The Bid Request Has Receipts: Eric Tilbury on Clean Supply, ID-Free Buying, and the Craft of Programmatic

Watch on Youtube

Listen on Spotify

Programmatic advertising has spent years getting smarter about audiences. But what if the bigger problem is that buyers still don’t fully understand what they’re actually buying?

Eric Tilbury, Vice President of Programmatic Ops & Solutions Engineering at Inuvo, joins Rio Longacre and Brett House for a deep dive into the machinery underneath modern programmatic advertising—and why better technology doesn’t necessarily produce better media buying.

Eric is an operator first. He has spent more than a decade inside campaign management, supply optimization, contextual decisioning, CTV, DSPs and the plumbing that determines whether a bid request ever reaches a buyer in the first place. That experience has also made him one of the more outspoken voices in AdTech on supply quality, measurement and what he sees as the industry’s bad habits.

The conversation starts with a deceptively simple idea: ID-free targeting is not necessarily the same thing as buying ID-free inventory. A contextual strategy may not require an identifier, but DSP settings around frequency caps, conversion tracking and optimization can still systematically favor impressions with IDs—or prevent ID-free opportunities from reaching the campaign at all.

From there, Eric explains why clean supply is not just a brand-safety exercise. Buyers need to know the publisher, seller and path behind an impression, especially in CTV and mobile apps. Cheap CPMs and attractive platform metrics can conceal low-quality placements, opaque reselling, bot traffic and inventory no advertiser would intentionally choose.

The discussion also gets into one of Eric’s biggest contrarian ideas: iOS and other ID-light environments may represent an overlooked buying opportunity. When advertisers optimize around what is easiest to identify and attribute, they can systematically ignore valuable consumers—and sometimes reward the inventory most capable of gaming their measurement systems.

Eric also breaks down Inuvo’s IntentKey approach, which uses content, concepts, sentiment and real-time signals to make buying decisions without building the decision around an individual user profile.

And then the conversation turns to measurement. Eric argues that platform ROAS, CPA and CTR should be treated as directional optimization metrics—not as the ultimate source of truth. If advertising is actually working, the question should be whether it created incremental business results that would not have happened otherwise.

Rio, Brett and Eric also explore:

• Why DSP auto-optimization can send campaigns toward terrible inventory

• The difference between attribution and incrementality

• Why CTV supply-path verification should be much easier than it often is

• How fraudsters benefit from the industry’s obsession with identifiers

• Why contextual intelligence becomes more valuable as open-web behavior changes

• Whether AI agents and MCP interfaces will weaken DSPs—or actually make them stronger

• Why buyer judgment still matters in an increasingly automated media ecosystem

• Eric’s skepticism toward probabilistic ID bridging

• And how a 4:30 a.m. training schedule and Brazilian jiu-jitsu fit the mindset of an AdTech operator

  
  

This is AdTech school for anyone who wants to understand what actually happens between an advertiser’s strategy and the impression that eventually gets bought.

If you work in programmatic, media, CTV, contextual advertising, measurement or AdTech, this one goes deep.

Subscribe to Signal & Noise for more conversations with the people building—and challenging—the future of advertising, data, media and AI.

#AdTech #ProgrammaticAdvertising #Programmatic #CTV #DigitalAdvertising #MediaBuying #ContextualAdvertising #Advertising #Marketing #MarTech #AI #Measurement #Identity #SignalAndNoise

---

Brett (00:01.198)

Hey everybody, welcome back to Signal and Noise. This is Brett House, joined by my co-host Rhea Longaker. And today's guest is Eric Tilbury. Or Tilbury Tilbury, I gotta write, I always have to do that. We're excited about this one. Yeah, and we had a good 10-minute intro, we which we should record next time because I think this is gonna be a great conversation. for those that don't know Eric, he's the vice president of programmatic ops and solutions engineering.

Rio Longacre (00:06.983)

Everybody.

Eric Tilbury (00:12.444)

Tilbury.

Rio Longacre (00:14.28)

We've been excited about this one, Brett. This is one that's been long overdue.

Brett (00:30.388)

at Innovo. he's spent a decade or more in the programmatic execution space, spanning campaign management, supply path, auditing, contextual decisioning and solutions engineering for display, video and C T V media. So you've been at Innovo for since twenty twenty twenty seventeen, right? So

Eric Tilbury (00:49.248)

So actually I have been there so Anuvo purchased a company called NetSear and I came with that purchase and I have been there since 2012.

Brett (00:56.417)

okay.

Wow, in in an ad tech, that's an eternity. I mean th oftentimes companies aren't allow around that long. And you guys are now publicly traded. certainly's been some a bumpy ride. but you know the ad tech, you know, you you you guys seem to be shifting away from what was really a legacy search arbitrage business.

Eric Tilbury (01:02.569)

It is a lifetime.

Rio Longacre (01:18.302)

Yeah, one ad tech year is worth at least two or three human years, I think.

Eric Tilbury (01:21.642)

Yeah. Yeah.

Brett (01:22.378)

Right. So so I can see why you make a pivot away from search arbitrage, because there's been a seventy percent drop in clicks from Google alone over in twenty twenty six, you know, due to the LLM infiltration. and you guys have really moved towards this product called Intent Key. I don't know if it's a product solution or or I think you probably just call it a product, right? and that's a patented AI system that targets audiences by reading the meaning and intent of content itself. So think of it as like

I guess a contextual type of solution on steroids in a way, right? So that you don't rely on cookies, device IDs, identity graphs and other ID specific linkages to to be able to target people. Right? Is that a is that a safe way of describing

Eric Tilbury (02:05.684)

Yeah. Yeah, I mean so I I like to describe it as very much so contextual two dot right? So it's it's using like all of ad tech really uses all IDs to to, you know, model, make decisions and all of that good stuff. We let content drive that decision, right? And content consumption patterns, trends, like what really is important to an audience at any given time. that's how we go and make decisions in real time. Completely.

Brett (02:29.666)

Yeah.

Yeah, and it's an it's it's sort of an audience intelligence in and technol and targeted technology that that sits on top of the the current ecosystem, the DSP specifically for ad serving purposes. I don't know your interplay with SSPs, but that that's where you sit versus actually being an ad serving player yourself, right? And

Eric Tilbury (02:44.49)

Yep.

Eric Tilbury (02:48.02)

Yeah. Yeah, we could sit w we sit in the DSP, you know, DSPs that we're integrated with, and then we could also we also sit on the SSP. It just depends on, you know, what the clients can and can't do at any given time.

Brett (03:02.656)

Yep. Yep. And so for those that also haven't seen Eric on X and LinkedIn and he s he mentioned before that he's moved a little bit more towards LinkedIn over the last couple of years, due to maybe some of the pollution that exists in the e in the X world, right? And you and you've been on a bunch a bunch of pods and industry events and and you're really known and I wanna you know, we're gonna certainly deconstruct this in the episode for really saying that most sort of claims of ID free targeting

Eric Tilbury (03:14.771)

Okay.

Rio Longacre (03:17.511)

Pollution.

Brett (03:30.242)

You know, which is la you know, the gum gums of the world, the contextual targeting pla players of world, isn't really ID free at all. Right. And I think that's an interesting point. And tell me if I'm wrong about that. Like a DSP buyer can select a contextual or concept based targeting method, and then when they route it to a DSP, there's an ID attached impression associated with it.

Rio Longacre (03:49.895)

Yeah. Yeah, he's just got like like he's talked a lot about the limitations of I ID based buying, right, Eric? I mean I think that's kind of sum sum it up.

Eric Tilbury (03:54.177)

Yep. Yeah. Yeah, so think of it like this. Every single like you could have a decision that, yes, the decision itself is an ID free decision, right? It's using some kind of content, it's using some kind of identifier that doesn't need that user ID, right? But when you're going and working in DSPs, right? D DSPs are just decision trees, right? Like w I'm simplifying it. Like there's a very simple way to do

Brett (03:55.565)

Yeah.

Brett (04:20.195)

Yep.

Eric Tilbury (04:23.53)

to define it, but is a decision tree, right? So if you have any type of setting at the DSP level, like frequency capping, conversion tracking, exclusion targeting, whatever, right? That requires any form of a user ID along that decision, your ID free strategy that you're that you're going and executing is actually an ID'd strategy, right? Because the DSP is forcing an ID when it comes time to bid, right?

So yes, you are using it.

Rio Longacre (04:54.543)

They prefer bids with IDs for a lot of those reasons. All those things you mentioned, it's it's hard if not impossible to do them without an ID attached.

Eric Tilbury (04:57.565)

Exactly.

Brett (04:58.614)

Yeah.

Eric Tilbury (05:01.044)

Yeah, so so if you have some kind of flag at the DSP level that you can say, Hey, I want to also listen to ID free bid requests, right? And I want to make decisions on them, like that's how you would execute it. But I would get I guarantee you like ninety-nine percent of buyers wouldn't know that, right? Like your average buyer wouldn't know.

Brett (05:01.315)

Yeah.

Rio Longacre (05:24.211)

Yeah, in fact, Eric, that's the type of those are the type of conversations we really wanted to have in this pod. And why, like I was really looking forward to this one, Brett, because Eric, he's gets in the weeds, he's a true master of his craft, and he gets out there and he shares this publicly, which I think is super cool. And you mentioned X before, Eric. I think we did meet on X, right? I mean, I think we started like interacting with each other's content. I think that's correct, right? Yeah.

Brett (05:24.386)

Yeah, so s it's a promise.

Eric Tilbury (05:48.107)

Yeah, yeah. It it I mean, literally the the birth of my social personality was on X. before X's algorithm went rogue for a while, right? Like X was an awesome place where a lot of ad tech went and met and and socialized, right? Like that that was the place. It was very fast, you could get ideas out fast, you could get interactions out fast. yeah, a immediately it was awesome.

Rio Longacre (05:54.727)

That was wild.

Brett (05:58.595)

Yeah.

Rio Longacre (06:11.697)

Get feedback right away. Yep.

Brett (06:13.388)

Yep.

Eric Tilbury (06:16.522)

And that's yeah, like that, like that personality that I built online, like that online facing personality was just my personality. And on X, you could be more yourself, right? So it was a perfect place for me to be myself online and talk about, you know, my profession.

Rio Longacre (06:33.831)

Yeah. Yeah. So we so we met I met a lot of other cool people like Joe Zappa is another one that I'm I've I'm I'm pretty friendly with, met him initially there, and we meet and meet and at conferences and stuff too. And so I th I knew this would be a fun conversation. And then aga again, the the thinking here was you're an operator, you're in here, you're in the weeds, you know these things, you've been very vocal about some things, and you have some pretty strong opinions, which we do like. Brad, fair to say we like strong opinions here at Single Noise. So we wanted to talk about a bunch of topics. I think some of the main ones were clean supply, ID free buying.

