top of page

The AdTech Discount: Alex Brownstein on Margins, Multiples, Moats, and Who Actually Gets Paid

Nov 5
72 min read





Why can an AdTech company create real value, generate strong margins, build sophisticated technology—and still get punished by investors?


Alex Brownstein, Founder & CEO of AHB Advisors and former Partner at 3C Ventures, joins Brett House and Rio Longacre for a wide-ranging, highly entertaining conversation about the economics underneath the advertising technology industry.


Alex brings an unusually broad perspective. His career has spanned finance at Credit Suisse, operating experience at LiveIntent, strategy consulting at McKinsey, investing and advisory work at 3C Ventures, and now his own firm, AHB Advisors. That combination gives him a rare ability to look beyond the technology itself and ask a more fundamental question: who is actually creating value in advertising—and who is simply collecting a toll?


The conversation starts with the “AdTech discount.” Why do independent AdTech businesses often trade at lower multiples than SaaS and other technology companies? Alex explains why high margins alone don't necessarily translate into premium valuations when revenue is tied to advertising spend, isn't truly recurring, and can disappear quickly when budgets change.


From there, they dig into one of Alex's biggest themes: AdTech made a lot of money from complexity. As supply-path optimization, transparency and direct relationships remove layers from the ecosystem, what happens to the DSPs, SSPs and other intermediaries whose economics were built partly around that complexity?


The discussion also gets into the enormous gravitational pull of Google, Meta and Amazon, why the walled gardens continue to capture so much industry growth, and whether independent AdTech can maintain 10%, 15% or 20% take rates when vertically integrated competitors can price technology dramatically lower.


Then comes AI.


Will AI create better AdTech businesses by reducing the cost of planning, buying, optimization and reporting—or will it commoditize the very capabilities vendors once charged a premium for? Alex, Brett and Rio explore agentic media buying, the future of DSPs and SSPs, what constitutes a real moat in 2026, and why simply putting AI on top of an inefficient process doesn't necessarily fix the economics underneath it.


They also tackle the future of agencies: the limits of the FTE model, principal media, in-housing, technology ownership, consulting, creative strategy and Alex's memorable prediction that the future may require “more Don Drapers and fewer Harry Cranes.”


And yes, there are detours into New York vs. San Francisco, Mad Men, AI overlords, consulting economics and why some conversations with Alex are apparently better continued over a beer.


In this episode:

  • Why AdTech trades at a discount to other technology categories

  • Why high margins don't equal high-quality recurring revenue

  • Google, Meta and Amazon's growing economic power

  • DSPs vs. SSPs and who really controls the money

  • The shrinking “AdTech tax” and pressure on intermediary take rates

  • Whether AI strengthens AdTech margins or destroys differentiation

  • What a real competitive moat looks like in 2026

  • The future of agencies, FTE pricing and principal media

  • Why independent measurement and verification still matter

  • Where value—and margin—will ultimately accrue as the industry consolidates


This is one of those Signal & Noise conversations that could easily have gone another two hours. Alex is one of the smartest and most interesting economic thinkers in AdTech, and we expect to have him back periodically to unpack the financial and economic forces reshaping advertising, media and technology.


Transcript:


Brett (00:01.173)

Hey everybody, welcome back to Signal and Noise. This is Brett House, joined by my co-host Rhea Longacre, calling in from San Francisco. right, from a hotel room.


Rio Longacre (00:07.636)

Everyone?


Rio Longacre (00:11.765)

City by the bay, I am here for unbox. That's the Amazon ads conference. You're sorry?


Alex Brownstein (00:14.698)

I'm really sorry.


Brett (00:17.373)

You're really sorry, why?


Alex Brownstein (00:19.222)

I'm sorry. I'm not a San Francisco fan. I'm sorry.


Brett (00:21.937)

Yeah. I I hear you. And we'll maybe we'll dig into that, Alex.


Rio Longacre (00:23.475)

I love s I love this city. I'm I'm I'm from here originally, but I know Brett you you don't believe me, he but I I actually am a New Yorker, but I was born in the Bay Area and my folks live here, so I always will have an affection for this city. But it is the city that it's funny, not like n not everyone is a fan of, including my wife, which is why we don't live here.


Alex Brownstein (00:27.96)

Got it.


Brett (00:42.995)

Yeah, it's


Alex Brownstein (00:43.382)

It's beautiful. It's beautiful. and you're getting my own bias 'cause I am a born and raised New Yorker, so like there are no other places, right? Like what what exists? The that like


Brett (00:53.429)

It's the New Yorker a it's the New Yorker magazine cover. Yeah, yeah.


Rio Longacre (00:55.015)

That's fair. New York's the best city. I can't argue that.


Alex Brownstein (00:57.448)

Yeah. It's a it's a New Yorker cover, but it's just it's so expensive and lacks city conveniences. So yeah.


Brett (01:04.978)

Yeah. And has like a really obnoxious bro culture. Let's not forget about that. And a terrible homeless problem. Yeah, I mean there's parts of New York City like that, but not the whole city.


Alex Brownstein (01:09.354)

Yes, yes, that


Rio Longacre (01:10.591)

And a terrible homeless problem. A terrible homeless problem. It's still pretty bad, yeah.


Alex Brownstein (01:14.378)

My first time I I mean, I think New York has gotten worse since COVID, but was way better a long time ago. I think, you know, my first time ever in San Francisco, I was visiting a friend who was a student at Stanford and we went down to the city or up to the city rather for a nice dinner and we were going to a great restaurant and it was San Francisco, so it was blazer weather, and multiple people stopped me on the street and were like


You're not from here. You're wearing a blazer. And I was like, the Yeah, stop me on the street. And like meanwhile, people were walking by in tutus like no one batted an eyelid, but like the blazer was too much. You can wear your $600 hoodie, but you can't have a blazer.


Rio Longacre (01:46.111)

People stopped you.


Brett (01:54.482)

Yeah.


Rio Longacre (01:56.913)

It's it it's it's funny, you'll appreciate this. So I I know a a friend of mine used to work for one of the big four consulting firms. I know you're as a cons a former ex-consultant. well ex big firm consulting person, you'll appreciate this. So he, you know, he he he went to you know B school, you know, worked and worked for one of the big firms. Yeah, I'm former Cap Gemini too. So yeah, that's where I started my consulting career too. But I remember he so he, you know, he put on a suit and then they his first project was at Lucasfilms.


Brett (01:57.396)

Yeah.


Rio Longacre (02:23.772)

Right, which is I guess, you n in the Bay Area, not too far from here. He says he got off the plane and the client knew he was a New York guy who was wear and and he knew what he'd be wearing basically. So the client met him he was like, It's weird the client wants to meet me at the airport. The client met him at the airport, picked him up, it's and drove him straight to Gap and made him change. He's like, You cannot go to the client site looking like that. They'll know you're a consultant, I'll get in trouble, you'll get laughed at. So funny.


Alex Brownstein (02:41.633)

Ha ha ha.


Brett (02:46.887)

Yeah, wow, wow. So so before we go any deeper, we haven't even introduced our guest, Alex Brownstein. No, Alex, I love it. This is a good sign that this is going to be a freewheel, fun, and informative and interesting conversation. So for those that don't know, Alex Brownstein, founder of AHB Advisors, where he works with companies and investors across ad tech, media, martech, and commerce.


Alex Brownstein (02:47.105)

Well, so it's


I know, sorry, I I jumped in.


Rio Longacre (02:52.35)

That's true.


Rio Longacre (02:57.086)

Good start.


Brett (03:12.041)

Background your background's unusually broad for this industry. It's certainly not limited to Abtech or Martech by any stretch of the imagination. You were in finance at Credit Suisse. After coming out of Wharton, right? You went to Wharton for for your undergrad, or was that your business business you double dipped? You went twice for undergrad and and MBA, right? Very cool. So definitely definitely in New York or through and through. You know, I've been here for about twenty six years. Rio was here for grew up in Long Island and was in the city for


Alex Brownstein (03:25.577)

I I double dipped. I did twice. Yeah.


Mm-hmm. Glutton for punishment, I guess.


Rio Longacre (03:32.136)

That's a good NBA program, that's for sure.


Brett (03:41.632)

you know, a couple of decades, right? Probably twenty years, right, before moving to Denver. and you've been you've worked in the ad tech, you've you worked for Live Live In Intent and I think you were in the finance department there, through Series D and and some acquisitions, and strategy consultant consultancy, which Rio mentioned at McKinsey. and and and you looked into a lot of the media network business and and and you've done some investing as a partner at three C Ventures.


Building commercial frameworks behind the firm's first year of growth, which was which was roughly upwards of 50 million, zero to fifty million. So certainly got a broad experience. AHB Advisors is obviously your advisory firm, right? You founded that in 2024. and you've been advising companies and investors across advertising media, ad tech, all these industries that we talk a lot about, commerce media, AI transformation, which I'm sure is a is a hot button topic with a lot of your clients.


and even commercial due diligence. But one thing that we really thought was interesting about your background is, you know, you you were recently on marketture, you've written an essay or an article called Why Ad Tech Stocks Keep Unpr Underperforming Their Fundamentals. So you've been sort of a strong voice in economics on ad tech, right? And and the core argument that I sort of got was you know, that that they're being sort of undervalued potentially for a reason.


Right, where you're seeing companies like like the Trade Desk and App Love and on the demand side that have been sharply punished in market, right? While supplied side platforms like Magnite and Pop yeah.


Rio Longacre (05:18.1)

Certainly the trade desk has been sharply punished. I think their stock is a fourteen dollars yesterday, something like that.


Alex Brownstein (05:21.293)

Yeah.


Brett (05:22.833)

Yeah, it was d down f like forty I've seen between forty and seventy percent in twenty twenty six, and depending on w you know, which research you read. But wow, wow. And then the supply side, like Magnite and Pubmatic have have really rallied and and but I think you've made an argument that I've read a couple of times around how w why really the reasons why behind l the a lot of these ad tech companies are are undervalued potentially.


Rio Longacre (05:29.662)

I guess down almost ninety percent from the top, Brett. It's it's been pretty bad, yeah.


Brett (05:51.565)

and it kind of ties into this topic of value creation and and whether or not ad tech has a kind of complexity tax, right? that that has the investors kind of running scared or if there's concentration with the Googles, the Metas and the Amazons in the space from a from an exposure to advertising perspective in in the investor community. So all those topics are super interesting. you seem super interesting, Alex. It's great to meet you and welcome to the show.


pleasure to have you. And did I get anything wrong? Did I did I egregiously misrepresent you to the public?


Alex Brownstein (06:21.153)

Thank you.


Rio Longacre (06:24.337)

Or did he miss anything?


Alex Brownstein (06:24.791)

So look I think that


You know, I so I guess I I started my independent advisory in twenty twenty four, but have been really focused on it for the past three months because I was at Three C Ventures and Three C Ventures is less of a investor shop and more of also a consultancy. So those would be the things that I would I would clarify. But, you know, person number four there helped build the original strategy, helped build the original model, was a great ride, you know, led initial client projects.


Brett (06:43.763)

Yeah.


Alex Brownstein (06:59.573)

So all fun, all good things. yeah, no, excited to be here and you know, excited to riff on all things media, marketing, nerd, San Francisco, New York, consulting fun facts. yeah.


Brett (07:01.491)

Yep.


Brett (07:12.946)

Yeah. Yeah.


Rio Longacre (07:15.924)

Well, Alex, we're glad thrilled to have you here. Like we like Brett said earlier, we can tell from the start this is gonna be a fun one. And and I heard you on Markitecture a few weeks ago. I thought this is a smart guy, he'd be great, he'd be great for a podcast together. So glad we can make this happen. And the thesis for this conversation, as Brett started talking about, was ad tech is not just really a technology category, right? It's in a lot of ways, and you know, like what I'm we're saying is almost an economics problem because of its complexity and other reasons that we're gonna get into.


