You get sent controller that literally has buttons on it, at
- 38:33least this is how it was when I was there.
- 38:35And the buttons are like Person A, Person B, Person C, Person D
- 38:38So you identify yourself as you start to flip through content
- 38:41so. They, I don't think it's changed
- 38:42very much. I mean, they want, they're
- 38:44building a panel. It's not a big panel.
- 38:45They, you know they. Just so they can tell who's
- 38:47watching you. You would have to probably back
- 38:50into that data I'm assuming. So we don't use panel for that.
- 38:54So, you know, we don't need to have a kind of a human input.
- 38:58So that's just different, different business model.
- 39:01Our strategy was bread is to have, you know,
- 39:04representativeness by just having a 25 type of global deals
- 39:10with OEMs to have to just have enough density for for for let's
- 39:15say local affiliate network to say in that particular town we
- 39:19have this much television, 150 TV's which are not noise which
- 39:24not in the pubs or not in the gyms which actually in the
- 39:26household. So removing the noise, right to
- 39:29be sure, OK, we have enough representatives to say that we
- 39:33have enough viewership signal to support XY and Z use case.
- 39:37That always was the case. That was always the case.
- 39:39To support investment because you know that you're going to
- 39:41get. That's right.
- 39:42If someone is buying on that Geo, on that GMA, on that zip
- 39:46card like Internet log KL specifically like you have
- 39:49representative data to do both measurement and an entity derive
- 39:52inside of that. So that was really kind of just
- 39:57the model Samba has has built and it kind of played in our
- 40:01favor of versus where you have maybe more high end TV's from
- 40:06Samsung or LG which are more media businesses versus Samba
- 40:10where we don't touch media. We we are agnostic in that sense
- 40:14and representative so. I like that angle, you know, you
- 40:18know, as you real speaking to real laughing about that in
- 40:21person, like there's just so many smoke and mirrors and I
- 40:23understand you guys know it. I've been there.
- 40:25Not your first rodeo. So I'm like, give me a
- 40:28substance, give me consent of data which is unique and which
- 40:32has utility and has, you know, in somewhat future proofed.
- 40:37You know, there's obviously legislation and all everything
- 40:39else which is flying around, but you know, that was kind of how
- 40:42can I make the best bet possible?
- 40:45And I was like, that's sound like a good bet and a good team
- 40:48so. You know, and future proof just
- 40:49to yeah, because it's in the hardware.
- 40:52It's not an IP address, which is not future proof, regardless of
- 40:55what people want to tell you. It's not, you know, even e-mail
- 40:58is technically not future proofed.
- 41:00But and you have the relationships, you're in the
- 41:02hardware and you have the relationships with the OEM.
- 41:03So I think that's that's very important.
- 41:05And plus, you know, like talk about future proof like TV, like
- 41:07you know, I was thinking of reading the other day, it's
- 41:09projected overall projected to grow 16, 17% over the next
- 41:13couple of years, which is pretty significant, right?
- 41:16CTV. Right.
- 41:17All TV obviously linear is declining, but CTV, the growth
- 41:21in CTV is making up for it. So it's a good industry to be in
- 41:23and. You know, like a Amazon slogan
- 41:27is, you know, your market is my opportunity for Samba.
- 41:30If I had to rephrase that, I'm just thinking on the fly.
- 41:32Fragmentation is an opportunity like really.
- 41:36So if you look in how, how the content is consumed in the, in
- 41:41the, in the living room, that's what's, that's what Samba solves
- 41:44for, right? So cross-platform measurement,
- 41:47you know, how many unique people did I reach across all TV's?
- 41:52And I think this is what was solved for.
- 41:55And he's still like to Brett's point, you know, measurement is
- 41:58fragmented. You have live TV panels we spoke
- 42:02about in a platform specific streaming data, digital ad
- 42:05measurement is different. So industry still needs a
- 42:09unified currency layer, right? You brought up that word, the
- 42:13currency layer. Dangerous one.
- 42:17OK. And by the more you go and I
- 42:19stand correct is CTV is growing at around 1617% and then linear
- 42:24there is declining by around 7%. But overall it is.
- 42:27Streaming is about 45% of total television consumption, with
- 42:32linear being about 44%. Yeah, yeah.
