← Marketecture: Get Smart. Fast.
Marketecture: Get Smart. Fast.
Episode 116: Brian Wieser on why his ad forecasts are heading downwards and what happens if pharma ads are banned
00:54:06
Loading summary
Transcript148 lines
- [00:00]
Brian Weiser
This episode is brought to you by Zeta Global. Do you know what it takes to.
- [00:04]
Eric
Transform marketing into a data driven profit center?
- [00:07]
Brian Weiser
Are you able to align the C.
- [00:09]
Eric
Suite around your AI vision and strategy? Zeta Global has the Playbook to help you get started Download Driving growth in the AI era today at ZetaCMO AI Book again, that's Zeta CMO.
- [00:32]
Ari
This podcast is brought to you by Audiohook, the leading independent audio DSP. Audio Hook has direct publisher integrations into all major podcast and streaming radio platforms, providing 40% more inventory than what could be accessed in omnichannel DSPs. What's more, audiobook has full transcripts on more than 90% of all podcast inventory, enabling advanced contextual targeting and brand suitability. Audio Hook is so confident that in addition to CPM buys, they offer the industry only pay for performance option where brands can scale audio and podcasting with peace of mind knowing they are only paying for outcomes. Visit audiohook.com to learn more. That's audiohook.com welcome to the market podcast. We have an exciting show today. We have Brian Weiser from Madison and Wall. Brian has been on the show. He's an expert in macroeconomics as it relates to advertising. He recently reduced his forecast for advertising in 2025. He thinks the political situation is pretty dire for advertising, so I'm really interested in hearing what he has to say about that and about pharma ads. Eric, are you worried about advertising at a macro environment?
- [01:49]
Eric
You know, I wasn't worried up until last week when I was at a cocktail party with Brian and he was talking crazy. Not crazy in terms of crazy. Like Brian is like one of the smartest people in our industry. But some of the downstream scenarios that he was painting that, you know, are not his base case but like the worst case and talking about them in detail, it was, you know, kind of scary. So suggested he come on and, you know, tell everybody so everybody could be as stressed out as I am right now.
- [02:15]
Ari
Yeah, we, we try to hide in the little advertising world, but advertising is part of the macroeconomic environment and you can't avoid it. So this week Mark Texture debuted another podcast. It's called the Brand Forum and it has got Jeremy, co founder of Mark Tector, along with Josh Palau, who's a well known brand expert. They're talking interesting guests about the concerns and activities of brands and marketers. A little bit different from what you get here where we're deep in the ad tech weeds. You should definitely check it out. I think we're going to give a bonus episode on this feed next week. So you'll be able to listen to it here. But we'd suggest searching on Spotify, Apple or YouTube for the brand forum and subscribing. It's really great interview show. With that, let's dive right in with Brian Weiser. All right, Brian Weiser, thanks for joining us.
- [03:01]
Brian Weiser
Thanks for having me here.
- [03:02]
Ari
You join the club of two time guests.
- [03:06]
Brian Weiser
I can't wait to get the five timer shirt.
- [03:08]
Ari
Yeah, robe, silk robe, things like that. Have a leaderboard. So you've been making news lately. So probably most of our listeners know your background, who you are, but I think you're most well known for doing sort of financial projections and financial understanding of the larger, larger advertising industry. Do you want to give us a little plug for your like what is your company and what do you do?
- [03:30]
Brian Weiser
Yeah. So background, exactly as you said, I'm known for these advertising forecasts. I was an investment banker, securities analyst, did two tours of duty in agency land, a little bit of strategic kind of work. But the forecasts were the public facing things that we did involved in the negotiations as well. So understand the, you know, mechanics of the business to a degree, but also a brief, brief stint in ad tech. Right. With Simul Media for a year. Otherwise, you know, just try to know everything there is to know about the advertising industry. Madison Wall itself though, which I started I guess running up in two years, is a consultancy advisory firm primarily, also with a syndicated research and data product trying to integrate those two things together. And we produce a couple hundred pieces a year right now of written research data underpinning it and then again advisory services that relate to it, to the media, advertising, retail industry.
- [04:24]
Ari
Okay. Madison and Wall is the name so you can sign up for the free newsletter as well. So you were in the news recently because you downgraded the advertising industry. Tell us what you said and what the reaction was.