The re I mean it people don't realize how big ad tech is, how it really does impact our lives. You open up your phone, you watch TV. I mean, if you're getting hit b hit by ads, and there's be there's a a ton of tech behind that that many people outside of the industry aren't really aware of. But when you look at public ad tech companies, they often trade at much lower multiples. Not e not even you know, trade the trade desk aside, they've been punished for different reasons, but just even beyond that, Brett Brent mentioned some other names. They trade at multi lower multiples than certainly than SaaS companies.


And certainly then in other other sectors, other sectors of the economy as well. so and again, part of the thesis is it is it the complexity, what's really driving that? So we're we're we think this will be fun because you can look at it from several different angles as a investor, consultant, operator, all those things. So let's start out. First question here is why is ad tech so often underperformed? It's apparent fundamentals, and what is the market seeing that the industry itself sometimes misses in its kind of navel grazing?


Alex Brownstein (08:43.103)

navel grazing.


Rio Longacre (08:44.787)

Yeah, the industry loves to look at itself and talk about itself. I this industry. So


Alex Brownstein (08:48.865)

You know, I was thinking of oranges. I think I think


Brett (08:54.612)

I you're making a joke about gazing versus grazing. but anyways, go on.


Alex Brownstein (08:59.723)

Well no, nable oranges, right? Isn't that like the


Brett (09:01.372)

Yeah, naval naval yeah, naval oranges for Yeah, that would be navel grazing. But i I think that's what came through in the audio is the grazing part of this. but go on, sorry.


Rio Longacre (09:02.483)

G that'd be naval grazing. That not not gazing. That that that'd be nable grazing.


Alex Brownstein (09:04.906)

Yeah.


Alex Brownstein (09:11.955)

No, I think look, I think


Alex Brownstein (09:16.863)

I think that there are


a few different components here, right? I think that


There's certainly a lot of nostalgia for the heyday of ad tech when like everyone made money hand over fist and the tech was easy to sell and it was, you know, the wild west of the internet and was underregulated and all these things. So I think that part of it is like we we we enjoy that history.


Rio Longacre (09:51.379)

So when was that? Was that you're like the late nineties or is this like the early two thousands? Like what what would you define as the heyday of ad tech?


Alex Brownstein (09:59.468)

You know, I so the late nineties might predate me a little bit. So I I that I that I have less less insight into. I think it was the the early aughts. yeah. Yeah.


Brett (10:02.483)

Yeah.


Rio Longacre (10:08.679)

Yeah. Up to twenty ten, roughly, yeah.


Brett (10:11.44)

Yeah, yeah, no, exactly. 'Cause late nineties was the advent of the internet, right? right. In the mid two thousand tens was when, you know, the excellates of the blue kys and all these programmatic either data management plant platforms or ad networks or data marketplaces kind of came to the surface, right? Yeah.


Rio Longacre (10:28.487)

It was the Wild West. That's when we started buying media, Brett, like way back then. Yeah. My first signing s signing assertion orders for emails and like run a network display and shit. So sorry, I actually cut you off, but keep going.


Alex Brownstein (10:39.799)

So I think, you know, I think that's part of, you know, I don't wanna say that we're as an industry like self absorbed, which to me is part of navel gazing, right? but I but I think that it's we enjoy looking at ourselves and coming up with like


medio not not mediocre. We come up with like moderate improvements over things that currently exist or take similar technologies that currently exist and like apply them to a new platform and then assume that it's novel and we can be the first person. And I think you know that that can create a little bit of an echo chamber. Now


Brett (11:20.006)

yeah, yeah. I mean Add Exchange, I think of the early days of Ad Exchange or when it was, you know, every other every other week or every week there was a new announcement from a new company that was putting out a new press release, you know, half of which were smoke and mirrors, right? Or or to your point, incremental improvements on something that, you know, was already there, so they're not quite revolutionary technologies despite their claims.


Alex Brownstein (11:41.76)

Right. And I think that like, you know, ultimately where that lands us is in a place where we have the largest companies in the world or advertising technology companies. And, you know, they are the ones that are growing. And they're growing. They're both gaining share and growing. And like most of the rest of the industry is shrinking.


Rio Longacre (12:09.508)

Referring to Google and Meta, correct? And Amazon.


Alex Brownstein (12:11.647)

Yeah, and Amazon. and I think


Brett (12:14.343)

Yeah. Which you which I think you've said they're they're about ten to twelve percent of the of the total S P five hundred value, right? In twenty twenty six. Something in that in that range.


Alex Brownstein (12:22.837)

I would be sh I I don't have that number off the top of my head, but I would think it was more, honestly. I mean like they drive so much of the growth. so I think therefore for a public ad tech company to convince an investor that it is worth the diversification. And by diversification I don't mean like revenue diversification. I mean like pure beta away from the SP 500. I


Brett (12:48.519)

Yep. Yep. Is it is it advertising exposure in terms of like they're trying to they're trying to not overexpose themselves to something that might have you know, downward pull, for example, if there's a recession, things like that.


Alex Brownstein (13:04.407)

Well, I think it I think it's both, right? Like I I mean it is yes, part of it is downward pull in the case of a recession, but also part of it is like, you know, s some of the largest owners of the public markets are just trading off of weights and making sure that they're like hitting their like contracted weighted allocation. And if you already are overweight in something and you call it approximately similar, then you


You know, they can't not own Google, right? They can't not own Meta, right? Like, but they can theoretically not own other ad tech if like they've lumped them together and see it as one trend. I think, you know, and that's on like the passive investor side. On the active investor side, I think that there's you know a little bit of an education that can happen for the investor community, right? I think investors can


Brett (13:47.933)

Yeah.


Alex Brownstein (14:02.099)

generally pay for the gap between return on invested capital and the cost of capital, right? So like we have we could borrow money at X rate and we think that this is going to return Y, right? if they don't understand or they don't have conviction that the that spread is really wide or or or more importantly getting wider, right? So they can make some more money off of it and the rest of the market realizes it, they investors won't bite.


Brett (14:12.329)

Yeah.


Alex Brownstein (14:31.997)

And I think that the the hurdle may be higher, but i if they don't understand the business fundamentals as well. and there's


Brett (14:32.242)

Yeah.


Brett (14:40.828)

And do you do you think this is like sort of a glass box? I hate to I hate to use a trite term, it's it's it's become common recently, but a glass box versus kind of a black box problem, right? Where they're where they I mean are they are investors because you mentioned education, are investors able to really see beneath the hood to understand sort of genuine value creation from let's say an ad tech company that they're looking to invest in, or or even Martech.


Rio Longacre (15:05.094)

Well, I would imagine, but it's it's hard because like it's hard for any industry to discern between some different tech, right? And almost no one understands all of ad tech. It's so freaking complex. But I even look at Google. So Google's Google's performed really well the last couple of years, Alex, right? But I think that Google's performed well because it's being s viewed more as an AI stock. They own data centers, they have Gemini, they have they have Waymo. You could say Waymo is as valuable or could be as valuable as all of Google, right? If it actually does go go does expand globally. So I think Google's not being valued as an ad tech company.


Alex Brownstein (15:13.397)

And I mean


Alex Brownstein (15:18.637)

Sure.


Brett (15:23.283)

Yeah.


Brett (15:29.641)

Yeah.


Rio Longacre (15:34.097)

really anymore. Although Meta certainly is, right? So I I I just think it's interesting. But when I look at all the ad tech companies in my I'm I'm more of an investor than a trader, right? I'm not doing day trades. I'm too busy. But like I but the ad tech companies more following outside of Google, they just and meta, they just suck. I like they've all gone down. Like even like the ones that have gone up. Like I I yeah Viant went up a little bit. Yeah I like them and I actually believe in them. But like it's only gone up ten, ten percent or something like that. So that these these stocks just they they they just get punished.


So is it the opacity what like Brett was saying? Or what do you think is it the the like that they they can't have they don't have a clear narrative about like why 'cause advertising continues to grow.


Brett (16:02.833)

Yeah.


Alex Brownstein (16:11.943)

continues to grow in Google, Meta and Amazon.


Brett (16:15.648)

Yeah yeah, yeah precisely. But within the walled gardens.


Alex Brownstein (16:17.741)

And like sort of and sort of CTV, right? Like within the wild gardens, advertising grows. so I I think and look, I don't know that I believe in the death of the open web, so like we can come back to that as a different topic. But I think look, I think it's a few things. I think first it's that.


Brett (16:31.175)

Yeah.


Alex Brownstein (16:39.327)

investors might not understand, right? I think that it is very technical. I think that anyone who has l looked at a Loomescape would say, like, there's way too many things on this piece of paper. what is the difference between all of them? And oftentimes the differences are quite minute, right? And it I think, you know, it's taken me many years to begin to try to mostly understand the differences and


Brett (16:53.672)

Yeah.


Brett (17:01.105)

Yeah, and the difference is like b across the the Lumiscape, but even between the separate categories that they break them down into, right? Within categories hard and then across categories hard. And so yeah.


Alex Brownstein (17:06.389)

Yeah, exactly. Like what like like if you're talking to Yep. I think that, you know, bring so then bringing that back to like the spread between return on invested capital and cost of capital, like if you don't understand the difference and you don't understand the moat, why would you believe that the spread is going to continue to exist? Right. I think that part of it is like correlation. So like whether or not Google is


Valued as a


mobility stock or an advertising stock, like I would hazard to guess it's valued as an advertising stock, but you know, I don't know. I think that, you know, it it's hard to say that there aren't like a lot of other options for the retail investor that are easier to understand and also that are very, very large and just by virtue of owning the market represent a large portion of your portfolio. So it's like beta argument. And then, you know,


Brett (17:46.418)

Yeah.


Alex Brownstein (18:10.763)

I think that from a transparency glass box, black box perspective, right, I think the


Alex Brownstein (18:22.157)

I think the industry does well when there's a lack of transparency from a revenue and margin generation standpoint, right? I think that can be sometimes challenging for an investor to understand as well, because it's intentionally obfuscated, right? I think that there's


Brett (18:41.957)

The the margins within ad tech, which which tend to be north of seventy percent and have kind of maintained that average for a a long time.


Alex Brownstein (18:42.465)

Yeah, clarity. Yeah, the margins correct. Correct. Great margins business, but it's not recurring revenue, right? So it's like, okay, so you have these high margins. I don't know where they come from and I don't know if this revenue's gonna come back tomorrow.


Rio Longacre (18:48.759)

Great margins, yeah. It's really good.


Brett (18:53.608)

Yeah.


Brett (18:59.045)

Okay. Well no but ha so that's not a like yeah, so they're not breaking down AR versus


Rio Longacre (19:02.705)

Well yeah, a percent of ad spend, Brett, a per yeah, a take rate is not the same as ARR. That those are considered completely different. Yeah, so


Brett (19:07.633)

Yeah. That's a very good point.


Alex Brownstein (19:07.861)

It's not. And I I mean I spent so much time in prior, I'm sent so much time in prior life trying to build a case that like advertising and advertising technology businesses were SAS. And I think it is very challenging. if not a little bit impossible or false. Like, look, you can model anything, like you can put anything onto an Excel sheet and it it'll exist. But if


Brett (19:21.053)

Yeah.


Alex Brownstein (19:35.859)

You have a commitment of a million dollars and all of a sudden, like, you know, the product gets recalled, right? You're just not gonna spend the remainder, right? So that commitment doesn't really mean anything. Whereas if you have paid for a year's of Salesforce licenses, you will pay for your year of Salesforce licenses. It is contracted. And it is is


Brett (19:44.52)

Yep.


Brett (19:54.843)

And it's contractually obligated, right? Whereas me media spin can be variable yeah.


Rio Longacre (19:56.978)

Yep. Or two years or three years. Some of these contracts are much longer than a year, even. I've even seen some that some that are cut that so yeah.