Brett (06:34.497)

Yeah.

Eric Tilbury (06:40.267)

Yeah.

Rio Longacre (07:03.773)

You've talked a lot about the iOS opportunity, how that's kind of an o overlooked, right, by a lot of people. And then your path into public commentary via X and now Twitter and and and podcast and different things. So with we thought that'd be fun. And also maybe touch on a little bit. I mean, I love the fact you you're I think you're at four thirty every morning, you're out you're up working out every day, right? you know, I'm I'm a r I run every day at five in the morning, so I can definitely appreciate that. So you you gotta, yeah. It's the only time to do it. So so maybe let's start out like like

Brett (07:26.733)

Before the sun rises. Wow, don't break an ankle.

Eric Tilbury (07:29.012)

Yep. Yep.

Rio Longacre (07:34.329)

You've been very vocal saying like a lot of people in programmatic suck a programmatic. Is that an accurate statement?

Eric Tilbury (07:38.57)

Yeah. Yeah, it's very accurate. It is incredibly accurate. It's too accurate.

Brett (07:39.135)

Ha ha.

Brett (07:44.406)

And what do you mean by that?

Eric Tilbury (07:46.593)

So how much time you got? Okay. So so think of like the entire ecosystem. Yeah, so think of the entire ecosystem, right? Like the reason why f like fraud and inventory, like bad inventory just runs so rampant. Like even like a real human could be on a piece of inventory that's just terrible. They don't want to be on it, right? And it's just filled with ads.

Brett (07:49.293)

Deconstruct that statement.

Rio Longacre (07:49.873)

Ha ha ha ha.

We we have a long form podcast. We got plenty of time, Eric.

Eric Tilbury (08:12.382)

And there, you know, there's there's video ads that are playing, no one's seeing it's in the corner, right? Like the industry rewards that type of experience because of the way that buyers go and think about how they're buying programmatic, right? Like they're going, okay, at the line item level, I'm setting some set of some sort of viewability, right? I have I have all of my different like you know, video completion rate, like they have these different optimiz optimizations set.

And they don't really check anything after the fact, right? Like as long as they have those those metrics that everyone says vanity metrics, right? And they're very much so a vanity metric, like viewability, CTR, CPA, did I say video completion rate, right? Like they are they're going after those metrics, but they're not checking like what am I actually buying? How is this decision being made? Like all they're doing is saying, Am I hitting this benchmark?

Am I reporting it to the client? Don't poke me. I'm doing my job. I have something to present on a weekly QBR, right? And that is actually incredibly wrong to do like doing that is not doing programmatic correctly, right? Like doing programmatic correctly is first saying, like, okay, this is the supply that I want to bid into, right? And I know that this supply is good.

And I you know, I got it from this source or I got it straight from a company.

Rio Longacre (09:42.215)

So it does start with the publisher or it does start with the audience, or both?

Eric Tilbury (09:45.643)

No, so the it really depends. You like in theory, it starts with the publisher. Like what, you know, what what publishers am I gonna buy from, right? And what is plugged into that publisher, right? But you can start with like a curated type of audience, right, that has some kind of audience solution built into it, but 100% you should ask that curator, like, what

publishers are you connected to? Like what am I gonna buy from this? Right? And if they say no I can't give that to you, they're F my They're a liar. Yeah, they

Brett (10:17.644)

Yeah.

That's a question w and it's the same thing with with like a P P buy, right? It's if you don't know the s the actual publishers within that environment and the quality of those, the chances that you're gonna be advertising against an MFA audience, against an IVT audience that may not exist, fake C T V inventory, all that stuff rises exponentially, right?

Eric Tilbury (10:28.372)

Yeah.

Eric Tilbury (10:33.472)

Yeah.

Rio Longacre (10:38.995)

Well, Brett, remember like we had a guest in this pod who s who compared curated deals to subprime mortgage backed securities, right? And you know, the you know the thinking is that you there's there may be some good inventory near, but it's a lot of shit inventory mixed in or shit or shit sandwiches. That's the other expression I heard.

Brett (10:48.854)

Yeah.

Brett (10:55.627)

Yeah. Yeah, and you don't know exactly what you're signing the dotted line for, right? And you realize, shit, I didn't realize my payments were gonna go up five X on month six. You know.

Eric Tilbury (10:57.056)

Yeah.

Eric Tilbury (11:00.49)

Yeah.

Eric Tilbury (11:04.99)

Yeah. I mean, and that's the thing is ad tech takes things that are cool and good and just muddies them up. You know I mean? Like curation in theory is an awesome thing because it solves a lot of things for the average buyer that doesn't know how to do programmatic well. So if you you have a good curator that can, you know, package something up that's very clean.

And good for you, or like, you know, you'd like say this PMP will give me very low density when I use this this curator, right? On good human inventory. I don't have to go and figure that out on the DSP. I I could just do that all in a PMP, right? Just buy against that PMP at the line item. That is awesome. That is a great way to do it. Where you get a ton of the just like ad tech muddying up this type of solution, right? Is

Someone getting some curation seat somewhere, right? They just do a run of an SSP, right? They just add a fee on top of what you would have gotten anyways, right? And now you're paying more for what you you could have just literally just set up yourself in the DSP. And that is incredibly common. It's too common, right? And that's why you see me on social media screaming.

about this stuff until I'm right in the face because I'm like, you guys are trying to ruin good things, right? Like there's a lot of good things in this industry, but a lot of people cut corners and you know, like I get it. People are trying to make money, they're trying to survive, but there's there's the right way to do things and there's the wrong way to do things. And someone that really knows the craft, they can they can say this is wrong. Don't do it this way. But there's not enough education and that's what I'm out there to try and

Brett (12:53.377)

Well and do it.

Eric Tilbury (12:53.418)

Spread is it?

Rio Longacre (12:53.489)

How much of programmatic is just sophisticated tech compensating for the fact that your average ad buyer or person who thinks they know programmatic has no idea where the ads are running?

Eric Tilbury (13:02.174)

Yeah, to just exploit it, probably the majority of it. Right? Like

Brett (13:05.931)

Yeah, yeah, and and a yeah, and how and how much do you think it's an incentive misalignment? We talk about this a lot, is with with brands saying, Hey, we want we've got these you're not gonna turn away budget, right? The agency's not gonna turn away budget, and we've got these these reach numbers, right? just for br for the the most reach we can get across these channels, find it, whether it's real inventory or not. I mean, how much do you think that that's a problem?

Eric Tilbury (13:18.39)

Mm-hmm.

Rio Longacre (13:30.099)

Yeah, that's where we get MFA, yeah.

Eric Tilbury (13:31.989)

Yeah, that's where that's where it gets crazy. There's not enough people having the real conversations of like, hey, you know, this this what you're trying to target, it's not an un unlimited list, right? When I go and target real inventory that humans wanna be on, there is a there's a point where I can't deliver anymore, right? obviously solutions just want you to keep spending so they bend those rules, right? And they they go and do these audience extensions and

Brett (13:52.075)

Yeah.

Eric Tilbury (14:01.92)

probabilistic ID bridging, right? Like

Rio Longacre (14:04.262)

Foo Foo was going on about that all the time, right? We had him on a couple of times. Like he he bashed that.

Brett (14:06.743)

yeah. Yeah, the second you see audience extensions, yeah. Like if you're within a walled garden. Yeah, if you're within a wall if you're buying direct p like publisher direct or walled garden direct and you go for an audience extension, that's right when you're gonna go into kind of a polluted universe of just f bots and fake inventory.

Eric Tilbury (14:08.021)

Yeah.

Rio Longacre (14:10.768)

He's like, uncheck that box right away, yeah.

Eric Tilbury (14:10.856)

Yeah. Yeah.

Eric Tilbury (14:22.172)

Exactly. Exactly. I mean and I mean to to piggyback on Foo, like Foo having something like Foo in place is a good thing for buyers, right? Because they can say, okay, I I tried out your audio ex extension, right? And I'm running post-bit analytics against it and none of it was human. So I'm gonna kill that immediately. And Foo like Foo you will be he gives you a dashboard. You can see that immediately of like, hey, this is all bots.

Brett (14:50.337)

Yeah. Yeah, based on based on the actual log files, right? Yeah, and he's he's he's pointing to it. And what and what I find so and we what's so disconcerting is how quickly the C T V media ecosystem, which is getting all of the rage and the upfronts and the new fronts, right, has been polluted, you know, with with a just a bunch of of kind of bot driven fake inventory that that isn't even appearing.

Eric Tilbury (14:53.366)

Exactly.

Rio Longacre (15:13.159)

f yeah, fake inventory. It's not even it's not even really C T V, right?

Eric Tilbury (15:14.602)

Yep. So let me let me give you guys a rule of thumb for CTV, right? There is infographics everywhere of the highest like the highest subscribed apps in the T V space. Right? If you look on that list and you see what's available outside of the walled gardens, right? Like Amazon Prime if you don't have a if you don't have a Amazon DSP seat, that's fine.

Brett (15:16.651)

Yeah.

Brett (15:30.337)

Yep. Yep.

Eric Tilbury (15:43.38)

But if you look at those lists that are available in any DSP and you're not buying those at your highest volume, something is broken. Something isn't something's not checking out correctly, right? Like what

Rio Longacre (15:55.165)

How many people are on that list? It's it's not a huge list, right? It's just we're talking just a couple of dozen, right? At most.

Eric Tilbury (15:59.039)

CTB is the easiest channel to get SPO correct.

Brett (16:02.1)

Yeah, because you 'cause you like you're saying it's the Netflixes, it's the Hulus, it's the Amazon Primes, it's the Key, it's the Paramount's, it's the Disneys. Yeah. The

Eric Tilbury (16:07.882)

The the yeah, the the Paramounts, the Peacocks, the Disneys, right? Even even like your spectrums and direct TVs, right? They should be on that list. Like why would you not be buying like the if I

Brett (16:19.052)

Mm.

And and it and if they what about yeah, so so Foos has a really great quote. He says there's no fraud in C T V because it's impossible. But that statement exists in the same universe as a hundred percent fraud in C T V. Because if you go outside of those top ten, you're getting generally a hundred percent fraud, at least his maybe it's not quite that much, but that's his claim versus going direct to those top ten, which is still being considered open internet, which is interesting, or open C T V content, right? Open web.