- 42:34Linear's declining. Yeah, CTV.
- 42:37'S grown at a much larger rate than any media channel, right,
- 42:40which is that 1718% level. So let me ask you this, when it
- 42:43comes to to viewership and measurement, right, because you
- 42:47guys have this deterministic sort of hardware specific ECR
- 42:50data that's built into a huge volume of televisions globally.
- 42:55Are you able to then definitively break apart the
- 43:00stuff that people are claiming is CTV, right?
- 43:02Really sit bifurcate other media that's might be called CTV or
- 43:07streaming, Streaming X being, you know, content that's being
- 43:10consumed on other devices that's not actually being consumed on
- 43:13the on the television screen. Because there's a lot of actors
- 43:15out there faking CTV audiences which are actually not being
- 43:20consumed on the television class.
- 43:22They're being consumed on on, on, you know, laptops and mobile
- 43:26devices, right. So you guys I'm assuming have
- 43:28have the ability to sort of help, you know, separate the
- 43:32wheat from the chaff in that respect.
- 43:35You're right. And like by definition this is
- 43:38what ACR is, right? We observe everything on the
- 43:40screen. Linear broadcast, cable,
- 43:43streaming apps, even gaming, fast channels, OTT advertising.
- 43:50So you have all that signal consumed in the real time.
- 43:53So you really have that true cross-platform visibility which
- 43:58is important. So if you're talking about
- 44:00breadth measurement, can we one of our biggest products
- 44:05obviously on the measurement is true rich and frequency
- 44:08cross-platform, right. So we could also can you solve
- 44:11for incrementality across linear versus you know your digital
- 44:15problematic. So we have partners who use that
- 44:17just for that. So that's why when I started
- 44:21with definition of Samba, although we dropped the TV and
- 44:24we have all the different data sets in our map together in our
- 44:28heart, we still measurement and data company and always has
- 44:32been. And I like the neutrality aspect
- 44:34of that. So we don't claim we remain
- 44:37agnostic and that's what's always our thesis and how can we
- 44:42evolve? And by position of semaz, you're
- 44:45kind of fusing like what other signal can we fuse in into this
- 44:48ontology graph? You know, if you're watching
- 44:51Desperate Housewives and then you're doing research on, I
- 44:54don't know, on you become travel intended at the same time, like
- 44:57you have very interesting mix of your intentionality.
- 45:01This is, by the way, why verve both captified and exactly for
- 45:04that, for that data. So we have that contextual
- 45:09understanding and meaning what you are reading, you know, what
- 45:13websites are you visiting, what this, what the sequential and
- 45:16the storytelling varies. And then what are you watching
- 45:18at the same time within that household and the graph to map
- 45:21it to an identity level. So you have that level of
- 45:23understanding. And then from there you can, you
- 45:25know, start building 2 outcomes, you can start doing measurement,
- 45:30you know. And does that identity graph
- 45:32does that? What's the scale and the scale
- 45:33and scope of that? I mean, is that multinational
- 45:35like your, like your television IEM footprint?
- 45:38So yeah, how extrapolated is it? Or is it a a data set?
- 45:43That's pretty pretty gigantic. It's I believe it's 50 million
- 45:48TV's in that US. OK, I probably should have come
- 45:52prepared. Yeah.
- 45:54So that sounds like it's an identity graph that gives you a
- 45:57seed audience to be able to carry, you know?
- 45:59So this is your no, this is, this is ATV, this is the
- 46:02households we have. So like this is not extrapolated
- 46:06number, right. So and then you have 2.5 billion
- 46:10URLs which was scanned. Right.
- 46:12OK, OK. Yeah, that's.
- 46:13That and there's like a 1,000,000 billion of maids and
- 46:16like you know, all IDFA. Connected to the 50 million is
- 46:19the actual OEM. Yeah, yeah, yeah.
- 46:20OEMs which are filtered out from, you know, and there's a
- 46:22pet. Well, you compare that to
- 46:25Nielsen, right, to those numbers, right?