Alex Brownstein (20:03.263)

Yeah, exactly. So I don't I like


Brett (20:05.307)

Yeah, and and and yeah, and SAS it's er or media spend is it's variable and it's and it's not sh you know, shock resistant. whereas CRM arguably, I mean both from a fundamentals perspective in terms of SaaS ARR, but also just from a tech in in use case and value generation perspective, it's sort of a you you know, advertising you could argue you could do it out, right, temporarily to to withstand some of the shocks of an economic recession. Whereas CRM, maybe not.


Rio Longacre (20:32.005)

Yeah, you're right. But if there's a sh yeah, but if there's a shock, if there's a shock, advertising spec you go down immediately, right? I mean those contracts are shorter term, right? They can say we're just not gonna you know, we mu we're not gonna but but you're right. I mean, like it takes it takes time to get out of a Salesforce license or a an Adobe license, like maybe it reops in a year.


Alex Brownstein (20:40.609)

Correct. Correct.


Alex Brownstein (20:47.561)

It's just not it's it can be highly likely to recur, but it is not contractual recurring revenue. So you now you're you get back to the investor, right? You say, okay, like this is not recurring revenue. There's no contract that says it's recurring. Yes, the margins are high, but I'm not sure I understand where the margins are. And it is correlated to other things that I'm already invested in. And like so, okay, great. There's like non-predictable high margin cash flow that


Brett (20:54.898)

Yeah.


Brett (21:06.309)

Yeah.


Alex Brownstein (21:17.791)

I don't know if that's what I need in my portfolio right now. And


Brett (21:20.005)

Yeah yeah. And and yeah, and and knowing that the meta's the Googles, the Amazons who it control forty five to fifty percent of total ad spend, right? You might feel like hey, I'm exposed I'm exposed enough yeah, I'm exposed enough there and and also they have a huge advantage because


Rio Longacre (21:28.645)

And at one point, Brett yeah yeah, it's it's


Alex Brownstein (21:29.247)

Again, I thought it was more than that.


Brett (21:34.481)

Like that value creation argument, they're able to there's like proof of performance because of a closed loop ecosystem or retail media, right? They they can truly they're not coming at you with a bunch of metrics that say, we do perform on behalf of our clients, ROAS, CTR, all these other vanity metrics, right? They're saying no, we can actually prove like causality.


Alex Brownstein (21:40.727)

Correct.


Brett (21:53.837)

of a purchase which is much more valuable I think in the minds of of an investor that's thinking, these guys service brands, right? What do brands care about? Incremental incremental sales, you know?


Alex Brownstein (22:02.731)

Exact well, I do think that's correct, but like, you know, again, the industry insider person in me says like, Yes, but they're grading their own homework and then the question is like the but d like does the does the investor care?


Brett (22:11.157)

yeah, that's yeah.


Brett (22:15.782)

Yeah. Does the investor care?


Rio Longacre (22:16.699)

Well the investor doesn't care if the platform metrics can be gamed. They don't care. They it it doesn't matter to them. Yeah.


Alex Brownstein (22:21.545)

Yeah, as long as it convinces more people to engage in the platform and creates more ad inventory and more demand, right? Like I mean that's the other thing that the platforms have that like most core ad tech, and obviously there are many exceptions to this, so with the grain of salt, like if you're talking about Google Meta and Amazon, right, they have the you know


They have the inventory, they have the tech, and they have the brand demand, which is, you know, theoretically to at least derive from the inventory and the tech. most other players don't have all three, right? Even if you're talking about most retail media, like 70 to 80 percent. I the latest number I saw is closer to 80% of retail media spent is Amazon, right? Like the scale, like is just


Brett (23:00.113)

Yeah.


Brett (23:12.327)

Yeah.


Rio Longacre (23:14.285)

Seventy five billion. It's a monster of a business, yeah, for sure.


Alex Brownstein (23:17.367)

Correct. And it's like everybody else represents like Walmart is large and then everybody else represents like ten percent of the market. And like I, you know, when I used to build these forecasts for what the growth in retail media could be, you know, we would do tons and tons and tons of growth off site. And it and that's where it has to come from because like most of these


Brett (23:46.798)

Which is so unfortunate. Right? Like meaning meaning out of network, right? When you go when you do audience reach extensions for incremental revenue. Yeah.


Alex Brownstein (23:47.083)

Retailers.


Well I I mean like yeah, it was exactly because like but there isn't the inventory, right? Like how the


Rio Longacre (23:57.777)

Yeah, they run out, yeah. Well meta's the same problem, yeah.


Brett (24:00.39)

Yeah, and we and we've had Dr. Augustine Phu a couple of times. He's kind of a friend of the pod, and he and he was he he was he said specifically like the minute you go out of network, outside of these sort of authenticated environments and you go off site, you're just running into kind of a you know, it's the forty four thousand websites where my campaign appeared problem, as opposed to really the the maybe Yeah.


Alex Brownstein (24:05.442)

Yeah.


Rio Longacre (24:20.4)

Yeah, that's w that's where you're getting bogus sites. That's where you're getting you know, invalid traffic, non human traffic, yeah.


Brett (24:24.324)

I should appear in maybe a hundred, a hundred and fifty like premium publishers, but somehow I'm on every AI generated recipe site across the entire world, right?


Alex Brownstein (24:33.996)

And it's also like, like, look, I don't have any numbers to substantiate this, so this is a hypothesis of mine, but I believe that just to take an example, the Best Buy prior purchases of Samsung TVs is a better audience for likely to buy TV purchasers than Google tech enthusiasts, right? Like an however


Brett (24:56.423)

Yeah.


Rio Longacre (24:57.976)

And how's that even defined? Yeah. Some of these some of these audiences. and by the way, just to like to put a number on, like Google, Meta, Amazon ads together, sixty two percent of US digital ad spend. So it's a it's a massive it's it's a and probably mo most of the gr of total US ad spend. Of total U total US digital ad spend, sixty two percent goes to those three or almost two thirds, which is a huge amount, right? And most of most it's most of the growth, it actually goes to


Alex Brownstein (25:08.526)

Yeah, that's what I thought. It was close to sixty, sixty five. Yeah.


Brett (25:12.486)

Sixty two percent of of total retail media. of you of US ad spend. Yeah.


Alex Brownstein (25:21.922)

And how much is


Alex Brownstein (25:26.966)

Yeah. And that's what I was saying earlier. Yeah, that's what I was saying. Like not only is total ads, total ad spend has grown slightly faster than inflation for the past few years, and Google Meta and Amazon have both grown faster than total ad spend. So everybody else is losing a share and they are gaining. And the pie is getting bigger. So it's like the double whammy.


Rio Longacre (25:27.554)

Is is going to them. So that number has gone up. That number's gone up.


Brett (25:41.543)

Yeah.


Yeah.


Yeah. Well and they're and and potentially some of that potentially some of that's growing is because they've diversified some of their product offerings, right? And and what they're bringing to market, whether it's AI or


Alex Brownstein (25:54.519)

Well, I mean, like look, I I think that we were both looking at the same e marketer or portrio and I it's I it's hard to


Rio Longacre (25:59.993)

E marketer, yeah.


Brett (26:04.912)

What's the source of that data? They're aggregating it from a couple different sources. Maybe yeah.


Alex Brownstein (26:06.966)

Well, I mean it's a it's like look, I was gonna s eMarketer's very good. and what I was going to say is like if you add on the growth in C T V to that, or even if you net C T V growth versus linear decline. So like if you assume that like dollars going from C T V from linear are going directly to C T V is just a h assumption.


Rio Longacre (26:11.406)

E marketer's pretty good.


Alex Brownstein (26:35.692)

I think that that is still growing as a share of the pie, which means everything else is really shrinking.


Brett (26:41.008)

Yeah. Yep.


Rio Longacre (26:41.882)

Seventy percent of incremental growth is going to those three, by the way, Alex. Seventy percent, which is crazy. So like for every you know, for every dollar that it gets added. So to your point, like y you know, you advertising grows five percent, but seventy percent of that five percent is going just to those three. So that you know that so they that that share will continue to grow based on current and then and even C T V like they're dominating, like, you you look at what Amazon's doing. I mean, they with with with Amazon Prime Video, for example. I know other people that there's there's I mean


Alex Brownstein (26:42.444)

Alex Brownstein (26:45.859)

Yeah.


Brett (26:46.503)

Yeah.


Alex Brownstein (26:58.584)

But it's


Alex Brownstein (27:07.82)

Right. Which is why it's not which is why it's not like perfect to just sort of add C T V 'cause I don't know where Amazon's including that and I don't know how Google and Z how YouTube is included in that. So like, you know, it's not a perfect, you know, analysis, but I think that like it's more yes, it might be seventy five per seventy percent of the five percent growth, but like given that other channels are in decline, it's in fact faster growth.


Rio Longacre (27:11.472)

Yeah, go ahead.


Brett (27:34.662)

Yeah. Yeah, which is which is your point about linear television and share being stolen or pushed over to C T V. Right? Yeah, so so yeah, so that yeah, it's it's kinda falsely represented. So b back to the topic, I'd love to get into your you know, we were talking you mentioned the death of the open web, and it's certainly a topic that we have heard about. There's a lot of hype, a lot of hand yeah, we've talked about it a lot. And and you know, I talked to a ton of people in the industry that say


Alex Brownstein (27:41.227)

Exactly.


Rio Longacre (27:53.828)

Talked about a lot.


Brett (27:59.874)

It's already dead, right? you know, thirty six cents to whatever forty seven cents of every dollar


is is o like if every dollar is going to publishers and the rest is being kind of siphoned off by middlemen in the ad tech industry, right? it you know, and then you look at what what's happening in the stock market, right? And you're seeing the the the sell side, the pubmatics and the magnets versus the applovin and the trade desk, and there's just completely different trends. One's trending down the demand and buy side, the other's trending up. Do you think that is


Yeah, what what are your predictions? What are your thoughts on on what's happening in terms of the compression of ad tech as a middleman between, let's say, premium publisher environments and brands andor agencies? What sh what are your thoughts on that?


Alex Brownstein (28:46.648)

There are like so many questions inside that question.


Brett (28:48.913)

I that's how I talk. Sorry. Let's break it down.


Rio Longacre (28:51.563)

That was a tough one.


Alex Brownstein (28:53.878)

Like is this like yeah, like like okay, so is it like what do I think about supply path optimization? What do I think about the death of the open web? Am I more bullish on D SPs or SSPs? What else was in there? What did I miss?


Brett (29:03.418)

Yeah.


Y yeah, yeah, I I th I think that about captures it. Well yeah, it's it's more about it so fundamentally it's about the compression of what's happening between between the sell side and the buy side, right?


Rio Longacre (29:16.451)

Maybe s yes, and how that that th that the supply path is part of that, clearly.


Brett (29:20.676)

Yeah.


Alex Brownstein (29:21.654)

Right. look, I think


I think fundamentally, right, everybody always is trying to maximize their working media. Right. And I think that


There have been, you know, again, going back to the heyday of ad tech, right? Like when money was flowing and there was like share to go around, that you know, mattered less. I think it matters more now. And I think that brands are more conscious of, you know, how effective is their media dollar that they're spending. And I don't think that that's unfair. I think that they should be conscious of how they're spending their money.


Brett (30:02.662)

Yeah.


Rio Longacre (30:07.823)

Yeah, they should


Alex Brownstein (30:08.094)

unless they're sp unless they're spending it on me, of course. I think no, I I just digest. I think I think


I so I like I understand the questions, in in trying to eliminate middlemen and the ad tech tax, right? And that's also part of why it's so attractive to just go to these platforms where like, you know


you don't know and you can't know because they're never gonna tell you because they own the whole thing, right? So it's like they're never gonna like Google's never gonna say like, well, you know, this percent of it went to like, you know, the advertiser side and this percent of it went to the publisher side. And this like they like why would they? Like it it's it's just going to Google. so I think


Brett (30:54.469)

Yeah.