- [04:37]
Brian Weiser
Well, okay, so when we came into December, let's go back then I would have said that the election outcome was the worst possible scenario for the advertising industry. And I desperately, desperately try to be apolitical. I mean you can tell where my leanings are. But I desperately try to analyze industry in a political lens. And if you just look at what the candidate for on the Republican side said, knew kind of what you're going to get. And I think there is all this willful optimism, especially on Wall street about oh, it's all going to be more rational. Yeah, maybe we don't like the social policies, fine. But the business Community, it's fine. And I'm really trying to look at it primarily through the business lens. And there's no way you could possibly have assumed that this was going to work out well, if you just took them at face value. You knew there were no meaningful constraints. They were going to try to run over any constraint. And we see this January 6, 2021, and so we kind of knew what was going to happen.
- [05:32]
Ari
Let me just ask you, let me stop you for a second. So what were the driving factors in that conclusion that the election was bad for advertising? Was it primarily tariffs or is it more to it?
- [05:43]
Brian Weiser
Yeah, well, I mean, tariffs were the first thing because it seemed almost certainly that we would have seen higher tariffs everywhere, which makes for a less efficient economy, which makes for higher cost of goods sold. Then you layer on top of that deportations. And again, you had to take the campaign seriously in terms of its intentions. Obviously, we're seeing this now, and I think that the massive disruption that this would cause to supply chains would cause a whole other range of problems.
- [06:11]
Ari
How is advertising related to the real economy? Like, is it one to one correlation? Like if the economy goes down 10%, advertising goes down 10%?
- [06:19]
Brian Weiser
It's not that direct, you know, And I think I can say I've studied this more than anyone else alive, that in the United States in particular and every country, it's a little bit different. There are strong correlations on a concurrent basis between economic activity and advertising. But what we've seen in the last two years in particular is a tightening of sentiment and economic activity with advertising. Meaning when I looked at this from the 1990s and 2000s, 2000s, you saw really strong correlations between annual data and advertising activity. Meaning if you looked at the quarters, it didn't tell you anything because budgets tend to get allocated on an annual basis. Right. There are some who would say, oh, advertising leads into. Or an economy leads into, predicts what will happen. Advertising, where advertising predicts what will happen to an economy. No, no, no, no. It's all concurrent. Right. On an annual basis. But what we started to see after February 2022, when are we still allowed to say that Russia invaded Ukraine?
- [07:15]
Ari
We are. I believe it for now.
- [07:18]
Brian Weiser
Phew. Okay, good. So when Russia invaded Ukraine, the way that Mark has reacted played off of what happened in the pandemic, where there was more of a hair trigger kind of reaction. Right. The pandemic messed a lot of things up. Right. From a budgeting perspective. But what it meant was that marketers were really well positioned to do things on more of a hair trigger. And so I have a theory that the correlations probably are now going to be a lot tighter to current quarters rather than years in terms of how budgets might get pulled.
- [07:46]
Ari
It sounds, though, from that description that advertising is a leading indicator that advertising drops faster and more aggressively than the real economy.
- [07:53]
Brian Weiser
Empirically, it doesn't. It doesn't over multiple economic cycles. That's not correct. That could be true going forward. Right, that could be true. But historically that has not been true.
- [08:04]
Ari
Okay, let's get back to my original question. So you started at the election, but now we're in March and you recently downgraded. So what happened this month?
- [08:14]
Brian Weiser
It brought it down even more because it got worse. Right. And so it's not the tariffs on China, which we were anticipating, but Canada and the massive disruption that this threatens for auto industry in particular. And we're seeing this play out by the hour in terms of what's happening with autos. But the fact that the US Would go after Canada and Europe in general is far more disruptive than going after China. On top of that, the policy uncertainty is an even bigger problem because you can't actually plan a business when you don't know what the policies are going to be. It would be one thing if the government said, we're doing 25% tariffs across the board, everywhere. We're just going to suck it up. And that's what the policy is. We think it's the best policy. Okay, we can disagree about that. But that would be better than, well, maybe it'll be 10, maybe it'll be 0, maybe 50. You can't plan businesses that way. And that has an impact on advertising because if you're a manufacturer, say you're in the auto industry, how on earth are you going to plan your advertising budget?
- [09:20]
Ari
Right. So what, so what did you. Let's get to what actually happened. So what do you move the numbers from and what do you move them to?