Eric Tilbury (16:38.464)

Yeah.

Eric Tilbury (16:43.999)

Yeah.

Yeah. Yeah.

Brett (16:49.794)

Right? E because it's appearing on these on these auction DSPs, right? For auction on these DSPs. So you're saying that the the the cleanliness of the inventory should be premium publisher subscription based, even if they're ad supported. and once once you now and there there's a problem with spoofing, right? Where people spoof a domain as ESPN or Paramount, right, to try to get into that bid stream so they're grabbing

Rio Longacre (17:13.402)

Or you said Brett, remember they'd have the JavaScript that would run that it would the video wouldn't even play, but it would like it would just be it would it would be like a fake f fake autoplays, right? Yeah.

Eric Tilbury (17:20.232)

It I mean no one no one's looking at their sellers.json, obviously, right? 'Cause you can see like who actually sold you the impression, right? Like that's Yeah, yeah.

Brett (17:21.184)

Triggering an impression, right?

Brett (17:28.96)

Yeah, it's like it's like SESPN dot com, right? You know, something something that stands out. And he and and Fu kinda says this is super obvious, and advertisers have to be certain you know, incredibly aware of this fact, right? Because they're chasing reach, but there isn't that much reach because there's only a limited number of actual premium providers that all the inventory that you should care about is is on.

Eric Tilbury (17:51.413)

Yeah, I mean there there is some fast apps that are that are completely fine, like you know, like the tubies of the world, right? Like those those are fine. They to be out of the Super Bowl. Right? So I mean you just need to really pay attention to what you're buying on CTV, who is actually selling it to you, right? Like sellers.json is isn't even available in most DSPs in just the UI to report on, right? That just blows my mind. Like

Brett (17:55.947)

Yep. Yep.

Eric Tilbury (18:18.366)

Why wouldn't that be they just have like map seller, which like the DSP map the seller, like come on, right? Like that's fine. But sellers dot JSON will tell you like, okay, who actually who actually sold you that impression reported, you know, from there. so if you're referencing that and you see something odd, like you need to kill that source immediately, right? And there's not enough Well, go ahead.

Rio Longacre (18:42.844)

But Foo but but Food, Eric, Foo did say that we should just buy for C T V. He was just saying go buy direct from a platform sellers or even buy through a DSP, right? He's even saying that because like there's so few places that actually are make up almost a hundred percent of all quality inventory. Why even go through a DSP? what would be your counter argument to that?

Brett (18:48.971)

Yep.

Eric Tilbury (19:01.53)

I mean so being able to being able to put my own audience on it, right? Like I don't want to just do a you know, a trust not just

I don't wanna use the the publisher's audience, right? I wanna use my audience, right? And I want to be able to activate across a bunch of different well, you know, the handful of publishers, right? I don't wanna just buy one publisher on C T V, right? So I want to have my buying platform where I could access all these different publishers. I wanna be able to put my audiences on top of it, test my own audiences on top of it. I mean, I can get that same inventory that they're going to buy if they just push me a P P from their per like

Like their preferred SSP, right? So like I would rather just do it myself.

Brett (19:46.038)

So there's th is there is there is so is there

Rio Longacre (19:47.73)

So you can get the same inventory, but you can attach your audience to it. You can do frequency caps across platform. You can do lot of things you couldn't do if you yeah.

Eric Tilbury (19:52.543)

And do whatever yeah, I have I I want the control, right? I don't want them to control it. I want to control it and I get it. Like programmatic guarantee deals and all that, like they're they're they're good, right? If you're trying to isolate some kind of inventory that you want to get in front of, right? Like that's fine. But I also want to go and control it and have my own audiences that I put on top of it. It's my decision.

Brett (19:55.936)

Yeah.

Brett (20:18.496)

Yeah, well, so you're saying that they you could do that technically with let's say at Netflix, you're buying directly from Hulu or Netflix, right? If they pushed their SSP back to you, you could technically layer in your audience. fr let's say you're a media banner on the a media buyer on the brand side or you're a media buyer within their agency group. I is that possible or does that only a capability that's only available via the DSPs?

Eric Tilbury (20:27.446)

Mm-hmm.

Eric Tilbury (20:30.964)

Eric Tilbury (20:41.014)

So yeah, Disney, Disney and like, you know, Peacock and all of them, they will let you put your so they'll push you the they'll put you Yeah, I mean, and you could you could you could apply it right in the in the DSP, right? They'll let you do that. Netflix, they they make you buy more well, unless it's changed recently, but they make you buy more of like a PG type style, right? Where they will say, hey

Rio Longacre (20:48.625)

Yeah, they have a clean room and you can upload an audience to them through that, right? But like but but they're more sophisticated.

Brett (20:52.075)

Yeah.

Eric Tilbury (21:10.75)

you can package up our Netflix audiences on this type of content, right? And it's this CPM, right? So you're buying biddable inventory, right? But it's controlled with their audience on top of it. Right. So you could Yeah. Exactly. Exactly. Right. And if I'm trying to if I'm trying to do some kind of omnichannel solution, right, where I'm targeting the same type of

Brett (21:13.504)

Yeah.

Brett (21:23.486)

Yeah, with their own segmentation schema, their own kind of audience profiles, all that sort of stuff, which may not match what you're looking to target from a you know

Eric Tilbury (21:39.678)

audience across multiple different channels and tactics, then I can't do that when they go and control it that way.

Brett (21:45.569)

Th this is the counter argument, Rio, to to the foo argument. That like why don't you just go buy direct? Well, yeah, there's all these other considerations and frequency capping being a huge one, the ability to control the and and you know, hi pull those levers that say, Hey, we don't want people to be exposed this much or we want to turn off this particular audience group or that particular audience. If it's frag if you've got ten different buys across ten different premium CTV providers, I don't see how that's possible. Right?

Rio Longacre (21:48.006)

Yeah.

Rio Longacre (21:54.225)

A huge one, yeah.

Eric Tilbury (22:09.61)

Yeah, yeah, I mean like a lot of it yeah, without a D S P. No, it's the I mean

Brett (22:12.906)

Wi without a DSP or some sort of

Rio Longacre (22:17.135)

Also controlling your budget and pacing, you can put you do that all through the DSP across all the places. So it aggregates all of that. So yeah, there's there's a lot of benefits to doing a DSP. I'm not saying you shouldn't, but that was Foo's argument, which I thought was interesting. and a little controversial, right? 'Cause he was basically saying just don't do it, go direct, do all your buys. Maybe even even AI can help you do that. but but you're saying there's still a lot of benefits to having to running through DSP.

Eric Tilbury (22:21.12)

Yeah.

Eric Tilbury (22:28.138)

Yeah.

Eric Tilbury (22:34.922)

Yeah. I mean, e even in that breath, like if you have an MCP connector from the DSP in your LLM, right? Like in your LLM workspace, you could still do that same thing that I was saying, right? Right in the LLM, right? that's why MCP connectors are so cool, right? They're they're incredibly cool because you can I mean, from from one I

Brett (22:56.513)

Yeah.

Rio Longacre (22:59.985)

Yeah, Viant's going totally headless. Everything they're going everything with MCP. I thought that was kinda cool.

Eric Tilbury (23:04.072)

yeah, I mean it's gonna get to the point it's not there right now. It's gonna get to the point where you don't you don't even need the the UI and the DSP, right? Like all you need is an MCP connector and then you can do whatever project you need inside of the inside of the LLM that will give you everything that you needed right up right right from the DSP, right? But I DSPs in my opinion, this is just going to strengthen them rather than hurt them.

Brett (23:33.732)

interesting. Because of like the rules yeah. Yeah, I tell yeah totally. The rule you're saying the rules logic and all these rules that we sort of named, I it's bidding logic, it's frequency capping logic, it's all this stuff. That stuff is defensible? Budget pacing, that stuff is defensible by the DSPs you're saying? You know, in terms of go forward?

Rio Longacre (23:34.641)

Well, 'cause we've heard the opposite argument a bunch of times.

Eric Tilbury (23:36.821)

Yeah.

Rio Longacre (23:46.373)

Budget pacing, all that, yeah.

Eric Tilbury (23:51.071)

Yeah, I mean think of like if you are able to do everything through one connect like the DSP solves a lot of being able like jumping through all these different hoops and all these different platforms, right? and realistically like DSPs being it like it all comes down to all of the different inventory sources and signals that are available to you, right? So like

Yeah, a lot of people are consolidating into different like SSPs and all of that good stuff, but you still want to have more than just one, right? Like I don't know. in my opinion, I think DSPs will be stronger because of this, because just because of that

Brett (24:34.475)

Yeah and

Rio Longacre (24:41.349)

Well yeah, but it's interesting because er Eric, the counter argument we've heard from a bunch of people is that a lot because of the DSPs can only ingest so much signal right from from supply, a lot of it's just been stripped is stripped out before they that goes into that decision tree you mentioned, right? And that's why ID's been prioritized because that's one of the only things they can prioritize. A lot of a lot of the contextual signals and other even first party signals from publishers, right? Don't make it in there. And that's why you've seen Chalice, with what they're doing on on SSPs and yeah

Brett (25:10.273)

Permitive.

Rio Longacre (25:11.281)

permitive and a lot of other people we've been talking to is okay well like a lot of there's been lots of there's been a lot of innovation I think on the supply side recently and then the the the argument go then goes as as a lot of this gets agenified and further decisioning potentially runs goes away from DSPs, do what does a DSP then do? But your argument is it it actually does a lot, even if even given that.

Eric Tilbury (25:34.869)

Yeah, I mean the theory that everyone is talking about, like you have to get a lot of so, I mean that's a like that's a rabbit hole in itself. Right? When you start agentifying everything, right, like what w w what are we actually talking about, right? Because when you're saying agentify, like for people are talking about for better outcomes, right? For it for example.

Brett (25:47.658)

Ha ha

Eric Tilbury (26:04.66)

Well what are people defining as outcomes? Right? Like the and I've said it in a post and I'll say it again here. The definition of an outcome is to make more money by serving ads than you would have without serving ads. That is an outcome. That's

Brett (26:07.167)

Yeah.

Brett (26:20.851)

Yeah, w yeah, that that and that's an incremental outcome. Right? That's saying, Hey, we wouldn't have made this otherwise and so we need to do advertising to drive some sort of incremental benefit for the brand.

Rio Longacre (26:23.129)

Incrementality.

Eric Tilbury (26:24.831)

Yes.