Alex Brownstein (30:58.754)

That


is all fair and all true. I think


Alex Brownstein (31:08.13)

I've gone in like massive ebbs and flows on like where I think the power lives between the demand side and the buy side and the supply side. you know, clearly the supply side has performed better in the public markets, at least of late. You know, there were many, many, many years where the demand side was the darling, right? So like let's not for like right now, like there's the right now and there's the prior 10 years. So let's, you know, give credit where credit is due for the prior X X years.


Brett (31:20.795)

Yeah.


Alex Brownstein (31:38.155)

And


I'm not sure, you like I'm not sure that all the tech is required. So like is it do you need both a DSP and an SSP to like oversimplify? Like I'm not sure. My gut tells me no. my


Brett (31:56.367)

Yeah.


Rio Longacre (32:00.579)

That's an interesting question too. Just cutting off to say, so and I guess related to that, so a lot of these ad tech companies have noticed they're all trying to become AI companies, right? And part of that I imagine is devaluation. AI company, you can get a 20x, 50x valuation, maybe 100x valuation. You can get a valuation on nothing now, right? so I see why that's very attractive. And I've been working with a couple startups recently, and like the like the advice from other ads from other advisors has been don't get value, don't get evaluation as an ad tech company.


Alex Brownstein (32:09.955)

Sure.


Alex Brownstein (32:14.339)

Sure.


Rio Longacre (32:27.075)

get out get evaluation as as is as an AI company, which makes sense to me. But the thing is what I think is interesting and the I guess the irony here is that if these like ad tech potentially could wipe out the entire supply path, right? All of these toll takers, Brad, right, that we always talk about in this pod, you know, you know, i i in theory, many of them could be cut out and they're cutting each other off. Thoughts on that?


Brett (32:27.429)

Yeah.


Alex Brownstein (32:40.792)

Well


Alex Brownstein (32:47.648)

Look, I think that there's a sliding scale of every single ad tech player will have an agent and it'll be the same players talking to each other's t just agentically. To there's one AI that'll rule them all and everything else goes away. And I kind of think that the truth will be somewhere in the middle, but I think that the general like strategic principles will need to hold, right? And I think that the question really is like are you creating more value for your client?


Brett (33:16.666)

Yeah.


Alex Brownstein (33:17.314)

than your costing and like Right. And it's like


Brett (33:19.823)

Yeah. Which goes back to that working media comment, right? Like are you actually deriving more working media from every dollar spent because you've simplified and removed some friction from this process?


Alex Brownstein (33:30.456)

Correct. Or are you saving somebody are you are you saving time, right? There are many ways to like quantify it, right? Like if you I'd be willing to give up some margin to save myself X time, right? Like that is effectively a lot of what the agency model is, right? Like, the like


Rio Longacre (33:30.507)

Out of transparency too, you need to prove it.


Brett (33:34.383)

Yeah. Yeah.


Brett (33:45.125)

Yeah. Which which g which you can back into headcount, right? Is because time means headcount. Yeah.


Alex Brownstein (33:48.815)

Right, which you can back into headcount. but I think you know I think the burden of proof is just getting higher. So you have to like really show that. So like if you think about like again, just taking the example of a DSP and an SSP, right? Do you believe that each one is acting as an agent in the pre open AI definition of agent, like Ari manual agent? that like


Brett (34:01.285)

Yeah.


Alex Brownstein (34:17.836)

That where they


Where the demand side technology is really optimizing for the buyer and the supply side technology is really optimizing for the seller and they have competing interests. And having both in the room, like, you know, either manufactured or truthful, like creates a more efficient marketplace. Like, do you believe that? Right. And


Brett (34:28.194)

Yeah.


Brett (34:33.424)

Yeah.


Brett (34:43.535)

Yeah. It's it's a yeah, it's a tension that's necessary for sort of these things to economically function the right way. Yeah.


Alex Brownstein (34:48.544)

Well and it's a belief in that tension, right?


Rio Longacre (34:51.054)

Well yeah, adversarial, but but but but the argument in ad tech is a lot of that the like like the DSPs don't always necessarily just represent the buyer. A lot of them just represent themselves. And we had Dr. Phil on remember Brad, he was talking about dyn dynamic take rates and DSPs stuff he was seeing, right, that he didn't think really made any sense and where they they were actually just using that to increase their take rate, you know, when they needed it to have a strong quarter, for example. Those are accusations that were brought up.


Brett (34:54.011)

Yeah.


Alex Brownstein (35:02.08)

And I and I


Alex Brownstein (35:16.642)

thousand percent. And I'm just saying that I and I agree with that. And I and I'm not sure that both pieces of technology necessarily need to exist for any given you know media buy because I'm not sure that that value is there or that value is commensurate. Like even even if we take the assumption that the both parties have the best intention, right? And they are optimizing for their respective buyers, right? Are they then, you know, like what would need to be true for


it to be worth 20% take rate, right? It'd have to be, or 10% take rate. You'd have to get like, in the simplest sense, like, you know, more than 10% better efficiencies on your media. And I just think that that's hard to prove.


Brett (35:59.279)

Yeah. Yeah yeah, it well it sounds like you're saying the original intent when you had the you know, the R Emanuel sort of the pre agent like AI agent, you know, these two representatives coming into the room, economically it made sense to create some of that tension in a competitive ecosystem to but people game the system, right? And the sis you know, and there there's enough obfuscation and black box sort of reality with this world that pe that brands aren't able to sort of see beneath the surface to say


Alex Brownstein (36:15.214)

Sure.


Brett (36:29.178)

You know, what what is the real value that you're incremental value that you're generating above the amount of money that you're taking out of my working media, right?


Alex Brownstein (36:37.186)

Totally. And I think that that's and that's where you need to like that's where I think that the market is gonna be heading. And if you can like show that, then I think that you're in a an okay place. I think and I I mean it's hard, right? It's hard. And like there are people who wanna do it all themselves because they wanna like control every level and there are people who wanna do none of it and like wanna outsource everything and they're gonna end up with different margin profiles, right? And that's you know the choices that they've made and that's the strategy that they're taking and


I think that's like a purely economic, you know, rational perspective. I think that the reality in this industry is that there's also the personality and the nuance of like the interrelated connected tissue. And you know, while SSPs are doing well right now, like DSPs are still closer to where the money sits.


Right. So like if I were a betting man, right, like even if the SSP is like if you don't need both and SSPs might have marginally better technology and DSPs are getting clobbered right now, DSPs are still have the money relationships. So like, you know


Rio Longacre (37:46.646)

Yeah, well, f for now they do, right? Unless unless that changes. But it's interesting what you mentioned too. So we've noticed like say you'll like you you mentioned one way to think about AI, like ad tech being AI fied, right? Would be you put a an agentic later on top of existing ad tech, and then maybe you begin to cut out some pieces, right? So we've seen PubMatic, Magnite, for example, create like a j a form of agentic trading where right where they go directly, they offer you can basically not it


They have their own kind of DSP-ish functionality they've added. So you can actually s yep. And then, you know, they always obviously have all the rich signal on supply side where there's contextual signal and even publish your first party data that is not used in Bidstream at all today, right? In theory, you can start to use in this, you could have better targeting, you could have better audience development, you could have better measurement, and you could have a lower and I think the reporting is like a slightly lower take rate, like it's it's a little better, right? So so that's one argument, right? But then it's your and then the counter argument is DSP's control.


Alex Brownstein (38:16.462)

Sure, yeah, a lot of the SSPs are doing that, yeah.


Rio Longacre (38:42.892)

You know, DSPs can still have the money, but the th then again, does that change? So where do you see within that dynamic? Like who is the advantage? So are you just saying don't count DSPs out yet?


Alex Brownstein (38:52.974)

I don't think you can count them out yet because they have the relationships. I think


I think that like from a pure tech perspective, I'm less convinced. but from a entrenchment in the ecosystem, how are they thinking through what you know, are are we gonna fundamentally change like how agencies and brands buy who are engaged with the trade desk? Like I


I think you know, from very sophisticated buyers, like I'm still told that it is materially easier to execute a buy on the trade desk than it is to execute a buy on D. And you know, I my experience with physically buying media is like I've spent a hundred bucks on most of these platforms just to say that I've done it and do what it is, but that certainly doesn't get me a DV license. so like I, you know, that's that's my research and hearsay.


But I


think that if you're optimizing for people's time and, you know, overall efficiency of like a process as opposed to like maximizing the margin of each individual point solution, like there is value there. And, you know, and then the question becomes like, how easy is that to replicate? And


Brett (40:21.359)

Yeah.


Alex Brownstein (40:21.782)

Historically it hasn't been that easy to replicate. Or, you know, no one's shown that they can, right? That's not to say that they can't be.


Brett (40:24.591)

Yeah.


Yeah. And well and then to your point about the power of these of the walled gardens. I mean, what I've read, and I think you might have published something on this, but that Amazon's take rate, I mean, they're basically completely underpricing, giving it away for free. Right? Classic like Microsoft bundle strategy, right? We're gonna sell this in z zero to one percent versus a take rate for the trade desk. So is that what the market sort of responds the take rate for trade desk is around twenty percent, right? Is that what the market's responding? Investors are going


Alex Brownstein (40:37.08)

Giving away for free.


Rio Longacre (40:42.125)

And own an operated inventory, they've been giving it away for free, yeah.


Brett (40:54.832)

Dude, they're gonna eat your lunch, right? You know, because advertisers are gonna to to go towards that knowing they can put more dollars into working media.


Rio Longacre (41:02.573)

Well well I think Brett to Dan to I before Alex jumps in, I think that tr the trade desk stock got pummeled because gre remember Jeff Green was saying Amazon DSP is not and they just renamed it by the way. It's not Amazon DSP anymore. It's gonna be a a j s something a gentic buying platform or something like that. But it but but I th but I think they it got it got they


Alex Brownstein (41:17.23)

Course.


Brett (41:18.478)

Hm. What is it? What Alex, do you know it the actual name? okay. Yeah, they're they're getting rid of the word DSP. Out of it d yeah.


Alex Brownstein (41:22.006)

No, I just said of course. I I


Rio Longacre (41:25.377)

Yeah. It's a gentle example. They just announced it today. I just I just heard it earlier, I forgot, but I'll pull it up in a second. But the thing is, or maybe you can search what I'm talking about. But like the thing is it's he said it's they're not a competitor, but then you know, they certainly are a competitor, right? And you they've expanded their third party supply dramatically. And and to your point, Brett, they're offering on some of their buys, like zero zero one percent take rates. I mean, yeah, they so like a lot of ad spend is certainly shifted over there. So the D SPs do have power, Alex. So I I think that's that's like a good point. So


Alex Brownstein (41:28.205)

Yeah.


Rio Longacre (41:54.913)

So thoughts on that?


Brett (41:57.54)

The Amazon ads agent.


Rio Longacre (41:59.564)

Amazon ads agent, that's right.


Brett (42:01.613)

Yeah. Which is which is interesting. I wonder if they what the what way they mean agent. Agent is an AI agenc trading or


Rio Longacre (42:08.717)

Well well what they showed today, Brett, it was like the meta today they pr it was the it's the meta model. You you tell us what you want to sell, you tell us your budget, and we'll do everything. It'll just we'll push it through this like you're not gonna be in a control panel, mess around with bids or we'll even do the creative for you. It's kind of like like what Zuckerberg was saying a couple of years ago. Like the his vision, that's what Amazon's going towards full speed.


Brett (42:18.606)

Yeah.


Brett (42:26.435)

Yeah.


Brett (42:30.061)

Yeah, all through a conversational interface of some sort or another.


Alex Brownstein (42:34.072)

So like sorry, to get to your question, which I want to make sure I'm thinking about properly, you ask you guys ask a ton of questions in a row with minimal pauses. And and and it's like, wait, dude I mean like I'm I'm I'm I'm doing the best I can. You know, you you gotta give a man a chance, like I I think like


Rio Longacre (42:46.977)

We do, we do.


Brett (42:49.143)

Yeah. Well I told ya I told you we're ADHD. We're all you know


Brett (42:56.643)

Take your time, we'll slow down. You do great, Alex. You're awesome.