- 46:26You figure like what's Nielsen 42,000 in the US, which is
- 46:32around 100,000, 40,000 household, which are 100,000
- 46:36people overall, right? So I think you compare it
- 46:39that's, that's, that's a sizable, that's a sizable
- 46:42identity graph. So, so yeah, it's been, you
- 46:46know, fascinating talking to you and, and seeing this sort of
- 46:49trajectory all the way through Samba TV and talking about some
- 46:52of the kind of hard problems, you know, that you guys are
- 46:55solving for in a pretty complex space.
- 46:57It's got a lot of, you know, and.
- 46:59Some detail on like how this whole thing works, right?
- 47:01I think a lot of people don't really understand that, so it
- 47:03was good to know. Yeah.
- 47:04I mean, you know, ACRII wanted to know exactly how that was
- 47:07implemented into the the the machines themselves.
- 47:11The hardware. And that was fascinating.
- 47:13So yeah. So tell us, I mean, as you've
- 47:15walked us through your career, I mean, what do you think is the
- 47:18hardest part about building a career in AD tech today?
- 47:24I think keeping up with the Change the Law study is the
- 47:28technology, you know, just going way too fast and consumer
- 47:32behaviour, specifically pushing that change.
- 47:34So if you're a learner and if you perseverance is your virture
- 47:39and you really, you know, offer a challenge, I think you should
- 47:42have a breeze. But if you are not, then that's
- 47:45that's. That's the sentence as far as
- 47:47I'm concerned. Yeah, yeah.
- 47:48The, the days of hiding in a cubicle and pretending that
- 47:50you're working for a large company are kind of those are
- 47:53over, but you sort of have to. I think your point earlier about
- 47:55being a builder is pretty important because you just
- 47:59don't, you can't just be safe in a in a structured sort of
- 48:03environment and system like you could in the past, because it's
- 48:06so rapidly being disrupted and. You better get control of tech.
- 48:11I mean, I think that's for sure, regardless what industry.
- 48:13I mean, I think especially in ours, it's so tech driven.
- 48:16You have what's 1 company or technology media that you
- 48:20think's measurably underrated? Maybe fly it a little bit under
- 48:22the radar that you want to call out.
- 48:27Well, never bet against Jeff Green.
- 48:29So there's a lot of heat on Jeff now, but I have deep respect
- 48:32what he has been building. So I would bet on trade that
- 48:35might be shorted for now, but that's my answer.
- 48:38Is shorted. For now with the publicist
- 48:40battles, yeah, the the, and, and I've heard that before that that
- 48:44just as a as a leader and somebody that's able to pivot
- 48:48and adapt and change and adjust yeah, that's it I think that's.
- 48:51It I don't know his battles with the hold codes.
- 48:53I I'm scratching my head over that.
- 48:54They built their entire business like really being by being easy
- 48:58to work with hold codes. When you look at the, you know,
- 49:01the, the UI they pushed out, no one liked.
- 49:03And so I don't really understand it, but I, I think you're right.
- 49:05He's, he's a good leader. So I wouldn't.
- 49:07I probably would be foolish to bet against him too.
- 49:09Yeah. And there is this whole sort of
- 49:11it's a battle of transparency, you know, with the the Holt Cos
- 49:15not having the level of transparency, but you could
- 49:17argue that that well, principal, well, principal.
- 49:20Based media, Brett. I mean it's like, I, I mean, I
- 49:22don't think anyone thinks it's great, but like when clients buy
- 49:25just based on CPM race at pubs and it's just race to the
- 49:28bottom. I mean, I think it you could
- 49:29argue from the Holt Co perspective, it's increasingly
- 49:31hard to hard to make money without it.
- 49:33So it's again like apparently some big, big, big expose next
- 49:39week about that. So I'm curious to see how that
- 49:41goes. Yeah, it's just once you start
- 49:43to dig beneath the surface of programmatic media as we've, as
- 49:47we've explored a ton in this podcast with Doctor Fu and
- 49:50things like that, you start to realize that transparency may
- 49:54not be the leg you want to stand on considering how much dirty
- 49:57bad data there is. Speaking of what we talked about
- 49:59with C. You call that the dark
- 50:01underbelly the. Dark underbelly of of the of the
- 50:04data targeting ecosystem. My father, we were just making
- 50:07fun of of sort of sort of older parents and their technology
- 50:11challenges. But he he said for years that
- 50:13I'm in the surveillance marketing business.