Eric Tilbury (26:29.822)

Yeah, so what people are gonna do with I I can already see it, because I have a crystal ball, I can see into the future, just the way the ad tech is gone. With agentic workflows, people are gonna go in, these things are like the Death Star, right? They point wherever you shoot. they're gonna go in and they're gonna say, Okay, I want to I wanna get rid of, you know, whatever DS Yeah, I want I want this CTR, go get me this CTR.

Rio Longacre (26:52.113)

I wanna destroy Alderan, right? You're gonna

Eric Tilbury (26:58.868)

Right? And they'll be like, I'll get you that C T R right. And then the the bad habits of programmatic kind of just keep changing, right? So like realistically, like with all these different workflows of what what is going to change, buyer mindsets need to change first before real outcomes and

Brett (27:01.61)

Yeah.

Brett (27:20.937)

Yeah, before before you expect these agencies agents to do anything correctly, like you've gotta change sort of what's broken behind the scenes, which is bi by is is is expectations and and strategic planning when it goes into this process.

Eric Tilbury (27:27.892)

Yeah, I mean could

Eric Tilbury (27:35.295)

Yeah, I mean the the the cold hard truth is that the when you start optimizing correctly and optimizing towards real incrementality, like those metrics that people are so married to don't look as good. Right? Because a CPA for a line item that's trying really hard to bring incrementality is not going to have the same CPA as a line item that's just sitting there for layups for users that are gonna convert anyways, right?

Brett (28:02.441)

Yeah. Yeah, yeah.

Eric Tilbury (28:05.172)

And the average buyer doesn't understand that, right? And the CTR is lower, but it's doing a lot of hard work for incrementality. And

Brett (28:13.705)

Yeah. And th so so they're fundamentally mismanaging their expectations. They're they're putting a much lower CPA maybe than what's realistic with that logic, right? Like you might you might need a ten X CPA to actually get incremental new audience 'cause otherwise that low CPA that you're looking for is just a is just a a a a pipe dream. It's not gonna happen.

Eric Tilbury (28:21.3)

Yeah. No. I mean the and they what

Yeah.

Eric Tilbury (28:33.844)

Yeah, well you you get all these these these like custom algorithms and custom solutions or solutions in general that are like we brought this much platform repro reported ROAS, we did this low of CPA, it improved this much, but it's like okay, were you just doing a better job of buying like finding users that were gonna convert anyways? Right? Like did you

Brett (28:58.666)

Yeah.

Rio Longacre (28:59.204)

That's the argument, right? Where were these people gonna buy a convert anyway? And you're just we're just we're just attributing them to this campaign and you're getting you're getting better CPA that's and that's it. There's been a lot I mean, there's that famous case study with Uber, right, where they I think they die they they they ran a whole dot test and they realized all of their I think it was their meta ad spend was driving zero incremental downloads. It's wild. And they stopped it completely.

Eric Tilbury (29:04.277)

Yeah.

Eric Tilbury (29:11.253)

Yeah.

Brett (29:17.504)

Yep.

Eric Tilbury (29:21.204)

Yeah, I mean it and what's funny is like yeah, they they saw no negative repercussions of that, which is yeah, it's it's crazy. and the thing is, is like, yeah, you're working in an LLM workspace, right? So you could what was hard before of like generating some kind of incrementality math, some media mixed modeling math, like that was hard and it was expensive and you needed really like PhD level data scientists.

Now you have one in your pocket. Right? So that should be implemented into workflows. But even even having that conversation on a weekly basis with the client is hard because you're like, well, you know, that type of analysis needs time, right? And I need to over time I need to spend and we need to prove out, hey, how effective is this? And that's not that's not how people are comfortable with marketing because they're so

Brett (29:52.586)

Yeah.

Eric Tilbury (30:20.072)

trained to be like what's my immediate roas that Google's giving me? What's my immediate ROAS that Meta's giving me? Right? And they try to replicate that in programmatic. And it's just like not not to knock Google and Meta. They're incredibly important for brands. They they are incredibly everyone likes

Rio Longacre (30:38.416)

But is measurement completely broken, Eric? Is like I i 'cause that's that's a point I think you you get it I think it is completely broken. But and I know you've said this quite a bit. Like why is it broken and what could be done to fix it? Is it just not relying on platform metrics? Is it getting a a like an independent source of truth? Do we need to just scrub the whole damn thing? Like what would you s what would you think?

Eric Tilbury (30:41.343)

It's completely broken.

Eric Tilbury (30:53.653)

Yeah.

Platform metrics are directional level optimizations in platform. They're not the source of truth. And what people get so incorrect is they use that as the source of truth of like my ads on my retargeting line item gave me a 30 to 1 return on return on ads better, which is crazy, right? It should be hey, these when I go and run these tactics, these are the platform metrics that I should expect.

Rio Longacre (31:18.712)

It's crazy.

Eric Tilbury (31:26.59)

And this is how they affect the outcome on the other side of did we actually make more money? Right? And over time, right, you can take how you're spending in these different channels and these different tactics and have an overarching mathematical equation that says how correlated are these tactics and channels to my actual outcome, right? And you can see that, like, hey, when I buy more unbranded search.

Brett (31:32.863)

Yeah.

Eric Tilbury (31:56.575)

It adds in more incrementality and fuels the entire ecosystem what I'm trying to

Brett (32:01.363)

Would would that be reliant on some sort of ID graph so you can match at least at the at the i like or can you match it like an an audience cohort level versus at the household level or the persons level? Because when you're cr when you're trying to do this cross either walled garden or cross platform comparison, how how are you matching the engagement activity and the conversion activity across all of this stuff?

Eric Tilbury (32:05.96)

No.

Eric Tilbury (32:25.162)

Yeah. Yeah, so that's and that that goes to speak to the difference of like platform level optimizations of what's available in a platform or like what can I do? Like is there some kind of household level target? What can I what can I track right in the platform and what can I not track and what can I get away with in terms of like not having a certain signal in a platform versus having a signal? What levers do I need to pull?

Brett (32:41.481)

Yeah.

Eric Tilbury (32:53.598)

And then at the top, you're literally just using math. I'm spending in this channel. This is my spend, you know, per hour, per day in this channel. And this is how these are the you know leads, revenue, new revenue, whatever the goal is on the other side. And over time, you know, you have those different data points, and it's going and looking and saying how correlated as the spend goes up and down in these channels, right, on a daily or hourly basis.

Brett (32:56.618)

Yeah.

Brett (33:11.36)

Yep.

Brett (33:17.824)

Yeah.

Yeah.

Eric Tilbury (33:23.082)

How is that correlating on the other side, right? So

Brett (33:25.267)

Yeah, yeah, and then you have like a confidence measure of that correlation to get closer to causation, but not truly causation. And and I think this yeah, yeah, and I and I think this lens itself, and we talked to Richie Glassberg, if you haven't listened to that episode, it is a hoot. It's gonna take you two sessions. it's about two hours long. and he's like, let's be real, personalization one-to-one is a complete lie. Right? It's never been possible, and nor do we need to be targeting or measuring one-to-one.

Eric Tilbury (33:31.284)

Yeah, no, it it's it's never exact. It's never exact.

Rio Longacre (33:39.332)

That's a good one. Yeah.

Eric Tilbury (33:54.228)

I I

Brett (33:54.963)

Right. You should it really comes down to like it's like an advanced regression analysis, right? And when you start to cluster and cohort audience groups, that should give you enough understanding to say, hey, there's nothing wrong with a a a female cohort twenty-five to thirty-four, or a male cohort, you know, XYZ. Those segments can be predictive of consumer behavior without having no any knowledge whatsoever of the ID of the right.

Eric Tilbury (34:16.978)

One Yeah, one hundred percent. Like the the big lie that was sold was digital was gonna track everything one to one. That was the lie that was sold. Like that was that was that's utopia, right? Like that's not that's not Yeah. Yeah.

Brett (34:27.08)

Yeah.

Brett (34:30.666)

Yeah.

Rio Longacre (34:31.509)

One to one marketing, right? We've been preaching that for decades.

Brett (34:33.479)

Yeah, totally. It's well it's either utopia t utopia or dystopia. Like if you go down the path of like a surveillance marketing sort of path, right, where they know exactly who you are and wherever you are.

Eric Tilbury (34:38.378)

Yeah, exactly.

Eric Tilbury (34:42.43)

Yeah.

Yeah.

Brett (34:46.879)

But but i like I think the point is is it's not needed and nor is it possible to actually get any sort of effective measurement outputs trying to measure at that level and that you need to do a cohort analysis to understand, especially when you're cro you're trying to compare across platforms and there's so much overlap, right, in terms of what consumers are doing and how they're being exposed.

Eric Tilbury (34:54.612)

No.

Eric Tilbury (35:04.628)

Yeah.

Yeah, and the the average buyer has it. Like the average buyer even like the average CMO doesn't even know that, right? Like the what what happens, like why the CMO position is kind of a revolving door, right? Is they come in, they go and kill everything in terms of like, hey, this isn't performing well, let's cut it, let's change this, let's they bring Yeah. They start nitpicking Yeah.

Rio Longacre (35:28.739)

They bring in their they're bringing their new agency, right? They redo everything, they kill everything up.

Brett (35:32.659)

Yeah, yeah, yeah. And it's the same old metrics, it's like Roas, you know, CPA, it's the same stuff, right?

Eric Tilbury (35:37.448)

Exactly. Yeah, they start nitpicking the what the other agency was doing on search and social. I see it all the time. Right? They start nitpicking what they were doing on search and social. we're gonna change this, we're gonna adjust this, P Max editing, right? Like whatever, right? And they kinda just do the same thing, right? Rather than being like, Where's Well, go ahead.

Rio Longacre (35:54.224)

But how many CMOs Eric, how many CMOs actually would understand any of these things? So I guess that's the question, right? Because that there's a lot of talk about how change needs to start at the brand side, right? Until brands are s are saying we're not gonna put up with with invalid traffic, we're not gonna put up with bull bull crap like vanity metrics, we're not gonna put up with forty, fifty percent of spend being siphoned off. Like and until they demand change, nothing will happens. The argument I hear again and again, which I don't disagree with, but like

Eric Tilbury (36:14.123)

Yeah.

Rio Longacre (36:22.041)

The stuff we're talking about, how many CMOs or heads of media would even understand many of the com maybe many of them would. I'm curious to get your take, like how informed are the buyers about these things?

Eric Tilbury (36:30.026)

You know You know what's funny is I've seen this I've seen this merry go around a lot. It's not even just the CMOs, it's the CFOs that are going, Okay, I'm spending this. What is my return in these channels? And the CMOs are reporting to the CFOs. Right? So they're like, Well, I have to report something to them on my ROAS and these different channels. And that is why, you know, the like Google obviously is such a huge, huge player, but that like

Brett (36:45.758)

Yeah.