Rio Longacre (42:56.781)

You're doing a great job, Alex. Don't worry about it.


Alex Brownstein (43:03.956)

Look, I just did some quick research, right? Like,


Alex Brownstein (43:10.571)

It seems an


that the trade desk and I'd love to pull up the actual filing, the actual thirteen F. but it seems like it's, you know, as I would have expected


somewhere between 60 and 80% owned by like BlackRock, Fidelity, right? It's also 11% owned by Jeff Green, right? Like I think I think the


Brett (43:36.868)

Yeah.


Alex Brownstein (43:47.587)

You know, these are businesses that yeah, I mean as of six thirty, the largest shareholder was BlackRock, then State Scre then Vanguard, then pension funds, two sigma invest so like these are mostly, you know, it's again minimal research in institutional investors that like are owning, you know, likely predetermined like


Brett (44:09.784)

Yeah.


Alex Brownstein (44:17.43)

If you buy the BlackRock S P five hundred fund, I guess Trade Desk is not part of the SP five hundred anymore, but like like BlackRock has to own them, right? Like so it's like it's I think that like asking questions around like are they looking at the take rates of you know Amazon versus the Trade Desk as an existential threat is, you know, the question that is sort of reserved for people who are actively investing and like managing portfolios and like a


meaningful way, which represents a smaller share of the total like pie there. Now, granted, like they're the part that is likely to sell rebalance move more because, you know, they have more active control. But, you know, those people are also likely to maybe do the research to try and figure it out and understand. I do think that like Amazon and


Brett (44:51.405)

Yeah.


Alex Brownstein (45:12.928)

as it differs from the trade desk like does own inventory. So part of like we're trying to get you on the Amazon DSP to buy everything and is also like to get more on Amazon inventory. And, you know, we talked I I one of your many questions that may or may not have been answered in the past was like about the death of the open web and like it's still where people are spending most of their time. even though, you know, like the the imbalance between time spent and money spent on the open web is still very large. So I think that


Brett (45:42.669)

That that's a really good point. Yeah. The imbalance between time spent and money spent on the is is there's an imbalance there, right? Based on the amount of time spent. Yeah.


Alex Brownstein (45:44.663)

Like I think


Alex Brownstein (45:50.167)

And like and I think that like the so now if you're buying on Amazon D S P, right, and granted or I've


Rio Longacre (45:50.196)

else have been balance, yeah.


Alex Brownstein (46:02.294)

I don't have an Amazon shop, so it's like hard for me to validate this, but I would be if if I were Amazon and there were a place where I could keep all of the margin and you know place the ad in my own inventory versus put it on somebody else's inventory, where do you think the ad is gonna go? Right? Like I so right, like keep it in the ecosystem. So like what


Brett (46:18.305)

Yeah. It's like circular economics right there, right? Keep keep it in the ecosystem.


Alex Brownstein (46:27.93)

y you know, and and again, I doubt that Amazon will ever tell us, like and why would they? But like I think that's that's


Brett (46:33.006)

Yeah.


Alex Brownstein (46:42.186)

Even if they're giving it away for zero, right? Like, how much is that certainty worth to you? So if you're a brand, right? It's not like a an either or it's like, is it like if you believe in the open web and you want to buy open web inventory and you might also be buying Amazon, but it may not be a core channel for you, like, is that 20% fee worth it to you to know for certain that you're going to be not on Amazon owned and operated inventory? Right? I, you know.


There probably some Amazon competitors, and Amazon kinda half does a little bit of everything, right, that might say yes. And that's like a strategy decision that like you have to sort of think through and what's that worth to you?


Brett (47:26.521)

Yeah.


Alex Brownstein (47:28.152)

I don't know. But


Yeah, the between Amazon, Google and the AI overlords, who knows? We're doomed.


Rio Longacre (47:33.6)

Yeah, well you can see


Yeah. I I wonder if we can maybe talk about a hold coast and hold coast stocks for a minute too. I think this would be you know this would be this would be kind of interesting. So they they've not done well the last few years, right? It's been all the major hold codes have been completely pummeled. Stocks are stocks are way down. I think like Publicis' stock has done the best of the bunch. but it certainly hasn't been yeah, it that they've been they've been up a little bit, but but it still hasn't been like a you know, it's not like it's it's not like it's followed the rest of tech, right? I mean


Alex Brownstein (47:42.36)

Sure.


Alex Brownstein (47:47.702)

Alex Brownstein (47:56.16)

Yeah. I thought it was up but


Rio Longacre (48:06.196)

Or follow tech, right, the last few years where the valuations have gone up significantly. So I mean, you could argue with what's going on in AI. This is this is this is a serious threat to the current model, which has been based on mostly in FTEs and even on in the even on the media side, right? There's been increasingly clients have been not all clients, but many clients have been demanding more transparency the last three years and understanding like where fees are coming out. And there's, you know, there's the re the WPP rebates scandal.


Alex Brownstein (48:08.344)

Yeah.


Rio Longacre (48:35.184)

in China where they were accused of taking rebates at and not disclosing rebates and then and just pocketing them basically. There there's a big lawsuit going on in China right now because of that. So I mean so thoughts on the future of the agency, you know, let's say the the valuations of these these companies w with these headwinds of AI, of potentially more transparency and all these other things, do you think that they they've bottomed out or do you I'd be or do you think or do you think that these are potentially like more shocks?


Alex Brownstein (49:04.79)

Okay. so


Just for the record, right? Like I'm a strategy advisor. I, you know, I I like thinking about these things and I am a you know, I and because my background started in finance, I try and ground a lot of my analyses in financial metrics, but I'm not a financial analyst and I please do not view this as like I and and any of this not not


Rio Longacre (49:35.338)

This is not financial advice. Like no one no one's here listening to this pod getting financial advice.


Alex Brownstein (49:38.83)

But like also just to like clarify like where I spend my time. I think look, hard for me to say if the agencies have bottomed it out from a stock perspective. I think that what is true is that the agencies have never been technologists themselves. They've gone on buying sprees. And, you know, Publicis obviously is the


Obvious one, right? You know, LiveRAM, BEPsilon, et cetera, et cetera. So like they're trying to build their own like quasi-Walled Garden mini ecosystem. and they're winning, both in terms of stock price, but also in terms of just net new clients and they're winning, which is great. I think that, you know, does AI represent a threat to a per hour model?


Sure, theoretically, I haven't yet seen anybody effectively replicate this at scale. like it was kind of funny actually. I read the other day that Meta is now charging a SaaS subscription for access to people. which I thought was like a wild turn of events. but


Alex Brownstein (51:01.9)

You know, I I think that it's hard to I haven't yet seen I've seen lots of really sexy demos of agentic buying. I haven't like seen a lot of real case studies of it happening. So you still kind of need people to like plumb the pieces together. And even with that, right, like the sexy case studies are exactly that. Case studies are PR and fun to talk about and say, like, look how cool and sexy we are, and they're not like


you know, handing over millions of dollars of real budget to happen without human oversight, right? Like I haven't seen that yet. and I don't know if or when it'll happen. So like there all


Brett (51:41.987)

Yeah, well in a and in the in the OS, the operating system model of of like the big hold codes, I mean, I you know, from what I've heard from insiders across the space is that there's a lot of AI veneer, right, and concern, but most of this work yeah, AI theater, but most of this work is still being offshored and outsourced and in right, to like small armies of people that are manually doing stuff.


Rio Longacre (51:57.738)

AI theater.


Alex Brownstein (52:04.44)

Totally.


Alex Brownstein (52:08.768)

Yes. I I agree with that. And I think that like, you know, if you're working on a major client, like it's like very high risk. and then who's gonna be the first person? so I don't know. Do I think, you know, to a lot of the conversation that we've had earlier, do I think that like media buying can become more efficient and tech enabled? Yes. Do I think that a per hour model is


Brett (52:17.859)

Yeah.


Alex Brownstein (52:38.624)

maybe not ideal. Also probably yes, but it sort of depends on the work you're doing. do I think that humanity is, you know, gonna have to learn to change and evolve and like figure out what's next and new? Like potentially, but like, you know


Brett (52:43.618)

Yeah.


Alex Brownstein (52:56.876)

When we went from the agrarian society to the industrial revolution when the steam engine, we had the same thought and fear. And then there was the assembly line, we had the same thought and fear. And we had the first generation of the internet and there was the same thought and fear. And humanity has existed. So like I I believe in the resilience. so I think that like


Brett (53:12.96)

Yep. we we could we could get into AI doomerism if you want, right? There's a I've just I just I've just watched it. Yeah, the AI doc, right? How I became an apocalypist. That's actually sounds like a snuffilothagus. An apocalyptolist was a really interesting documentary that kinda looked at both sides. It's on Netflix. It's called the AI Doc. You can kinda forget but but that term is the funniest term I've ever heard. Apocaloptimist. Because


Rio Longacre (53:18.027)

Ha ha ha.


Alex Brownstein (53:18.498)

Yeah, it's


Rio Longacre (53:27.125)

Yeah.


Alex Brownstein (53:27.629)

Yeah.


Alex Brownstein (53:32.492)

Wh what what what platform is it on?


Rio Longacre (53:39.871)

Yeah. I I I've I haven't watched it yet. But this pod we've had a lot of discussions about AI. And I think generally, Brett, both you and I are like pro AI, pro data set, or like, you know, we're not AI doomers. And I I we and Alex I published an article on singletnoise.ai, our website that talked about is it it was like is AI going to kill us all? And then my conclusion was no it's not. You're a bunch of like you're butch is a bunch of fear mongering and a bunch of disinformation or misinformation was kinda my but kinda my conclusion. But I guess


Alex Brownstein (53:41.134)

Hallo.


Brett (53:45.686)

Yeah.


Rio Longacre (54:07.883)

I don't know, but looking at the agency model, I guess my my my opinion on like like I I do think the FTE model is is gonna go away. Like is there's no way it doesn't go like it is gonna get compressed. And that's probably a good thing, but it's been going away for a while, right? It's gonna get But you're not but you're not gonna be able to like no you can't charge for people's like hours of work anymore. What is an hour work anymore with AI? You it's like you can compare apples to apples with anything. Go ahead, Alex, I'm gonna cut you off.


Alex Brownstein (54:08.632)

I but I think I it's a


Brett (54:19.062)

It's gonna get compressed, not go not go away entirely. Yeah.


Alex Brownstein (54:21.166)

But but I think that


Alex Brownstein (54:31.768)

Sorry, I think I like I'm not sure I disagree with you. I do think that it is harder to do that than, you know, it's not like you snap your fingers and it goes away, right? Like it's like re re beyond just like the recontracting everything, it's rechanging like how teams are structured, how you know people engage with each other, how people engage with clients. Like and then I think even it's a delivery model shift. I think that it's like I think that, yeah, I think that it's a


Brett (54:40.578)

Yeah.


Brett (54:51.02)

Yeah. It's a deliver it's a delivery model shift. How are you delivering solutions, right?


Alex Brownstein (55:02.048)

it it it also requires, I think, a level of standardization and like technical sophistication and enablement within an agency to have an own that I haven't quite seen exist yet. And you know, I think that there's some agencies that are maybe more unified than others, but ultimately, right, if you think about the classic holdco model, right? It was like


We would acquire people who had big personalities and big books of clients so that we can increase our media spend, right? And the thing about people with big personalities is that they have big personalities. So like there's lots of little fiefdoms in these agencies that like, you know, exist under their own names and like a lot of the names have been wiped away, but like we still say them and you know, and that's the way they are operating. So in order for you to get like true scale and efficiency that way, I think you have to standardize processes, standardize tools, like


make people work in a more unified fashion and get rid of a lot of that like personality, which I think is hard. And I'm not saying that it is the wrong answer. I'm just saying that it's like a lot more complicated and, you know, challenging than people are giving it credit for. I I also


Brett (56:03.115)

Yeah, yeah.