- [09:27]
Brian Weiser
Well, okay, so it doesn't sound like a big difference now, going down from four and a half percent growth for this year to 3.6. And we were at five and a half prior to the election for 2025. So we're coming down, you know, a fair amount. But here's the thing to keep in mind. Stagflation is a very real scenario, much more real. And for those unfamiliar, stagflation means you have low or no economic growth or even decline and you have inflation. So basically the base case scenario that we're saying Is kind of no growth.
- [09:59]
Ari
Right. Because your percentage growth is not a real number. It's a. Inflation might be 3%, you're growing 3.5%. So there's virtually no growth.
- [10:07]
Brian Weiser
Exactly, exactly. And that's kind of the base case scenario. But as someone put it to me when I explained the scenarios, and this might be what got Eric a little concerned when I talked to him last week.
- [10:19]
Ari
Well, he is a venture capitalist, so he gets very concerned about this.
- [10:22]
Brian Weiser
Tail scenarios are getting quote waggier. Hat tip to Dan Salmon for frame it that way as I characterized it. Yeah, no, the tail's getting waggier. The range of possible scenarios can be really, really bad.
- [10:35]
Ari
How does. Before we go into the tail wagging scenarios, is your model to at a high level to figure out what the GDP is likely to be and then apply that to advertising? Or is it more bottoms up where you talk to people and understand where their budgets are going and figure out a forecast?
- [10:53]
Brian Weiser
It's way more macro. If I've learned one thing from doing this over decades is talking to people doesn't really tell you that much about what will happen or about.
- [11:01]
Ari
That applies to a lot of life.
- [11:02]
Brian Weiser
I mean, it's true. But to be clear, like when I was at Magna and when I was at Group M, there was nothing that I could learn internally that told me what was actually happening to the market as well as actually just studying the data and forming opinions about what the real data told me and then understanding what markers would do separately from what they would say. Those are different things.
- [11:23]
Ari
And what data drives the forecast the most? Is it gdp? Sector gdp? Do you look at the earnings reports of the major holding companies as an input?
- [11:33]
Brian Weiser
Yeah. Well, to be clear, the single most important thing, if we're going to pick one data point for the United States, it would be economic growth and consumer spending in particular. Consumer expenditures are more tightly correlated with advertising than GDP in the United States. Again, different in different countries. But here's the real thing that blows or drives advertising, right? It's the creative destruction of the economy and the replacement. I don't mean the end of J. Walter Thompson. Right. I mean, you know the Schumpeter concept, the idea that new businesses replace old ones. And when you get new businesses that have higher advertising intensity replacing older ones that had lower advertising intensity, like E. Commerce spends a ton more on advertising than a traditional bricks and mortar company does. Right. That alone drives more growth anyways. Those are factors that are underpinned by underlying economic growth.
- [12:20]
Ari
Okay, so you mentioned Magna. They Also reduced their forecast from 4.9 to 4.3%. Do you think that's in line with your. I mean it's not the exact same number, but is their model similar enough to yours that we'd see correlations like that?
- [12:35]
Brian Weiser
Yeah, the structure. So the GroupM and the Magnum models are still based on data that I built and the frameworks and structures are very similar. So the great thing about looking at the data set and they're fantastic people doing both of those products, you get to see different opinions using the same framework. I would argue that only those three of us are producing sort of that best in class kind of advertising forecast model. Everyone else has a different process. Ours is rooted when we're looking at historical data. We are doing financial analysis to be clear. Right. It's not based on economic data from government. Mostly it is based on analyzing public company data and doing real financial analysis. So you know, it's a comparable from a historical perspective.
- [13:15]
Ari
So one theme that's come up is that 2024 may have been worse than we thought. It looked good, but it had political. And there's anecdotally, or actually more than anecdotally, several earnings reports have kind of hinted that the advertising market fell off pretty dramatically after the election just because the lack of spend. Do you have any thoughts on that?
- [13:35]
Brian Weiser
Absolutely.
- [13:35]
Ari
Yeah.