- 50:17He has merit. I was like, he has merit, you
- 50:19know, absolutely. And Brett, to your point, like,
- 50:22you know, the even basic transparency on the auction
- 50:25level, like the premise of programmatic, like, you know,
- 50:30little stuff there, you know, so.
- 50:33Yeah. Yeah, let's talk about rebates
- 50:35later. Let's talk about either the do
- 50:36the launch or. Supply path, all that.
- 50:38Yeah, yeah. Let's just start with the
- 50:40auction. Yeah.
- 50:42What do you think is being sort of overhyped?
- 50:44And you know, we could all say AI, but I'm not sure that's
- 50:48overhyped, right. If you had like an over under
- 50:50bet here, I would be like, I think that's actually probably
- 50:53not being sufficiently hyped, even though it is being hyped.
- 50:56What do you think of the most overhyped thing in the industry
- 50:58that you're seeing right now? Mar Tech.
- 50:59Ad Tech. It's a good one.
- 51:02You know, I wouldn't say AI, but I think AI agents doing
- 51:05everything is overhyped. I think it will happen.
- 51:09And the timeline for that, you know, based on some estimation
- 51:12is about 18 months from now. But you know, that's I guess
- 51:16high versus reality. It has a timeline.
- 51:19And also another thing is automation is better via UX,
- 51:23right? So what is true at Gen. tech
- 51:25versus what is true as AUI user experience in the UI itself?
- 51:31Because a lot of agentic workflows, automations we have
- 51:34seen, which I think is overhyped for sure is a recreated similar
- 51:38experiences, but just with agents.
- 51:41But what's the difference in going to booking.com for
- 51:43example, type in those fields versus telling the same fields
- 51:46to your agent to book that viathebooking.com?
- 51:49So I think that dichotomy is definitely overhyped.
- 51:55Yeah, that's, that's interesting.
- 51:56Like what would be your advice to someone who's just like
- 51:58getting out of B school or just starting their career who wants
- 52:01a career in, in advertising? And, you know, what would you
- 52:05like where, Like what would be your advice to like where, like
- 52:07what they should focus on, where they should start is their type
- 52:09of company, a type of technology?
- 52:11Like where would be a good place for them to dig in?
- 52:13You know, I will say if I would give myself advice on the same
- 52:16thing, where would I start my career?
- 52:18I think the answer would be the same.
- 52:20I would start it with data infrastructure and measurement.
- 52:24Someone still has to measure your homework and if you have
- 52:26independent measurement or you understand data infrastructure
- 52:30and that's called cloud infrastructure, I think you
- 52:33have. So OK, go back to your note,
- 52:36name the first word in your forecast.
- 52:38That's where I would go signal and maybe add trust to it so.
- 52:42Yeah. So, so like if you're in
- 52:44college, I mean, I've got a, I've got a son at Syracuse right
- 52:46now and he's, he's in the media space and you know, should I
- 52:50push him towards things like a common metrics, right?
- 52:53Understand the fundamentals of data and data statistics and
- 52:56data analysis and, and how this stuff is, is sort of integrated
- 53:01into decisioning, right? Which whether it's happening in
- 53:04AI or in a data warehouse or, or in a professor's spreadsheet,
- 53:11right. Do you think those types of
- 53:13areas are kind of important for understanding how data works
- 53:16and. I would say, Brett, you know,
- 53:18maybe I'm biased because this is kind of my upbringing, my
- 53:21background. My father always said you have
- 53:23to know mathematics and do applied mathematics, applied,
- 53:27meaning you can apply to anything, biochemistry, to
- 53:31epigenetics, to advertising and future marketing.
- 53:34Like it's universal language. Look what we see about
- 53:37embeddings, about future, about LLMS, how they work.
- 53:40It's all math. So if you understand it, you
- 53:45might have merit. We talked about automation part.
- 53:49Do you need to understand in the future to move away from AGI to
- 53:53as GI super intelligence and do you have merit in that or will
- 53:57it be 1% of population who needs to really control that
- 54:00narrative? I don't know.
- 54:02But I think you have a good chance and better bet if you do
- 54:05know versus not knowing. That helped me.
- 54:08You know, all my career was kind of best on that from, you know,
- 54:11doing sales, doing analytics, doing whatever, you know,
- 54:14anything in another day still ended in a foot.