Brett (56:10.805)

Yeah, yeah, th it's yeah, Rasha Rashad Tabakowala said exactly that. He said pe companies don't transform to use that that that bad word transformation, right? People do. And I'm seeing this even with clients in the large indie space, not just the the the ginormous holdco space, is the organizational do not underestimate the size and scope and fragmentation to your point about silos.


Alex Brownstein (56:19.809)

Exactly.


Alex Brownstein (56:26.925)

Yeah.


Brett (56:38.109)

of the organizational change. It takes education, it takes reskilling, retalenting, the delivery model changes. You know, people immediately think, my God, we're well are we just becoming a SaaS company? And you're like, no, this is actually something completely new. It's sort of a a a you're offering a managed service to sort sort of guide your clients through something, right? Whether it's a creative use case or a measurement use case or an or right


Alex Brownstein (57:04.524)

And I think I and I think that like it yes, I mean yes, and right, like it starts even more basic than that, right? Like if you are working with people who some of your stuff is saved on Microsoft and some of your stuff is saved on Google and some of it isn't an email and some of it is in Slack and then like it might end up right, like and you're trying to like move it around to like different locations, right? You're


Brett (57:06.699)

But you're doing that with tech, but you're not a SaaS company.


Brett (57:24.405)

And some of it's in your chat GPT and your markdown files and your


Alex Brownstein (57:33.127)

you have to do all of that work first and at scale across an agency, right? Like so it's like you you you have to be able to convince like all the people who have personalities that you acquired because they have personalities, which means they have clients, right? to change the way that they've done things that has given them the personality in the client to create scale for an organization that they no longer own fully. And I think that that is, you know, a hurdle. And then on top of that, right, like but I think often


Brett (57:51.522)

Yeah.


Alex Brownstein (58:04.13)

doesn't get talked about is where is the margin coming from for these agencies? And a lot of it is hidden in these like principal media vies, not the hours. So like so like so like if it is in principal media, not the hours, like does getting rid of the hourly rate like necessarily change anything? And what is the outc


Brett (58:13.229)

Yeah.


Rio Longacre (58:13.61)

Media, yeah.


Brett (58:21.794)

Yeah. And yeah, and Rio, you've talked about the float too, like where they're basically either they're offering a d you know, third party technologies and and helping to kind of orchestrate and manage across those technologies, whether it's through C P connections or if it's just some other mechanism and then just adding a bit of margin to each one of those layers, right? Which is another way. But then but then that's not particularly defensible.


Alex Brownstein (58:32.759)

Yeah.


Rio Longacre (58:33.736)

Alex Brownstein (58:42.978)

Yeah.


Rio Longacre (58:44.242)

Yeah. Well you can see why agencies yeah, for agencies that would be very attractive. 'Cause in for by all for all intents and purposes, these big hold codes really are like financial institutions, right? They're taking in tons of media spend. A lot of it's principally spine, rebates coming, all these other things, they're managing that and they're buying media, you know, across all these different places and they're they're they're they're managing the float of that. They they have lines of credit. So that there's a lot of money flowing through those pipes, and that's how the lot the agencies, Alex, to your point, really, really make make the most of their money, right? Not not based and then they have to support F


Alex Brownstein (58:52.738)

Totally.


Rio Longacre (59:11.358)

hire FTEs to actually execute the work that they're then getting paid that way. So it's a complex model. And I think that but as like AI is going to change l many industries, probably every industry, right? It's going to change SaaS, because guess what? You can't, it's going to be harder to like bill clients for seats or for licenses. The same way, there'll still be ways to build clients, but it'll just be different, right? And the cost structure will be different too. Token prices will fall, but you still have to pay for tokens, right? but then you look at agencies, as as AI becomes bec


gets used more, they'll need fewer people. But I I my guess is clients will start will start asking for more discounts. Okay, you've been going on and on every year about it with this AI theater, show me the money at some point they're gonna say, okay, we want to and that we're already he I'm already hearing that. Clients are saying a 10% discount, a fifteen percent discount or whatever discount because you've been saying you have AI, which you know look if I were a client I'd probably be asking the same thing, right? But but I think it is but I do think though that one opportunity people aren't seeing is as we if we go to a gentic


Brett (59:52.333)

Mm.


Rio Longacre (01:00:07.025)

If AI starts doing a lot of these buys, does some of this ad tech get dinner meet get disintermediated? Agencies are paying it out in tech tax today. It only ultimately comes out of working media, but a lot more money becomes available. So that's also an opportunity for agencies. I think not everyone's talking about. That's kind of interesting. And and so so and and it's before it passed. Go ahead, sorry.


Alex Brownstein (01:00:23.585)

I think


I I mean I okay, so there's a lot in there and I wanna bring up one final point about agencies. I agree with most of the things that you said. I think that if an agency goes to a brand and says, Hey, like we're just intermediating tech, we're lowering tech tax, like I'm keeping the margin, right? I don't think that that would land well with the brand. Like an


Brett (01:00:49.888)

Yeah.


Rio Longacre (01:00:51.817)

I wouldn't fly, probably no.


Brett (01:00:52.427)

Yeah.


Alex Brownstein (01:00:53.942)

So i I mean like i did I misunderstand your statement?


Brett (01:00:56.467)

No, th that you nailed it. You nailed it to like six words, right? That was awesome, right? No, no, because that's exactly what it is and that's exactly what the agencies are aiming for.


Alex Brownstein (01:01:00.014)

Alex Brownstein (01:01:06.24)

I but like look, I said this, I said this on another podcast, and I believe it. I think that if we go back to what LLMs are, right, they look back on the corpus of everything that has ever been done and predict the probability of a next thing, word, token, pixel, what have you, based on what you know, you kind of guide it towards and adjusting probabilities, right? Nowhere in there is net new creative.


Right. So like my hypothesis on agencies is more Don Drapers, fewer Harry Cranes. Right. Like, you know, I don't think that and I think a person working with AI and like to iterate on ideas with AI and to version a creative with AI can be very, very creative. But I I don't know. I haven't yet seen the AI generate the net new, never heard of big new.


Brett (01:01:43.372)

Yeah.


Alex Brownstein (01:02:04.788)

idea you know in advertising is that the next new WhatsApp or Clydesdale's or or you know I'd like to teach the word just where's the beef, right? like right like so I I I think you know similar to what I was saying about the tech tax and what are they doing for brands and agencies and who are they being an agent for, right? Like agencies, you know, will


Brett (01:02:11.242)

Yeah.


Where's the beef? Right? Yeah, just do it. Right?


Rio Longacre (01:02:16.594)

Love that, yeah.


Alex Brownstein (01:02:34.998)

in my view, need to show like what are the value added services that a human can do, that a human can provide that you're willing to pay me for, right? And are they and like I do think that this is a have to be a shift to their model. I do think that other people are gonna have to be retrained. I do think that they're gonna be, you know, i like changes. But you know, how can we create


value add if like the actual act of media buying in those hours are more automated and if you know the principal media becomes more transparent and we can no longer make money that way. Like that I think is like the fundamental question that agencies are going to have to grapple with. while also, you know, likely buying technology because they've never been the best at building it. But I think


Brett (01:03:09.665)

Yeah.


Brett (01:03:14.166)

Yeah.


Brett (01:03:19.617)

Yeah, and I and are they really competing against, you know, becoming more like consultancies potentially. in terms of adding value and human judgment and strategic acumen and direction and knowing Yeah. And knowing how all this stuff works, that's the challenge. If you've been doing kind of a lot of production work, just base level production work and a lot of media buying work, that stuff, you know, there's only so much strategic value you can add to that.


Alex Brownstein (01:03:25.528)

Sure.


Alex Brownstein (01:03:29.549)

I mean


Rio Longacre (01:03:30.747)

Accenture's getting back into media buying, by the way. Like they're they've announced that.


Brett (01:03:47.394)

mix, right? A media mix. Well that's that's what we do on this show, right? But yes. Yeah


Alex Brownstein (01:03:48.386)

I have many, many thoughts on this that are more appropriate for a beer than a recorded line.


Alex Brownstein (01:03:56.622)

I mean I I see I see no beverage.


Brett (01:03:59.8)

How do we move how do they move up the value chain to offer more strategic advice and knowing where it fits into the current ecosystem, whether it's the the clients you're talking to that have procurement teams that are trying to drive price down, the role the technology plays. And if you come from an older world, the d you know, and I do love your your example of Don Draper versus Harry Crane, and we've actually had people that are in industry insiders that have never watched Mad Men. And I I I say to them, you must watch every season.


Alex Brownstein (01:04:10.114)

Yeah.


Alex Brownstein (01:04:14.423)

Yep.


Rio Longacre (01:04:29.043)

Shame. Yeah. Th there's no there's no excuse for that.


Brett (01:04:29.92)

of madmen. I mean in the industry that we're in, not only is it terrific television, terrific writing and acting, but it does show how technology shifts from the rise of television to the the the first computers that filled up that giant room, right? Where they were trying to automate a bunch of this stuff and how these things are they kind of they're cyclical, right?


Alex Brownstein (01:04:43.864)

Mm room, yep.


Alex Brownstein (01:04:50.134)

I mean, I I I love the show. I my hot take about the show is it's not about Don Draper, it's about, you know, it's about really Peggy and Joan, but also about industry shifts, as you're saying. But like even if you go back to like the original Sterling Croop Sterling Cooper, right, when they got sold and I don't know for those of you who are


Brett (01:05:06.636)

Yeah.


Alex Brownstein (01:05:16.826)

never seen the show and are listening. This is a little bit of a spoiler. so I would I would hope so, but I like, you know,


Brett (01:05:19.934)

Yeah, spoiler alert, spoiler alert.


Rio Longacre (01:05:22.921)

I think most of our listeners have seen Batman, but keep going.


Brett (01:05:27.852)

Richie Glassberg, you've gotta watch Madman if you're listening. Yeah, Richie Glassberg, who is an incredibly smart and very well read man, does not watch Madman because he doesn't watch much television apparently. But go on.


Rio Longacre (01:05:31.133)

He called him out.


Alex Brownstein (01:05:31.414)

Yeah.


Rio Longacre (01:05:37.182)

Yes.


Alex Brownstein (01:05:41.004)

Well, I think I think that like when Sterling Cooper was getting acquired by the British firm the first time and they, you know, they became Sterling Cooper Draper Price, you know, they brought in Harry Crane and you know Roger Sterling was like, Why are we bringing this guy in like we just need like account men?


Rio Longacre (01:05:41.225)

It reads a lot of books, so


Brett (01:05:51.681)

Yeah.


Alex Brownstein (01:06:06.302)

And Harry Crane looked at him and said, like, I control TV and this is X number of media dollars and revenue. Like that's where like the money's coming from, right? Like I think, you know, that to me was like early beginnings of Holdco, early beginnings of shifting of influence from the Dom Drapers to the Harry Cranes of the world. Like, where is this nascent business getting money? Where is it getting revenue? And


Brett (01:06:13.611)

Yeah.


Alex Brownstein (01:06:30.636)

you know, a a lot of what I do is look for patterns across history and see where like businesses can be comparable to one another. Like I kind of go back to those days and say, like, hmm, are we unwinding that a little bit? Like are we getting to a place where, you know, the media is now commoditized, right? Like people who want the direct access to the direct publishers can the premium publishers can get it. You don't have to go through like the tech tax. Like if you're a large agency, a large advertiser you maybe even in-house


all or some of your like easy buying, right? And you leverage your agency for the complicated things were linear, which is freaking, right? So like are we going to a place where you want to have like that direct relationship with a human who can like help inspire you with that creative idea? And I I, you know, I have not yet seen AI do that.


Brett (01:07:04.94)

Yeah.


Brett (01:07:13.794)

Yeah.


Brett (01:07:22.315)

Yeah, cre creative is a new targeting for a for a quote that's been thrown around the industry for a while, right?


Rio Longacre (01:07:23.608)

Yeah.