Eric Tilbury (36:58.548)

That is another reason why they're so dominant, right? Because they tell you that. They're like, hey, you spent this, you got this many clicks, it generated this much ROAS. You know, it doesn't matter if it doesn't matter if I brought it or if I was just sitting at the bottom of the funnel, right? So that is it's it's so ingrained, that's why it's such a problem. And that's why like the the entire industry is so trained on the wrong thing.

Right? And I know like

Brett (37:27.88)

Yeah. Well and and aren't the CFOs gonna call bullshit on on and one of the reasons why you probably see low tenure times for for CMOs in general less than two years, because if they're seeing this stuff not translate to the bottom line, you know, it's just like in B to B marketing where they're like, We created all these MQLs and SQLs and yet you're actually not s yeah, nothing's converting to the bottom line and the head of sales goes, What is going on here? Like you're facing you know, marketing's claiming all these great numbers.

Eric Tilbury (37:40.875)

Yeah.

Rio Longacre (37:46.295)

Well none of the leads are converting, right?

Eric Tilbury (37:49.524)

Yeah. Nothing changed.

Eric Tilbury (37:56.277)

Yeah.

Brett (37:56.325)

is is that that's kind of fundamentally the problem, right?

Rio Longacre (37:58.116)

Well well again well I think Brett, it's like if the fact that CMOs and heads of media can't even like have allowed a lot of the media to just go to like fraudulent sources. I think many CMOs, if you asked them how much are you spending on how much media is going to working media, they couldn't tell you. I think many CMOs have yes them how much are you spending on tech? What's the total cost of ownership for your Martech stack? They couldn't tell you within twenty five, thirty percent, right? I mean, I think that that is outrageous. And they wonder why.

Eric Tilbury (38:15.563)

Yeah.

Rio Longacre (38:27.299)

Why are you only at the job two and a half years? Well, if I were a CFO and I would if someone couldn't report basic things to you and had not been a a a good steward of the company's finances in this way, I mean, I think it's not surprising ten years are pretty short.

Eric Tilbury (38:29.75)

Yeah.

Eric Tilbury (38:40.286)

Yeah. I mean, if I was a CMO, right, of a like a a brand, the what I would pay major attention to is what supply sources am I buying from? Right? What what are the where am I buying from? Right? All they care about, like a lot of times is like what audience are we reaching? What audience are we serving ads to? But like it the the

Brett (38:41.139)

Yeah.

Eric Tilbury (39:07.284)

The Wizard of Oz type of scenario is like, well, don't pay attention to the supply that it's buying from. More so just care about this this addressable audience and our TAM, right? Like that's what we're going out and that's what we're going and reaching, right? When it's way more complex than that. It's so complex. There like you need to first know, okay, what supply sources am I buying? Am I buying from the New York Times? Am I buying from like Minecraft kids?

Right? Like that that distinction is like that will make or break a job, right? But that does that stone doesn't get uncovered and that's crazy to me.

Brett (39:49.246)

Yeah, no, for sure. For sure. So let's let's talk about y your company. I mean not to not to to shill innovo, but t tell us a little bit about what you're doing from a a contextual, what'd you say, two point three point and and and what that looks like going forward, man. You with Gum Gum and a lot of these other companies, the the contextual talk was all the rage. A few years ago it seems to have quieted down, right, and it was all kind of like I think a reaction to

Eric Tilbury (40:09.002)

Yeah.

Eric Tilbury (40:12.799)

Yeah.

Brett (40:17.071)

ID deprecation and data deprecation in the ecosystem. Right? They said we have to have so you know, and so contextual made a comeback from you know, what it was in the in let's say the early two thousands or mid two thousands. So where are you guys pushing in that in in that direction and and you know, it like how how fast are you running in that direction and away from the search arbitrage business that you came from?

Eric Tilbury (40:28.021)

Yeah.

Eric Tilbury (40:37.174)

so the search arbitrage that so there was actually there's two there's two sides of Anuvo, right? There was there was Bonfire and there was Intenkey. and Bonfire was the search side that had no like it they're two completely different lines of business. so the whole emphasis of the company now is just on Intenkey. and we're actually sprinting in that direction.

Brett (40:56.552)

Yeah.

Eric Tilbury (41:07.442)

In terms of like where the industry's going, they don't even realize that they're going this way, right? Because they're trying to hold on to their IDs so tightly, right? Like that all these different solutions are going and you know, creating different ways to to map a user ID. And the the the reality of what's going on on the open web, or even just on you could even say CTV.

Or just really every programmatic channel is like QPS is drying up, right? Like it's not the same that it was before. And the reason why No So the the I would say overall, yes, but the the actual intent of the user, right? Like, let me explain this. And it's something that I explained a while ago.

Brett (41:41.597)

Yeah.

Rio Longacre (41:45.357)

Has QPS volume fallen?

Brett (42:03.305)

Well and q and QPS is queries per second, right, for just the the audience.

Eric Tilbury (42:06.442)

Yeah, just like you know, just think of like inventory available.

Rio Longacre (42:08.557)

Yeah, and by in eighty percent of eighty percent of overall queries or bids are actually duplicate duplicate. I mean only very few are unique. I about fifteen percent are unique and they're just they 'cause they get they hop between different SSPs and they get they get push they get pushed on. So continue, sorry.

Eric Tilbury (42:14.891)

Yeah.

Eric Tilbury (42:21.662)

Yeah. So before what what you were able to do is like users on the web would probably they were they were casually browsing, right? Like they were going from web page to web page to web page. And think of it when you start thinking of like QPS and modeling or or just picking out IDs

You could have a like a really good idea of what that user ID is when you see it 20 different times on 20 different sessions across various different publishers, right? as you have that the behavior of the open web changes, right? You don't have that.

Brett (42:53.544)

Yep.

Rio Longacre (43:03.827)

People are browsing for your websites 'cause they're just staying in staying in ch L L chat interfaces. They're not going to the sites themselves.

Eric Tilbury (43:07.08)

Yeah. Exactly. So when they do land on the open web, they are there for a very specific reason. It's the content that they're reading, but the ID is not stable or not even present, right? So the reason why ID is so valuable, right, is because legacy ad tech can't make that decision right now, right? Because they're so

married to their CPAs, their frequency capping, right? The tracking everything, right? And that that opportunity is completely blind to the mo the majority of ad tech. And fraudsters, right? So fraud, they make much more money when a ID is present versus when an ID is not present. Right? So it's more lucrative for them to just to

Rio Longacre (43:56.868)

Y ye y y yeah, traffic with IDs for viewers, traffic with IDs monetized is significantly higher than than ID less or ID free traffic. I think it's about like thirty, forty percent higher, correct?

Eric Tilbury (44:05.878)

Exactly, exactly. So A you have a

Brett (44:09.265)

Is is that a va is that because of validation? You're able to validate that impression was served to a specific audience? Yeah. Okay.

Rio Longacre (44:12.975)

Well, 'cause D S Ps prioritize traffic with IDs.

Eric Tilbury (44:16.222)

Yeah. You a and they're just they're ch like they're literally just dro like those IDs are actually probably real user IDs, right? And they're just dropping IDs that they scraped from somewhere into the bid request. You're buying you see as valuable, right? And that's, you know, fraud's just

Brett (44:22.919)

Yeah.

Eric Tilbury (44:31.818)

Going just chugging along. In an ID free environment, they can't do that, right? So they they're trying to, you know, spoof whatever they can, but they're they're not getting the they're not getting the bids on it because ad tech doesn't even see that auction, right? Doesn't even reach the DSP, right? So you're in an environment with a user with very high intent, they're there for a specific reason, right? And you're able to make a decision where all of your competitors can.

The reason why that's not bid on now or it's not as common is because you the buyer, the average buyer can't report on that on the other side in their platform metrics. So they just completely, you know, or they just don't know what they're doing.

Rio Longacre (45:09.519)

So Eric, you just made you made a really good point. So basically, in the future, or in very near future, if we go to a a place where there's way far fewer bids, because people are just not spending as much time on these websites, and IDs are fewer because to your point, people aren't going there. So IDs become less stable. Well, you have fewer IDs overall in the bid stream, and you have like far fewer bids overall, that actually sounds like it'll be it'll be

Eric Tilbury (45:30.164)

Yeah. Yeah.

Rio Longacre (45:38.841)

Good for advertisers and publishers.

Eric Tilbury (45:41.161)

Yeah, very good. I mean, but that's that's why I always like you hear me shouting from the rooftops, ID free diff ID free decision making should be the number one thing that a program audit buyer is thinking about right now. Even in CTV, even you have CTV publishers that cannot monetize their ID free inventory because ad tech is chasing that stable IP address, right? But when you could use a combination of the geolocation

and the content that you're getting from that from that that that ct vib bid request you can still make a very good decision and i've seen a lot of times id free ctv outperforming ip or user id based ctv because getting that the getting the id correct in ctv is so hard right and if you're saying hey this this this type of content and these geolocations resonate with my audience

Brett (46:31.677)

Yeah. Yeah, yeah because

Eric Tilbury (46:38.816)

Chances are you're gonna hit the right person more than you know, your spoofed I think.

Brett (46:44.039)

Yeah, yeah, d yeah, that was my next question. Does it show it does it show up in performance? Do you have signal that's strong enough that basically makes the whole point of having IDs moot because it's not really driving better relevance, right? Better engagement, better outcomes, you know, from a from a a conversion perspective. Right. So what is the signal behind what you're doing and how is it richer than in in does it drive better outcomes than let's say an ID full solution?

Eric Tilbury (46:57.398)

Mm-hmm.

Eric Tilbury (47:11.978)

Yeah, yeah, it's it's all on the it's all on the back of our concept graph. So it's just saying you we we analyze content, we see what's trending with us. So clients come in, they create their custom models, they they describe what is important to what they're trying to target. We stay in front of what is important to that audience by saying, you know, what's trending with certain content, right? That matches this audience. And then as bid requests come in, we're literally scoring every single bid request and saying,

Brett (47:13.497)

T.

Eric Tilbury (47:40.875)

How important is this BitAQuest to this audience right now? Right? If it if it is very important, serve the ad. If it's not, pass on it. We don't need like that's that's the thing. It's like we don't need to hold all this user information or anything. Our system doesn't even know what a user is. it we don't

Brett (47:58.757)

Yeah. And so and so and you say concept graph, w wh that's an interesting positioning play. Like what do you mean by that?