- [13:35]
Brian Weiser
And this is actually an important point. 2024 was ridiculously strong year I'd argue and I think that so many people got used to these unsustainably fast growing markets, um, especially coming out of the pandemic when these are just ridiculous rates of growth that we saw. And there are other factors going into it. Cross border trade is an important issue which we can get back to. But last year I calculate that we were about 8.6% growth for the full year. Ex political to be clear. So we're excluding political advertising. All of our analyses we're talking about. You've got to make sure you're asking ex political. Including political, ex political. 8.6% growth, which is pretty good for a 5% nominal growth kind of economy. Right. But it did slow down in the fourth quarter. The year over year growth rates in the fourth quarter were only about 6% on our estimates. Now we estimate 7, which was strong. 6 still pretty good. I think it's important to note that especially when we think about the trade desk or others where they think they missed to their credit. Far too often companies like to blame external factors when things go wrong. They like to give themselves credit when things go right. Even if the external environment was what drove it right to the trade desk credit. They blame themselves when things went wrong. Right. Good for them. But I think the overall economy, the ad market was probably more of a factor.
- [14:55]
Ari
Yeah. So a hot topic right now is RFK Jr. Who, you know, we'll leave our opinions by the wayside when we talk about him, but he is threatening to ban pharma ads. I have a multi part question. Part one is do you know if he's allowed to do that? Like, would it take an act of Congress? Second question is how bad would it be if that happened? And third question is how. What would the distribution of the damage be by like media types?
- [15:24]
Brian Weiser
Oh yeah, let's go into this one. So going back to the December forecast, we were a little surprised that the likes of an RFK Jr. Would wind up as somebody considered. That's why. Yeah. And it was a. Absolutely a part of the forecast where we assumed back half of the year that there would be some impact on pharmaceutical advertising in particular on television. That hasn't happened yet. There's no signs yet. It is happening. That doesn't mean it can't. But if there's any quote unquote positives from the first couple of months for advertising that is that there haven't been any moves in that direction doesn't mean it can't happen. Now, to the extent that I dug into it, first of all, Pharma is roughly 5% of all advertising.
- [16:05]
Ari
Wow. In the U.S. it's like 0% outside the U.S. right?
- [16:08]
Brian Weiser
Yeah, most countries it's very tiny.
- [16:10]
Ari
5%.
- [16:10]
Brian Weiser
5% of all advertising.
- [16:12]
Ari
Yeah.
- [16:12]
Eric
What percent of cable television?
- [16:14]
Brian Weiser
Exactly where I'm going with this, 10% of television and probably 20% of cable news. And so my thinking back in December, not to get all conspiratorial, but I was not, not that, not that this is what the administration had in mind necessarily.
- [16:31]
Ari
I think they did okay.
- [16:33]
Brian Weiser
And you're, you're the conspirator here. But it was like, okay, what a. When you think about a CNN and an MSNBC who probably have, let's call it 20% operating income margins, you have Fox with about a 40% operating income margin probably. And yes, they had non ad revenues too in both cases. But if you whack 20% of the revenue from CNN and MSNBC and you whack 20% of the revenue from Fox, which one of them comes out best?
- [17:00]
Ari
If it killed the peacock, RFK Jr. Would eat it.
- [17:03]
Eric
All right, the other side of this. So you didn't answer the question, Brian, before I ask another question or make a point, can it happen?
- [17:11]
Brian Weiser
Yeah. So to the extent they've dug into it, they'd have to get pretty creative. I mean I've gone through the policy making process I went through in the late 90s and reading all the documentation I can, it does sound like there are various tactics that they have at their disposal where they can say like okay now there must be farm ads, must now have like 10 minutes of disclaimers if they're going to be on TV. And a counterpoint by the way to the negative view. And this is, well, wait, wouldn't that be great for like TV advertising? Because now they have to buy 10 minute slots. But I think on balance that they could do all sorts of things to tie up the sector. And so you have to assume that it's got a negative spin but you have a different view.
- [17:47]
Eric
Yeah, I mean I've dug into it as well. I've talked to Josh Walsh who is the founder and CEO of Branch Lab, a portfolio company. This is their business believe Ari is an advisor to the company as well. You know where he comes out and he's been super prolific about this I think on online publicly, number one violation of First Amendment. So like this could end up you know, being very quickly like a bigger issue than just like you know, running ads on television. It's like, you know you're violating First Amendment is thing one and then thing two on the other side of it in some of the levers that can be pulled. It's a little bit more about data than anything else. Right. And there's like all sorts of either like national or state level policies that vary that could make it like very difficult if they tend to, if they go in the direction of like state by state. Because state by state the guidelines are very different.