Yeah, and looking at the agent, the hold coast stocks. I pulled up at AI created chart here. So s last five years, Publicis up two hundred two almost two hundred and four percent. So it's been a pretty good risk. Five years, right? Omnicom went 120% of where it was five years ago. Avas 121%. Densu ninety eight percent. WPP fifty percent. So some of them, you it's five years. Some of them, you know, last couple of years it's not quite as rosy, but so


But it's been okay performance, great performance for publicists. Omnikah I'm okay and then Havas okay.


Alex Brownstein (01:07:59.427)

I don't know where you I don't know what you're looking at. My my my Google finance is telling me it's up sixty six percent is up sixty six percent, but


Brett (01:08:08.327)

yeah, maybe he's got an L L maybe it's an L L that's hallucinating.


Rio Longacre (01:08:08.648)

This was from a baseline. L L made this, yeah, so maybe this is yeah, maybe this is L L is hallucinating and lying to me. I don't I don't know Nate Bullets' stock, so I w I don't really care. But but


Alex Brownstein (01:08:13.422)

And I know.


Brett (01:08:14.549)

It's telling you what you wanna hear, Rio. It's it's a


So so yeah, so so maybe those numbers are put off, but


Rio Longacre (01:08:25.202)

No, those are percent. This is a baseline okay, it's a baseline of a hundred. I see. Okay, got it.


Brett (01:08:30.603)

Yeah. It's a it's like an index value.


Rio Longacre (01:08:31.176)

So okay, so yeah, so it's not a percent. Okay, I see what it created here. Yeah. So now so the stocks though they it's done okay. I mean that's not six up sixty some percent. That's not bad in five years. It's not great, right, compared to like some of the tech stocks, that's for sure.


Brett (01:08:45.889)

Yeah. So so how does this tie back to the the rise of creative, right? You y you seem to be saying that the that decoupling that happened a few years ago, quite a few years ago, between media and creative shops, you know, are they coming back together? Are we seeing a a focus back on the value of creative in in in advertising that is the Don Draper? Well yeah, there was a decoupling of media and creative.


Alex Brownstein (01:09:07.896)

Sorry, you said decoupling.


Brett (01:09:13.451)

or you know, organizations, right, from an agency perspective. You had you had a bunch of agencies that went and did pure creative and project work, right? And then you had your media buying agencies that yeah.


Alex Brownstein (01:09:18.648)

Sure.


Yeah, sorry, yeah. I I I I I follow. I think look, I think time will tell. I think that we're gonna need


Alex Brownstein (01:09:37.807)

I I again like looking at it from a variety of different perspectives, right? I think that there's what does the brand want? Like I think the brand wants an integrated marketing story, right? I think that the brand is probably not organized internally to accept the integrated marketing story. but I think that ultimately they're gonna need and want an integrated marketing story, which I think involves having


Brett (01:09:58.593)

Yeah.


Alex Brownstein (01:10:06.646)

an integrated team working on their behalf. I think that what does the agency want? The agency, my guess, is like the type of person who is buying lower funnel is different from the person who understands branding and upper funnel is different from the person who is like a creative thought leader. So like, you know, separate your people a little bit. Also trying to understand like where the profit centers are from where the cost centers are. And like there are many ways of slicing that that could be on a client basis, that could be on a team basis.


Brett (01:10:10.752)

Yep.


Alex Brownstein (01:10:36.199)

but I think that as those elements are continuing to shift, right.


My guess is that like media may be like look, I think if an agency is like bringing along creative thought and strategic advice and media becomes easier to execute and requires fewer hours, what have you, right? Then, you know, and becomes more transparent, then I think that like logic would indicate that the media becomes a lower margin business and the strategy becomes a higher margin business. so it you


Brett (01:11:11.391)

I exactly. Readjust your pricing model, right, to accommodate for that.


Alex Brownstein (01:11:15.358)

we might read readjust the pricing and also like the organizations that like it's clear what you're bringing that to the client, but then like they're gonna wanna have to they're gonna have to like kind of match what their clients want. Like they're ultimately in a client services business so they have to, you know, service the needs of their clients.


Brett (01:11:28.535)

Yeah.


Yeah, one way or the other they're not becoming and I was talking to Adam Gitlin.


the other day at a at an event and and and we were we were tr talking about like can agencies become or the challenges let's say of becoming technology companies and a lot of them from a a valuation, those that are PE backed and PE owned, and I'm seeing this with with some of my own clients, are under a lot of pressure to change how they are self-defining themselves or self-defining. Because as a managed service without going through some sort of transformation, whether you're gonna say it's becoming a tech company or not, the point is


Alex Brownstein (01:11:35.937)

Yeah.


Rio Longacre (01:11:37.65)

I don't like you.


Brett (01:12:05.583)

Is that they're gonna they might get a two X multiple on a typical FTE model, but as a te if they're tech enabled, let's say, like in in a unified tech stack type of model, you know, those valuations are gonna go much, much higher. And so it seems like is that is that an accurate read of


Rio Longacre (01:12:21.447)

Yeah, it's a valuation yeah. Well yeah, I mean that yeah, but I mean if you're if your services or g it's gonna be evaluation of EBITDA, right? But whereas if you're a tech company it's a valuation of overall revenues, right? Of ARR, right? So it's a it's a totally different calculation. It's a five to seven X, you I maybe it's gone down a little bit recently. And if AI company could be twenty X, right? But so sorry, over to you, Alex.


Brett (01:12:33.163)

Yeah.


Yep.


Brett (01:12:41.079)

Yep.


Alex Brownstein (01:12:42.99)

I mean I think okay, again, lots of questions, lots all at once, what lots of thoughts. I think I think look, I


don't know that I have seen an agency build its own proprietary tech without acquiring it.


Alex Brownstein (01:13:09.25)

very well. Now I'm sure that there are missing examples and I really don't want to, you know, I I don't want to like alienate folks. So like happy to hear any and all like examples, but I think that like most of the agency tech has been acquired. So I think then the question becomes like, can you buy your way into a tech valuation? And I don't know. okay fine. Now you have like this combination of like


Brett (01:13:10.156)

Yeah.


Brett (01:13:29.815)

Yeah.


Alex Brownstein (01:13:38.105)

People enabled technology, which is, I think, like maybe a future that we can envision with AI. You know, I don't know what exactly that looks like yet. So, like going back to the past, if you think about the spectrum of services revenues or services margin, right? There's on one end like the McKinseys of the world, which are really, really, really profitable and like very expensive and high margin, and all they charge for is people.


And then there's the sales forces of the world on the other end of the spectrum that give away the services for free to get you lo locked into the software. Right. And if we go back to like, you know, what we've said about organizational change and service change, right? Do I fundamentally believe that like agencies are gonna stop charging for people at all and become a sales force model? And I'm not saying that that necessarily is necessary, but like just to take a the the extreme end of one argument, right? like


Brett (01:14:11.627)

Yeah.


Brett (01:14:28.588)

Yeah.


Alex Brownstein (01:14:35.532)

I think that that would be a fundamental shift to how they operate their business and I I just I don't know that I see that happening.


Brett (01:14:40.522)

yeah.


Brett (01:14:44.075)

Yeah, maybe an impossible shift to your point. They're not gonna become Salesforce or that model.


Alex Brownstein (01:14:48.94)

And I think and


Rio Longacre (01:14:49.223)

Yeah, well it's but it's interesting with the SaaS companies like Salesforce Alex. So like like I've worked with them for a long time in Adobe and like most of the big SaaS companies. So like yeah, generally this like the street will punish them if their services revenues go over like ten, fifteen percent. Yeah. But they need some professional services to your point in order to get people to use it. That's why they've relied on systems integrators to be s yeah.


Alex Brownstein (01:15:04.812)

And to make it sticky, like it's it's it's forward deployed sales in like, you know, the AI speak.


Brett (01:15:08.767)

Yeah. Yeah.


Rio Longacre (01:15:10.299)

Yep. Four deployed sales, four deployed engineers. Yeah, I mean that's but that this there's there's more of a need for it. And and in fact, like you see companies that you know like OpenAI, Anthropic, they're actually creating f teams of four deployed engineers or partnering. Yeah.


Alex Brownstein (01:15:20.75)

partnering with and creating their own. Yeah. Yeah. I mean like look, I think that that's


I I I I I I do see a future in that, right? And now how successful will the agencies be at doing that? Like I'm not sure 'cause at the end of the day, like media is still not a recurring revenue stream, right? It's just not, right? Like I can't imagine


Brett (01:15:42.23)

Yeah.


Alex Brownstein (01:15:47.577)

Public is creating like the Cloud Max subscription where it's like, hey guys, like you normally spend, you know, ten million dollars or on media, pay us twelve for the year and you get as much as you want, right? Like I just like I don't I like like I just I don't see that happening, right? So


Brett (01:16:01.069)

All you can eat, right?


Yeah.


Alex Brownstein (01:16:08.602)

they're still gonna need they're still dependent on advertising budgets, which are by nature, you know, generally like IO driven. And even if they are commitment driven, they're not contr contracted. Like it's just not recurring. So like then you have these expensive people as forward deployed sales or consultants or what have you driving against non-recurring revenue that is, you know,


Brett (01:16:25.772)

Yeah.


Alex Brownstein (01:16:36.642)

gets a lower valuation because it's less predictable, that margin just becomes challenging.


Brett (01:16:39.234)

Yeah.


Alex Brownstein (01:16:45.376)

Again, it's like what's


Rio Longacre (01:16:45.466)

And Brett remember our conversation with Rashad Tabacawala, he was even saying that for agencies, his advice to he was saying to publicists is your competitors are not just the other hold codes, it's Accenture, Deloitte, the the consulting firms, as well as it's the platforms. You know, and yeah, I would argue the AI companies are their partners as well as their competitors, especially as they deploy more forward deployed engineers and other specialized roles.


Right. And then, you know, I think there's gonna be more of a blurring. Like Accenture I mentioned earlier is getting back into media. I don't know if it'll be successful or not. Deloitte may take another bite at the Apple. it's certainly, you know, if clients are asking you for one thing, they can ask you for another. And like they they've historically they've been turning them down. But what's to stop these these systems integrators and and consulting from I know McKinsey's getting in your for your alma mater's getting more into like digital data stuff. They have a UX team now. They they do a lot of things I never would have imagined ten years ago, where they were like at tier point just


just doing the highest level big ass strategy, right? That's change.


Alex Brownstein (01:17:41.686)

Yeah, I mean again, lots of thoughts on that for a beer. We we should keep it running tab. I think like


Brett (01:17:46.113)

Yep. Yep. And you're in New York City, so you know, when Rio visits we should we should certainly get together. Yeah, totally.


Alex Brownstein (01:17:52.106)

I'm I'm I'm here for it. I'm here for it. I think I think that like


I yes, I a hundred percent agree. And it's also like what are people paying for, right? Is like Accenture being paid to be button pushers and lever pullers, or are they being paid to be a systems integrator and build the systems that enable their clients to do it for them? Now obviously Accenture is likely going to want to continue to do that for a long time and you know there's things there, but I think


it's a different it's a different world. Like it's a different thing that they're selling. And it's a different price point. It's different like hiring model. It's a different, you know, a little bit of everything from the bottom to the top.


Brett (01:18:42.571)

Yeah. Yeah, which is which is that that that logic applies to agencies as well, which is makes it a very complex and hard to solve problem. Right?


Alex Brownstein (01:18:51.276)

I mean it's I mean but the but the end of the day, like agenci we we've been talking about the death of agencies for as long as I can remember, right? Like, you know, they're not they're not going to. Like they've continued to exist. It will evolve, but they'll continue to exist.


Brett (01:18:56.781)

Yeah.


Rio Longacre (01:18:57.912)

Agencies aren't gonna die. Headline, yeah, they're not.


Brett (01:19:02.113)

They will evolve, yes.