Eric Tilbury (48:04.82)

Yeah, so here here's a very ver here's a very basic definition, right? Think of like your average just full URL.

Right? It and it contains content, right? When you understand at the concept level, like what's going on, you could understand very quickly what's going on on that page. So, like, for example, when we go and a lot of people think of keywords, but it's not keywords. So think of like a sentence saying the Toyota Tacoma is a lemon, right? We would understand the concept of that of like truck performance and the sentiment is negative, right? And then when it's surrounded by a bunch of other concepts.

Brett (48:31.869)

Yeah.

Eric Tilbury (48:40.97)

We group them together and we say, like, what is this page talking about? Right? Or what is this piece of content? Like, how is it relating to this audience? Right? And that's what we're going and making decisions on.

Brett (48:44.743)

Yeah.

Brett (48:49.916)

Yeah, so so there's a sentiment analysis, it's not just purely a keyword analysis, which gets you a bit a little bit closer to to and then and then I'm assuming you rank that or you you categorize that in a bunch of different dimensions.

Eric Tilbury (48:54.012)

No, no, no.

Eric Tilbury (48:59.338)

Yeah. And and that's constantly updated and refreshed. So actually I have a good example. So we have a client that was in the Wi Fi space. They're in the Wi Fi space, right? And there was a day where Verizon went completely down, right? Like Verizon Wi Fi just went way down. And then we noticed a huge spike in intent in that audience for that day and we were serving ads in front of it, right? And we're like, What happened? I don't even know what happened. And we have an

We have an MCP connector into our models and we could speak to our models. And we're like, what happened? And what we reported back was, hey, Verizon's Wi-Fi went down. The concepts that matched this model, right? And the concept grouping that matched this model went way up. So your audience size like quadrupled in one day, and you were able to stay in front of it immediately, because our models are constantly updating and refreshing in real time, right?

Brett (49:55.485)

Yeah.

Eric Tilbury (49:56.213)

were like a legacy ID solution, right, would not be able to pick that up, right? It would just

Brett (50:01.434)

Yeah, and speaking speaking of like outcomes, if your outcome is to is to conquest Verizon Wi-Fi purchasers, right? people that might consider switching, right? How how do you find how do you reach those people at the at the opportune time? I mean obviously at the end at the end of their contracts, but it's not really like Verizon, it's not like buying a car, you know, they don't have a lease. But like when something like that happens, in order for that to happen, I mean for that to be automated.

Eric Tilbury (50:10.462)

Mm-hmm.

Eric Tilbury (50:16.948)

Yeah.

Eric Tilbury (50:20.416)

Yeah.

Yeah.

Eric Tilbury (50:28.148)

Yeah. Yeah.

Brett (50:28.3)

is is really intelligent, right? Because that's when people are gonna be most likely to be like, Yeah, I'm not sure Verizon's working for me.

Eric Tilbury (50:34.804)

Yeah, and and that's an example of why content really is keen, right? Especially if you're like in a digital space, is like people are even even in different apps, like they're there for content, right? Like they're they're not you could be five different type of users throughout the day, and even though you showed signals for a specific type of product like five hours ago, you could be completely out of that space and that ad could be mute to you at that point, right?

Brett (51:02.429)

Yeah.

Eric Tilbury (51:03.24)

So if you're able to instantly be like, hey, this is relevant right now, go ahead and serve an ad, right? That's way more powerful. I like I've said things of like a lot of ad tech is like you're making today's decision based off of yesterday's data, right? And that's very much so like how ad tech kind of operates because they need

Brett (51:19.943)

Yeah.

Brett (51:24.166)

Well, and that's how that's how machine learning traditionally operated, right? It was always batch batch files that were uploaded and then predictive ag algorithms based on old data. And until you started getting real time streams of data pumping into even DSPs like Excelite did this, you're not gonna be able to make like current decisions, like like with current data and current information, which might be very different. Yeah.

Eric Tilbury (51:27.326)

Yeah.

Eric Tilbury (51:32.233)

Exactly.

Eric Tilbury (51:37.28)

Yeah.

Eric Tilbury (51:40.692)

Yeah.

Eric Tilbury (51:44.725)

Yeah.

Rio Longacre (51:45.015)

You're fighting the last war. Yes, funny. Yeah. So Eric Eric, you talked a lot about about I IOS. We need to pivot a little bit here about how it's undervalued and I think because there are fewer signals, because they Apple does strip things down quite a bit with app tracking transparency, restrictions on Safari. there are no third party cookies. IDs are I think, you know, quite rare. So thoughts on that? On like it it's under monet it's under monetized, I think you could argue. but what are marketers missing?

Eric Tilbury (51:47.23)

Yeah. I mean and

Eric Tilbury (51:51.67)

Mm-hmm.

Eric Tilbury (52:06.709)

Yeah.

Eric Tilbury (52:11.158)

One hundred percent.

They're missing that signal that that so because Safari strips up and and think of it. If you look down and you look at what phone everyone has, chances are someone has an iPhone, right? As a marketer.

Brett (52:27.184)

Yeah. Probably seventy, seventy to eighty percent, right? At least in the US, maybe ninety percent.

Eric Tilbury (52:31.124)

Yeah. And

Rio Longacre (52:32.276)

Of of affluent consumers, yes. I mean I y you I think A Apple's t is below the majority. I believe we could we'll check that in a second. Globally Android has the prevalent market share. US Apple has a very high market share, but if you look at affluent consumers, it's the vast majority.

Eric Tilbury (52:39.254)

Mm-hmm.

Eric Tilbury (52:44.149)

Yeah.

Yeah, yeah. So let's let's just take US for example, right? So if you're going and buying ads, now you're saying, okay, I'm going to not target that person because it doesn't favor the metrics or the the the technology that I use to buy ads. That is such a missed opportunity. It is incredibly missed. There's so much out there that they could go and be serving for a lot cheaper. Like when you start talking about CPMs.

IOS CPMs are way cheaper than

Rio Longacre (53:19.372)

And by the way, sixty point seven percent IOS market share to US, which is pretty impressive. It's gone even gone up the last few years, Brett. Yeah. Yeah, you're pretty close.

Brett (53:24.22)

Yeah. Hey, seventy percent that wasn't bad. That wasn't bad.

Eric Tilbury (53:24.83)

Yeah. Yeah, I mean literally just go look like you could do this simple exercise of like go look at your Google Analytics. Go look at the sessions that are cut like by browser, where your sessions are coming from, right? And then go look at your media buy. Where are you serving in terms of a browser, where are you serving your ads? Why is there a massive difference there? It's because you're you're favoring where the IDs are. So there's a

There's a massive piece of your audience that you're just that you are neglecting because you can't reach them, you can't put that information into a QBR, you can't report that to your CMO, your CFO, and people are afraid to get fired.

Brett (54:10.831)

Specific specifically mobile and sp and mobile is the what eighty percent of time spent? I mean it's a it's an overwhelming majority of time spent amongst US consumers. And I mean you have companies like Digital Turbine and these these that that say well th this app ecosystem where people spend most of their time is being radically underspent on and undervalued by brands, right? It's sort of in it's sort of independent, not independent of, but slightly independent.

Eric Tilbury (54:16.192)

Yeah, I mean it's Yeah.

Brett (54:37.275)

the of the operating system 'cause there's some other things. I mean how do you think about that just generally? What are your thoughts on the on the app ecosystem and the ability to take advantage of it as an advertiser?

Eric Tilbury (54:46.95)

app ecosystem is a funny ecosystem. It is if you can if you ensure you're buying a real good app inventory right like you know you have your you have your socials right the like the obviously you're gonna buy those but outside of those right like it gets messy it it gets it

Brett (54:49.453)

Ha ha ha ha.

Brett (55:07.431)

Yeah.

Brett (55:13.061)

Yeah. And well it's kind of like the open web, right?

Eric Tilbury (55:16.602)

It is it I you have to be so so hyper locked in to what you're buying in app, right? You cannot like it is the you you shouldn't do it on the open web, you really shouldn't do it in app or you just go and blindly buy an exchange, right? Like that should never

Brett (55:37.222)

Yeah.

Rio Longacre (55:38.08)

Especially if you're trying to like tr especially if you're trying to look at installs or some metrics like that where I mean, you know, there's so much so many things can be gamed.

Eric Tilbury (55:43.53)

Ugh.

It's it is it's crazy. Like you we we see it every week.

Brett (55:47.43)

Yeah. Yeah. And is and is it b is it like the eighty twenty rule where where, you know, twenty percent of the apps are driving eighty percent of the total traffic and then there's this long tail of just apps that are either being removed or just sit unused on an operating system?

Eric Tilbury (56:03.796)

The way that I buy apps is I hand pick the apps that I wanna buy from. And I hand Yeah, yeah. I do not Yeah, I will not go and blindly just say, Hey SSP, give me a give me a run of these apps, this app category. I'll never do

Brett (56:08.443)

So it's like a premium publisher play, right? Like what you do with C T V.

Eric Tilbury (56:22.44)

I will go and hand pick the publishers that and the publishers and apps that I'm buying from and say, Okay, I only want this, don't ever put anything outside of this in here. I don't care about the new apps anymore.

Brett (56:22.652)

Yeah.

Brett (56:33.935)

Yeah, and is and is that because

Rio Longacre (56:34.03)

Well, interesting is number of number of the number of people apps people have is fallen quite a bit. I remember Brett we're talking about this with Mick Rigby a few weeks ago where it used to it's gone from a few dozen to like people only have like ten, twelve apps they spend like ninety five percent of their time in now. So actually that makes a lot of sense, Eric. And I and you know, but and I'm sure too, if you go if you went to one of those SSPs and bought, you know, some specific audience sports enthusiasts, you know, ninety like I can't imagine like ninety five ninety five percent of your of the inventory you bought would be totally bogus.

Brett (56:40.678)

Yep.

Eric Tilbury (56:45.215)

Yeah.

Brett (57:00.645)

Yeah, whereas you should just go to ESPN, right? And go directly to because that's where er ESPN and maybe a couple of other kind of key app locations, right?

Eric Tilbury (57:01.268)

What? Yeah, no.

Rio Longacre (57:07.198)

A couple others, yeah. Go to Strava, right?