- 54:16This project, you know, it was a little people don't excel.
- 54:20And if you don't understand that, you know you have a
- 54:22problem. Yeah, yeah.
- 54:25I think in order to disrupt or to deconstruct, you really and
- 54:29you really need to understand how things are built from the
- 54:32ground up and if they're built on mathematical principles or
- 54:35technical principles. A lot of the people that have
- 54:37succeeded in our industry have have have been technical people
- 54:42that then build out some of the commercial knowledge and
- 54:46experience. It's a classic CEO role.
- 54:48I was on product and now I'm in sales.
- 54:50If you can play both of those worlds and bridge those worlds,
- 54:54I think you're just you're you're advantage, right
- 54:57definition. And you've certainly done that
- 54:58in your career. It helped me tremendously.
- 55:01It helped me tremendously moving from tech product to operations
- 55:04and then just finishing now with the commercial.
- 55:06You know, when you set, when you understand things, when you can
- 55:09talk to it in simple language, you just call authentic salad.
- 55:13You don't need to read slides. You just explain how it works
- 55:17and you know you you know, you become natural authentic seller.
- 55:20And I think that was my journey. It helped me tremendously cheapo
- 55:24and read understanding technology and kind of using
- 55:27that as a spine and then dress it up nicely with with cool
- 55:31slides, which can be a jet generated.
- 55:33The your differentiation is not the presentation layer, but but
- 55:38understand it. But also, I was thinking about
- 55:40that something I would love to learn more and kind of, you
- 55:43know, practice is one thing which I think will truly
- 55:49differentiate great from good in the future is ability to
- 55:53articulate succinctly anything. And I think that that's a human
- 55:58element, right? So McKenzie style speech, when
- 56:01you come in and you really can capture the audience with your
- 56:04narrative, with your storytelling, and you can grip
- 56:07them with that attention, that skill will be universal and I
- 56:12think will be more prominent than anything else.
- 56:14So this is something I'm working on now.
- 56:16And if it's not the ad for McKenzie training?
- 56:19But no. It's funny you say that because
- 56:22that's come up a few of these other podcasts with leaders who
- 56:24said that like it's if anything, maybe more important now, right?
- 56:28Because you know, you can use AI to crank out code and to turn it
- 56:33a lot of maybe content. That's OK, right?
- 56:36But being able to understand what it means, be able to write
- 56:38good promise, be able to take what the AI gives you and then
- 56:41communicate that in a great way. And also to your point, be able
- 56:44to talk to technical people as well as business people, be able
- 56:48to bridge that gap. I mean, I think that that that's
- 56:50an incredibly important skill that not many people have.
- 56:53Is that ingredient definitely a superpower?
- 56:55Well, yeah, but it's been phenomenal talking to you.
- 56:58I wish we could go all day on this, on these topics.
- 57:03Yeah. So we always ask how can people
- 57:05reach you? Yeah, if.
- 57:08They want to get hold of you like what's the best way is
- 57:11there? Yeah, the best way.
- 57:12Social media. I think LinkedIn like I'm not
- 57:15Omni channel, you know, it's probably by like he'll is he'll
- 57:18LinkedIn is my is my poison. And you know, I'm always always
- 57:22open to ideas to, you know, to collaborate with people.
- 57:26So please do reach out, always reply so and thank you Brad and
- 57:30Rio. I was absolutely amazing talking
- 57:32to you. A very engaging conversation,
- 57:35both thought letters and. Friday when we're when we're all
- 57:38running on fumes after a busy week.
- 57:40But but everybody, thanks for making it this far in the in the
- 57:43podcast. If you've made it this far,
- 57:46signal and noise. You can find us on YouTube,
- 57:49Amazon, not Amazon. We're going to be on Amazon
- 57:51soon, though, Rio and I've talked about that, but Apple?
- 57:54Pie Yeah. YouTube, Spotify, Apple Yeah.
- 57:56Subscribe if you haven't. Yep, and subscribe.
- 57:59So they download automatically? Yep.
- 58:00Yep. And we will see you next time.
- 58:02And thanks again. Yeah.
- 58:04Thanks you for. Being a pleasure to be here.