Rio Longacre (01:19:03.95)

They're they're cockroaches, as we've c as we've talked about in the show many times, Alex. You they they will survive, they'll transform. They you know, like it's like they'll change, yep. Everyone counts them out. Like every few years, like I remember like you know, being in consulting for many years, or they're like, we're gonna create this in housing team, we're gonna destroy the agencies. No, the agencies just adjusted, right? Which they will to this too.


Alex Brownstein (01:19:10.901)

L Transform


Brett (01:19:24.173)

Yeah.


Alex Brownstein (01:19:25.9)

And in and like the even in housing, right? Like I can count on like one or two maybe two hands, the number of companies that have really successfully in housed at scale.


Brett (01:19:35.435)

Yeah. Yeah.


Rio Longacre (01:19:37.382)

I worked with one client, they in housed everything, then d the second they finished in housing it, they spun ad buying right back out because it just didn't make any sense financially or operationally.


Alex Brownstein (01:19:46.145)

Yeah, it was it was an i it's a it's a fundamentally different thing for which you don't have like the scale or the expertise and it can get expensive unless you're doing a ton of media. And like then I don't even know if it's better if it's a ton of similar media or a ton of very differentiated media, right? Like


Rio Longacre (01:20:04.806)

If it's just digital and it's just in like developed countries, you can maybe make an argument you can do it. But once you start getting tr into like you know traditional media, like how are you gonna buy like out of home like Zimbabwe, right? Or like do a ri terrestrial radio buy in Ethiopia or like, you know, good luck. You have to know someone, you have to know the person to go to. Like this like so that's the thing. Like if it's global like especially for global companies, big ads spend across dozens of countries. I mean, how on earth are they gonna build in house teams to do this? There's no way.


Alex Brownstein (01:20:33.79)

It it's very challenging and it's it's very challenging. And for for me it's fun, but like I'm also I get to watch it, see, learn it and watch it, see it, learn it and then leave. But I think


Brett (01:20:33.953)

Yeah.


Brett (01:20:43.296)

Help companies navigate this challenge.


Yeah.


Rio Longacre (01:20:48.654)

And then there's and then there's paying the platforms like this to for a big company, they don't realize you've got to play platforms everywhere, net thirty, like it's they'll shut you off, right? So agencies to that that's part of what the service they provide is they will make sure those platforms get paid every month, right? And you can pay your your agency net ninety, whatever whatever the you know the big enterprise will take to pay. The agency will just they'll they'll carry the float for that or they'll they'll have a line of credit. That's part of the business model.


Alex Brownstein (01:20:52.482)

Totally.


Brett (01:21:11.393)

Yep.


Yep. As financial institutions in a way. Yeah. It's i in a sense you're sort of leveraging debt. You're indebted to the agency in order to be able to do this stuff with volume and scale that you couldn't Yeah. So it probably changes the economics of your finance department if you do the if you try to in house this. You get to bring in a small army of finance people, a small army of of very hard to do across, you know, very big agency marketing budgets.


Alex Brownstein (01:21:14.082)

Yeah, totally.


Rio Longacre (01:21:29.254)

Yeah, it does.


Alex Brownstein (01:21:34.773)

It's it's yeah, it's not straightforward, right? Like you have to make decisions like what verification platform am I using? Am I trusting the ad server or am I, you know, getting my own third party? Do I have to pay for that? Do I want to pay for it? Is it worthwhile, right? Like, what and that's just even even within like the simple display, right? Like what you know, there's the question of am I getting the best rates or not? Which, you know, is sort of an existential one. There's


Alex Brownstein (01:22:08.204)

You know, all of your financing points are quite true and valid, and like there's so much flow in the ecosystem that's crazy. but


It's it's hard. It's a lot. It's a lot.


Brett (01:22:24.918)

Yep. Yeah. What do you yeah, yeah, so I guess last question, 'cause we are we're we're we we could probably do this for like two hours. Alex, you could you c you could Alex, you could do you Yeah, we can do the Joe Rogan full on. I got a I gotta I got a client deliverable that I have to deliver. But I do wanna ask you, 'cause you mentioned the verification company. So double verify my friend Mark Sigorski. If you're listening, hey Mark. just so


Rio Longacre (01:22:30.778)

We probably could. This is one of those we could keep going.


Alex Brownstein (01:22:31.488)

I mean I I I I I can keep going if you want to.


Rio Longacre (01:22:48.761)

He was on the pod like about a year ago, yep.


Brett (01:22:50.814)

Yeah, Jay it was in the very beginning of of the signal and noise experiment. And they d you know, Nielsen, who I used to also work, so Mark has exited two companies, Exele, and I was part of that acquisition and now double verified for two point one five billion. and then Yes


Rio Longacre (01:23:06.447)

Yeah, Brett, that guy is a smooth operator, man. Like our interview with him, like he we asked him some tough questions. He he knocked everyone out of the park, great answers. He doesn't sound rehearsed even though he like yeah, yeah.


Brett (01:23:13.748)

Yeah. Very authentic even ev yeah. Yeah. Very very auth he would come across very honest and authentic without


Alex Brownstein (01:23:15.661)

my god, I hope that's how you describe me.


Brett (01:23:21.924)

even if he was staying on script. It never sounded in any way scripted and you know, and I think at the end of the day he's he yeah, it's there's a truth to that. He's kind of a he kind of believes in in and strategically lays things out in a very smart way. But but yeah, and then IAS w is is I think have they already been acquired by P E? Or are they or are they in the process of being acquired by PE. Right? Yeah. Yeah. And but also at like that two billion. Where do you where do you see that


Alex Brownstein (01:23:44.046)

I think I think they're owned by NovaCap, right?


Rio Longacre (01:23:46.639)

Yeah, they got acquired.


Brett (01:23:52.064)

that space going and evolving into. That ver out of verification.


Alex Brownstein (01:23:56.523)

Alex Brownstein (01:23:59.959)

Look, I think


Alex Brownstein (01:24:04.423)

I think there's


room for independent measurement. I think independent measurement is important. I think that the


Alex Brownstein (01:24:24.694)

I think that it requires


A


Buy-in from the buyer, right? Why like enough people have to believe that it is bad for people to be creating their own homework, which sounds sort of obvious until, you know, it I mean it sounds obvious, but like but enough people also have to believe that it is worth the incremental cost.


Brett (01:24:46.239)

Yeah.


Alex Brownstein (01:25:02.914)

you know, if like there's probably a point at which it's so expensive that you don't care, right? And that I think is, you know, maybe a theoretical limit, but like one to grapple with. I think agencies


Brett (01:25:14.41)

Yeah.


Alex Brownstein (01:25:23.098)

it sort of behooves agencies to like provide this extra layer of scrutiny and analytics, so that could be a value that agencies provide. but then it's also which tries transparency, but it's also like


Brett (01:25:31.658)

Yeah.


Which is which ties to transparency of like where is this media going? What's the quality of every dollar that I'm sp or every impression that I'm serving across this complex ecosystem? Yeah.


Alex Brownstein (01:25:43.107)

But it also kind of depends on like what is the like where in the value chain does the measurement exist, right? There's the measurement, like I think about it as there's a priori. So like I'm thinking of running a campaign, I wanna understand the demographics or the psychographics or what have you of the people. Like that's measured, right? Then there's like how am I buying? There's a currency, that's a measurement. Then there's like, did it work?


Brett (01:26:06.527)

Yeah, yeah. And Nielsen definitely plays in those two areas. It's like, you know, you know, like reach and frequency across certain demographics, Asian demographic. We can give you that sort of read so you know what you're spending your money towards, right? Go on.


Alex Brownstein (01:26:20.96)

Right. And then there's also like the did it work measurement. Like and then, you know, maybe this is kind of bundled with did it work. So I'm not it's totally me see, but that's like, how are we charging for a measurement? Like, was it a human, right? so I think each of those things are important and can be similar vendors or different vendors. I also then think there's different methodologies of actually calculating the measurement.


Brett (01:26:24.362)

Yeah.


Brett (01:26:34.292)

Yeah.


Brett (01:26:47.2)

Yep.


Alex Brownstein (01:26:47.212)

Right. Is it probabilistic? Is it deterministic? Like is it based off of a panel? Like, you know, all of the all of those sorts of questions that you know we we've all talked about ad nauseum. And then finally, right, there's the does the market agree upon like the unified air quotes correct number that we can point to and think about and draw back on. and I think that that is


Brett (01:27:13.161)

Yeah.


Alex Brownstein (01:27:17.046)

you know, important for types of media buying that is maybe less auction based or at least as


Brett (01:27:23.795)

Yeah. No, 'cause th you make a really good point there because I mean think with linear television you knew that there was a finite universe of total number of ad spots across all sort of news or or television publishers. Think of like the madmen days, right? So you had an idea of how large the space was, how much media was available in the in the in the upfronts, right? And and you auctioned and and paid for those those spots, right? Based on total number of of Yeah.


Rio Longacre (01:27:49.465)

You had confidence you were actually getting what you paid for, Brett. Like like you with the measurement currency you knew like you were you're tar you were paying for this this audience, you're getting it, right? The ad would run.


Brett (01:27:53.514)

Yeah.


Brett (01:27:58.41)

Yep, and and that's what Nielsen sort of set up to reduce friction from the from the ecosystem to allow for that purchase to happen because they trusted they would hit this many people this many times of this age and demo, right? Now but I think with the the the problem with programmatic and auction based media to your point, Alex, is that it's a total black hole. Nobody knows how large the actual reach universe is. And there's arguments out there that that brands are reaching basically inventory. They're they're they're


There's buying more inventory that actually exists with principle based media, right? They're they're they're actually buying shadow inventory that really doesn't exist. So there's there's a there's a gir a a hu yeah? What do you think?


Alex Brownstein (01:28:34.431)

Well


Alex Brownstein (01:28:39.106)

There's I think there's all right, I so two things in our last statement, devil's advocate to something in middle because again, too many questions all at once. Devil's advocate to something in the middle is like, doesn't an auction create a price in a way that like setting a rate card up front like like if the auction is cre if the auction were efficient and the auction were creating a price that like we're mutually agreed upon, then like the upfront currency is


maybe less relevant because like the clearing price is the price. Right. So that's that's thing one. and then like if you're going in an upfront and you're buying it that way, then like you as a buyer are accepting like I'm overpaying for this and part of the overpayment is blocking anybody else from winning. Right? so that's that's thing one. I agree with your statement about inventory.


Brett (01:29:13.353)

Yeah.


Brett (01:29:30.144)

Yeah.


Alex Brownstein (01:29:37.609)

I think that there's a a bunch of nuances there though. Like the first is inventory is basically unlimited on many platforms. Like you can always add another ad slot. Like there was always like, you know, back in I don't know, 2015, 2016, maybe 2017, like Reed Hastings said that Netflix's number one competitor was sleep. Right. Like there's always another place. And this was back before they did advertising, but like they like there's always another place to place an ad and like


Brett (01:30:01.941)

Yeah.


Alex Brownstein (01:30:07.086)

created impression, right? So like I think you are correct in that like the there isn't that like finite amount of inventory that existed in Mad Men or linear television or even in a magazine, right? Like the magazine was printed and it was done. Right. now


Brett (01:30:19.273)

Yeah. Yeah.


Alex Brownstein (01:30:25.656)

I also agree that brands are paying for impressions that are fraudulent or going to bots or, you know, not within their targeted area. And that is a a horse of a different color and a an issue that like, you know, there are many, many firms that are trying to solve for. Part of them are, I guess, measurement or verification firms. and I think that that is a an important thing to think through, especially as like the percentage of traffic on the web goes more and more and more towards AI bots.


Brett (01:30:44.522)

Yeah.


Alex Brownstein (01:30:55.314)

and then finally, I think the last point was about how brands are buying impressions that exceed the value exceed the amount of inventory that's there. Now, my question is that like if we think that of inventory is more or less unlimited, which I kind of do and on many channels, not all but many channels, right? Then what I think you're actually describing is like multiple measurement software claiming



Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page