Eric Tilbury (57:08.622)

ESPN yeah, ESPN is one of the big apps that I buy from and you know, I buy it in C T V, I buy it in mobile apps, right? Like yes, like you should definitely, definitely, definitely go straight to them and then pay attention to what's

Brett (57:23.653)

Yeah, I I do I do wish, just an aside, I do wish ESPN's ad experience. I think their whole app experience kind of sucks. And I don't think it's something something with their abil I don't know why they're not making updates that are common like with places like Google YouTube, right? So like look at the YouTube model from an ad serving perspective and duplicate that on ESPN because ESPN, I'm watching a fifteen second, you know, replay of like of like the Immaculate Inning but you know, like right, you know, from from X on

Rio Longacre (57:47.48)

Yeah, but Google, they're the masters, Brett. They are the masters at ever at ad delivery.

Eric Tilbury (57:47.754)

Yeah.

Brett (57:53.599)

on the Red Sox and and and I have to watch another thirty second ad. Why am I watching a thirty second ad for a fifteen second piece of content? And every single time, so you you th they'll break down all of the we've all seen this, they'll break down all of the highlights that you want to see for whatever sport you're looking at, right?

Eric Tilbury (57:57.365)

Yeah.

Eric Tilbury (58:01.91)

Yeah.

Eric Tilbury (58:09.035)

Mm-hmm.

Rio Longacre (58:09.42)

Yeah, with a thirty second ad. It will literally be that long. It's so n annoying.

Brett (58:11.095)

Every single time. And there's like five highlights, right? And you're like, well, this is ridiculous. Why do I have to continue to watch an ad through this whole experience? So to me, it's a really bad ad experience. And I'm like, take some lessons from YouTube and upgrade your your app a little bit to all those ESP listeners. Corp. Yeah. Yeah, for sure. So so yeah, so the app ecosystem, I mean, are you saying that because

Eric Tilbury (58:24.462)

Yeah. Yeah. Yeah. Or make make make some like ESPN shorts, right?

Brett (58:37.089)

you're seeing just terrible performance once you go outside of of the premium apps.

Eric Tilbury (58:40.19)

Yeah. I mean if if you just randomly go buy the or like a run of exchange or a run of an SSP, right? Like you'll get like and they're legitimate apps, like the like the utility flashlights or the the sleep tracking apps that are just like refreshing ads. Like think of like the the the sleep track

Rio Longacre (59:01.9)

All night long, right. You never no one sees a single ladn those those those the the the night light ones, some of them are the worst offenders. They were all night long pumping ads through behind the

Eric Tilbury (59:05.771)

Yeah.

Eric Tilbury (59:09.502)

Yeah. And they're just they're just refreshing, refreshing, re and you know what's crazy is they they probably convert, right? Because your user ID is there, ad tech can find user IDs that convert. They're going, you go and make that purchase. That ad had no say in your purchase, but the DSP or the whatever buying platform they have set up, it's it's rewarding it, right?

Brett (59:29.145)

Is is gonna recognize it. Yeah. Yeah, it's it's cookie stuffing all over again.

Eric Tilbury (59:35.157)

Yeah, so math would be able to figure that out. The the the the platform tracking wouldn't, right? You would make a very wrong decision, right? Like say you had an iOS targeted line item that reported zero conversions, and then you have this line item that's targeting utility flashlights, this one has a five dollar CPA, this one has an NA CPA, you turn this guy off, the iOS one that brings in all your incrementality, you turn on the fraud, and you're just burning money.

And that's that's literally every day. That's literally every day.

Brett (01:00:05.349)

Yeah.

Yeah. Yeah, I it's it's amazed that that hasn't been cleaned up, right, or improved, right? And why why do you think that is? I mean

Eric Tilbury (01:00:17.544)

It's buyer, it's it's okay, so it it comes down to like literally what we were talking about. That's CFO, CMO. CFO wants to see what the CMO is doing. CMO reports back, hey, these different channels, I got this ROAS, I got this CPA. The buyer that's actually executing it, right? The CMO doesn't know what's going on in the weeds, right, at all. The buyer that's executing is being like, Well, this this

Buying this way helps me keep my job, right? I have something to report on back to my CMO, right? So this is what I'm gonna buy, right? Even if they know better, right? They're gonna be like, well, I don't know how to explain the iOS buy, right? Even though I know it's good, I don't know how to explain it.

Rio Longacre (01:01:02.423)

But but Eric, you probably saw that Nike case study recently, but you they act they're actually out of the the S P one hundred now because they just totally tanked it. And I remember like four or five years ago, like 'cause I I used I I used to work there, right? And I I was not doing their their media strategy, but I was working on like different I work with the email team. I did a lot of work with different different category owners there. I remember people back then were talking about how they were doubling down on Google and performance advertising. Remember everyone's talking about it there.

Eric Tilbury (01:01:11.424)

Yeah.

Rio Longacre (01:01:29.005)

I mean, I bet like all these traps you're mentioning right now, I bet they fell for every single one of them, right? Short term they're getting incredible ROAS. They were driving tons of new conversions, but they weren't. They were paying for sales they already would have made. They're advertising to a lot of bots. I mean, I think they fell for it I felt I think they fell for that trap.

Eric Tilbury (01:01:33.578)

Every single

Brett (01:01:40.804)

Yeah.

Brett (01:01:45.924)

Yeah. And was that wasn't that your argument, Rio, about P Max, right? That P Max is just an auto exposure that everybody sees and so you're you're sort of getting non incremental actions taken by consumers?

Eric Tilbury (01:01:46.141)

one hundred percent. I mean that's

Rio Longacre (01:01:55.119)

Well, yeah, well, I I was running campaigns this led like about a year and a half ago for a client. I remember like it like p P Max and then the audience extensions and I remember it was just o the actually this is the meta perfor there's the meta one, right? I remember they the audience extension flipped on automatically and the the you the meta a the meta AI went o automatically and then w instead of advertising to a specific audience that we were targeting based on IDs on meta, we knew who these people were. They had actually bought a car.

Brett (01:02:23.236)

Yeah.

Rio Longacre (01:02:24.684)

It started blasting out these ads to like ten times the amount. The CPMs actually went down quite a bit, so our budget didn't really change that much, but we were pissing it all away on not even real people, Brett. Or if they're real people they because they were they were off site, right? many of them probably were real people. If they were, they weren't people who bought the car, didn't even didn't even avail of the offer. So it was like it was a big compliance problem. And like just meta did that, just burned through most of our budget that one month. We caught it pretty quickly, but it was crazy. So

Eric Tilbury (01:02:49.866)

Yeah. Yeah. And they would have been better off just buying out of home in Times Square. Right? Like that's that like that's

Brett (01:02:51.046)

Hmm.

Brett (01:02:55.878)

Yeah yeah. Yeah, yeah, we're it's back to the future. We're gonna start buying linear television out of home. Or direct mail.

Rio Longacre (01:02:56.343)

Yeah, seriously.

Rio Longacre (01:03:01.228)

Or direct mail. They could have just bought direct mail, right? I mean they knew where these people were.

Eric Tilbury (01:03:04.256)

They could have done anything but that. They they literally could have donated it to charity and did a write up on donating it to charity. And that would have been way better. Yeah. Yeah. Right? Right? Like that's that's that's what like

Brett (01:03:13.102)

Yeah, d the PR itself would have driven better better awareness and and halo effect on their on their product sales.

Rio Longacre (01:03:13.107)

It is true.

Rio Longacre (01:03:21.768)

They could have done a bus wrap in Times Square. Yeah.

Eric Tilbury (01:03:24.436)

Yeah, exact that's why you guys see me on social media literally red like think of all my posts, I'm red in the face just screaming at people. I'm like, How do you not see this? Why am I the only one that sees this, right? It's it's crazy.

Brett (01:03:36.496)

Yeah, yeah. Yeah. Well this has been an awesome conversation.

Rio Longacre (01:03:40.974)

So I got a couple more questions. So okay, so you're you you're a self professed ad tech hipster.

Brett (01:03:46.54)

yeah.

Rio Longacre (01:03:48.16)

So I I lo I I love to just like comment on that.

Brett (01:03:52.253)

do you you vape and blow out huge clouds of vape, right? Outside of outside of like dive bars and like the sunset strip. No, you're in San Jose. You're in San Jose.

Eric Tilbury (01:03:55.358)

No no

Rio Longacre (01:03:56.547)

Your beard's not full enough, Eric for the hipster.

Eric Tilbury (01:04:01.43)

What's

What's funny is that I am actually the least hipster person alive. I I originally put ad tech hipster in my ex bio because I was hipsters do stuff that's trendy before it's trendy, right? So I was like, you know what? I do ad tech different than the rest of everyone doing ad tech. So I was like, I'm like an ad tech hipster, right?

Brett (01:04:19.365)

Yeah, yeah.

Eric Tilbury (01:04:28.31)

so I just put that in my bio and I did not think it would take off, but like people started recognizing me as that and I was like, first off, it's cool that you recognize me, right? And second off, like, you recognize me for the ad tech hipster? Like that was something that was just like an afterthought that I put in my ex-bio.

Rio Longacre (01:04:43.255)

That's funny.

Brett (01:04:44.398)

I'll take an adtec hipster over an adtec therapist any day.

Eric Tilbury (01:04:47.742)

Yeah.

Rio Longacre (01:04:49.165)

Well we have a we had a net tech therapist on here. We've had Messer on before. Yeah. Yeah, yeah. Mess here's a good dude. Yeah. But now he's competing podcasts, but we we'd welcome on again. Like we we we look at this as a friendly ecosystem. So okay, so you're also you work out a four thirty in the morning. Is that still accurate? Let last time we spoke you that

Eric Tilbury (01:04:54.467)

really? Yeah, and Meso's cool. I like Meso.

Eric Tilbury (01:05:07.166)

Yeah. Yeah, so w so I do four thirty, I do weight training and then six AM I do jujitsu. So yeah, so I I gotta stay dangerous, you know what I mean? Like I got I got two daughters, right? So you know I don't need some sixteen year old kid, you know, years from now challenging me on my front lawn. Right? So Yeah, exactly, I gotta win that fight.

Brett (01:05:15.098)

Wow.

Brett (01:05:27.695)

Yeah, yeah, well

Rio Longacre (01:05:27.981)

Well if he does you better if he does you wanna win, right? that's also what kind of jiu-jitsu is it? Brazilian jujitsu or

Brett (01:05:32.9)

Yeah, yeah. But you know

Eric Tilbury (01:05:36.372)

Yeah, yeah, Brazilian Jiu Jitsu. what's yeah, what's funny is if you if we did this podcast last week you would have seen my ear like out to here. yeah, I gotta drink like they they price it.

Rio Longacre (01:05:38.797)

Like Hoist Gracie.

Brett (01:05:45.45)

you get yeah, you get the cauliflower ear, right? my my only my only now do you do MMA because my only my only gripe with with jujitsu and and the reason why mis mi mixed martial arts sort of wins every time. Yeah, it's it well no no no is is that y you see a straight BJJ person going to MMA today with no other skills and they get murdered. They're on their back grabbing at people's ankles, right?

Rio Longacre (01:05:58.818)

Has it in shitified the MMA?

Eric Tilbury (01:06:07.476)

Yeah.

Rio Longacre (01:06:10.113)

Well now they do, yeah, yeah. But remember when like what remember remember UFC when it first started, th like those the straight the straight th those guys would come in and just smoke everyone 'cause no one knew how to deal with it, right? Once they once you get in grappling.

Brett (01:06:15.491)

Yeah, the the gr the Gracie family. Yeah, it and now you've gotta you gotta Yeah, you gotta have boxing skills, you've gotta have kickboxing or Muay Thai skills, you've gotta have wrestling skills because jujitsu doesn't outside of the judo influence, doesn't really teach takedowns. And then you've gotta have the on your back grappling, I'm gonna yeah. So do you do all of that?

Eric Tilbury (01:06:29.451)

Yeah.

Eric Tilbury (01:06:33.972)

Yeah.

Yeah. Yeah, so I I so I don't do MMA. I started out wrestling, like so wrestling shoes are hanging behind me. yeah I yeah, so I I started I started wrestling when I was like seven until probably like twenty two. and then you know stopped. You know, as you get older or whatever, life happens. And then my son started doing jujitsu.

Brett (01:06:47.436)

yeah. there we go. So you have the wrestling skills. That's yeah.

Eric Tilbury (01:07:05.928)

And then I I started taking the jujitsu and I like, I could do this. Right? So I I started doing the adult classes on jujitsu and you know it's you know, I haven't looked back so trying to be. Yeah. Yeah, I just

Brett (01:07:13.933)

Yeah, now you're a cold-blooded killer. I mean she could just put someone to sleep in like ten minutes. Right? You know that's the danger. I bet yeah, they get a a jujitsu person gets on your back if you can take someone down, get on your back and then they'll just choke you out and they'll put you to sleep in like th five minutes.

Eric Tilbury (01:07:28.352)

yeah, I mean like it's it's crazy, like you could come in really strong and I've seen it, like people have just been like TRT out, like bigger than me, a lot bigger than me, and they you it's you could just do anything you want with them because they have no no sense of what they're doing in jiu-jitsu, right?

Brett (01:07:47.482)

Yeah, yeah, yeah. Well and that that is incredibly disciplined. I mean to do four thirty four thirty to to s an hour and a half of weightlifting every day and then you do jujitsu? Wow, you must mix it up a lot because to prevent kind of str you know, overwork injuries. Wow. You're still young and young and nimble, right? And flexible Yeah

Rio Longacre (01:07:48.491)

Yeah. Yes.

Eric Tilbury (01:07:57.226)

No, do do do too.

Eric Tilbury (01:08:02.484)

Yeah. Yeah, you know, you gotta stretch, all that good stuff. But you know Father time Father Time is coming for me, you know, I compete in the masters division, so

Brett (01:08:14.361)

Yeah, r like I g yeah, that much working out I work out like five days a week, but not not for that amount of time consistently.

Rio Longacre (01:08:20.065)

Well well but Brett, like an outrage, like the amount of workout you need to do in order to just stay the same, not even get in better shape is is kind of crazy too. It's funny, I had a friend like in college who got into he was a big big brawler, big big Irish guy, right? And he ended up getting I might have told you the story, Brett. He he got into MMA and you remember you you're like you're you ever hear a guy named Irish Cop?

Brett (01:08:42.725)

I think so.

Eric Tilbury (01:08:42.838)

no.

Rio Longacre (01:08:44.085)

Way back, right? He he was like an MMA fighter, like that that was his nickname and like earlier he fought Kimbo Slice in that like underground yeah, there's a famous if you YouTube it'll come up, right? Like so he fought Kimbo he beat Kimbo Slice in his under and Kimbo Slice is a maniac, right guy was a beast, right? he beat him in some underground like basement matches, like a bunch of like townies hanging out drinking beers. It's it was crazy, right? So he got really into it. and then he started working out like crazy. I I think he was probably doing steroids too, 'cause he died he died of a heart attack. It's crazy.

Eric Tilbury (01:08:51.995)

okay.

Brett (01:08:52.004)

Yeah.

Eric Tilbury (01:08:59.028)

Yeah. Yeah.

Eric Tilbury (01:09:13.184)

Yeah, all those guys. What's funny is Kimbo Slice Yeah, Kimbo Slice kinda made YouTube. Like if you think of it, like everyone was watching those Kimbo Slice videos. Like Yeah.

Rio Longacre (01:09:14.362)

so not not good. Like father time catches up.

Rio Longacre (01:09:20.855)

He kinda did, yeah. He he was he was he was intense. Yeah. So

Brett (01:09:23.959)

Yep. Didn't quite translate over to the U UFC, but again it shows you multidisciplinary interdisciplinary is the only thing that that works in in MMA, right? Like you gotta have everything.

Eric Tilbury (01:09:35.468)

yeah, I mean everyone's so high level now, you have to have everything. Like you can't before, yeah, you could be a jujitsu guy or you could be a really good wrestler and just dominate. It's not like that. Yeah.

Brett (01:09:39.599)

Yeah.

Brett (01:09:45.657)

Yeah, yeah.

Rio Longacre (01:09:46.242)

You gotta be able do it all. Yeah, that's cool. So well, Eric, this has been fun. Yeah, this is like the like this is like we thought we got a little nerdy. and like you know, it's funny so we've had some of these podcasts that are you know we we've interviewed some non technical people, and then there's there's some of them have gotten really into weeds and like we were joking last time, Brett remember like my barber, like he he he wanted to listen and like we I I gave him one episode with and it was so nerdy he was like, I didn't understand a word of that. This is probably one of those for someone who's not like in in programmatic

Eric Tilbury (01:09:53.834)

Yeah.

Brett (01:10:12.259)

Yeah. If you're a generalist, if it's someone that you know just in the everyday walk of life, we'll recommend other episodes. This is the ad tech this is the ad tech deep dive, right? Where it's like, what are you guys talking about?

Rio Longacre (01:10:17.869)

Yeah. This is this this yeah, this is ad tech school right here. So this is cool. This is fun. do you wanna maybe wanna maybe Brett may do real quick hits real fast before we before we before we go?

Eric Tilbury (01:10:17.876)

Yeah.

Brett (01:10:25.625)

This is awesome. Thanks, Eric, for joining us.

Sure. Yeah, so I'll I'll I'll start. What is the fastest way to tell a buyer does not understand its supply path?

Eric Tilbury (01:10:39.668)

Huh. they'll come they'll come asking about their CTR immediately. They'll be like, why is my CTR not, you know, point six in display ads or you know, why is my V CR this and CTV? It's like all CTV is like 98% V CR, right? So like if they start asking those questions, I'm like, you have no idea what you're talking about.

Brett (01:10:57.838)

Yeah.

Brett (01:11:02.084)

Yeah.

Rio Longacre (01:11:03.689)

Eric, if advertisers actually understood blank, half the programmatic industry would be out of business tomorrow.

Eric Tilbury (01:11:11.902)

If buyers actually understood measurement and what actually leads to an outcome, I would say half the the tech industry would be screwed.

Brett (01:11:24.045)

Yeah. What what default setting in a DSP deserves the most speculation or suspicion I should say?

Eric Tilbury (01:11:31.156)

The most suspicion? Mm.

Brett (01:11:34.607)

D di d like default settings that should be avoided at all costs.

Eric Tilbury (01:11:38.536)

I would say auto optimization to a CPA.

Brett (01:11:41.923)

Yeah. Why?

Eric Tilbury (01:11:44.841)

Or CTR, right? CPR, they could be interchangeable. because it's like again, it's like the Death Star. It will point wherever you shoot. So if you don't do any other settings on top of that, which the average buyer doesn't, like with great responsibil with great power comes great responsibility, right? So you apply this setting and you don't do any

Brett (01:11:46.33)

Yeah.

Eric Tilbury (01:12:08.37)

Any supply type of you know supply path optimization or anything on top of that, that will just buy the most god-awful inventory and you'll buy it by all bots. It will get it. It will get it. I mean, at the end of the day, DSPs want to perform, right? Like people think that they want money, right? The only way they get money is by performance, right? So they want to perform, right? So if you say that's my performance goal, they will hit that goal.

Rio Longacre (01:12:18.761)

It'll get you it'll get you that C P A. Come held or high water, yep. Yep.

Brett (01:12:18.766)

Yeah.

Rio Longacre (01:12:38.541)

All right, Eric, last one. What is the biggest piece of complete bullshit and programmatic advertising everyone in the industry knows is bullshit, but we all keep telling ourselves this is real?

Eric Tilbury (01:12:48.006)

Ooh I would say probabilistic ID bridging, right? Because that's so that's so gray, right? Like I'm sure in some instances it's okay, right? Yeah, they're just saying, Well, I I think this I think your Yeah. Yeah, yeah, I I think this ID

Brett (01:12:52.98)

That's

Yeah.

It sounds like it's just guessing.

Rio Longacre (01:13:02.189)

ID bridging, you it's complete PS, right, I mean, come on.

Brett (01:13:07.542)

I mean probabilistic was a little questioning. You had to like look into w how is this architecture, how is it built, what is the model deciding on, and then when you y you add the ID bridging piece, you're like, come on.

Eric Tilbury (01:13:13.995)

Yeah.

Yeah. Like yeah, it's again all IDs that are outside of like a a publisher's O and O is like there's a there's a cookie match that happens or an ID match that happens, right? And then you start throwing like a bunch of other just ID matching on top of that, like that sounds gross. I wanna throw up talking about that.

Brett (01:13:21.806)

Brett (01:13:37.699)

Yeah, yeah. Hey, well this was awesome. So thanks for joining us, Eric, and everybody that made it this far. Visit us at debdebdeb dot signalinose.ai. Also subscribe. that we don't ask for that enough, Rio. Subscribe to our YouTube channel, subscribe to our our our podcast, subscribe to our newsletter. Yeah, our newsletter too.

Rio Longacre (01:13:40.461)

Great.

Rio Longacre (01:13:53.101)

Our newsletter too. I slight the newsletter has it all summarized every week.

Brett (01:13:57.473)

And we are everywhere you can you can absorb media from social to all of the core platforms like Apple and Spotify. and we will see you next time. Thanks, Eric.

Rio Longacre (01:14:08.215)

Ciao.

Eric Tilbury (01:14:08.544)

Chew.