- [18:39]
Brian Weiser
Yeah. And again, I don't claim to be either a First amendment lawyer or pharmaceutical lawyer. I've read counterpoints to the First Amendment defense and so it's far from assured. But the point is it could cause a lot of disruption. And if you're a pharma based advertiser, again like with the tariff issue, the lack of certainty is the problem because you end up in a situation where you start to prioritize other channels and start to do other things because you've got to work on what you can rely on.
- [19:04]
Eric
Yeah. And you know, just this being the year of you can just do things. Let's assume something happens and you know, somebody just does something and TV becomes Difficult. What then happens? Ari's second question, which is, you know, what are the knock on effects? What other channels might benefit, might, you know, kind of grow as a result.
- [19:21]
Brian Weiser
Exactly. And I think digital platforms probably benefit if nothing else. I mean we've seen this with, again, back to the auto industry. It's fascinating how little autos are spending on television now. Like it's maybe 10, 15% of their budget. Really. Used to be, yeah, used to be massively skewed towards television. So in that case with autos, the impact is on the digital platforms first and foremost. But in the case of pharma, unless they're made illegal, can't imagine that happening at all. You can imagine that they would shift their spend. But the bigger issue is from a planning perspective, you get some curtailment of spend and some shift of spend.
- [19:54]
Ari
Yeah, it seems to me that pharma, the most likely avenue of a pharma ad ban would be through the FCC and the broadcasters, which would leave the digital folks in good shape because they wouldn't be subject to any regulations with regard to autos. Just as we're recording this today, on Thursday, Trump announced, I think a 25% tariff on importing autos and it destroyed all the European stocks of the automakers and the Canadians. It really only benefits Tesla because they manufacture in the US So that it could be a really bad year for autos, basically.
- [20:27]
Brian Weiser
Correct. And this is the other thing we haven't really touched on. I mean the knock on effect around the world. Again, as you may be aware, I'm Canadian and I was not aware. Oh well, there you go. I'm Canadian, I'm also British and American. But it's hard to understate the sentiment in, it's hard to overstate rather the sentiment in Canada around this issue. And if you're a business owner of any sort, you are planning around the United States going forward, full stop, there's a decoupling, it's happened, it's happening. And I think the same is going to be true with Europe. So what happens to the digital platforms in this environment? You think they're being hit hard now? Just wait. So that's going to have a real impact in terms of both creating opportunities for domestic players where possible. Right. And I was joking with Eric last week again, I said, hey, if someone wants to give me a billion dollars right now, I'm not actually asking for a billion dollars from anyone, but if I were to do a startup right now, a billion dollars on a super app for the world outside The US would be a huge opportunity right now.
- [21:30]
Ari
You'd be like the Google of Canada or Europe.
- [21:33]
Brian Weiser
These apps exist outside of or Southeast Asia and other places.
- [21:37]
Ari
Yeah. All right. The last topic I wanted to talk about was holding companies and their economics and how they adjusting. I know you cover them probably better than anybody and I'm particularly interested in just how are they kind of moving forward in this uncertain environment. And any indications you're getting about the role of AI in improving their economics?
- [21:59]
Brian Weiser
Yeah, I mean, okay, start by separating what's going on with policy uncertainty and what that means for agencies. We really haven't seen how this is going to impact them. You may have seen S4 reported this week. MC Sachi just reported this morning. I think take that, they're feeling like they're getting hit at this minute with the exclusion of the policy uncertainty issues. Agencies are okay this year. It's a low single digit growth business right now. The more complicated a market, the more complicated a business, the better it is for services far and away. Just you need more people, so that's good for them. I do think that with that said, the international orientation of agencies is also really helpful in this environment. And as long as we don't have tariffs on services and data and watch out for when that starts to happen, agencies are really well positioned because they can provide services cross border and frankly, if you were cross border and you're not an American citizen, you really don't want to be traveling into the US right now.
- [22:59]
Ari
Yeah. I think this is a total side note, but Eric Sofer pointed out that the all that kerfuffle around EU US data transfers that have been going around for 10 years, that might actually be in jeopardy because of some of Trump's actions. So we could be back in the situation where you can't use European data in the United States, which has been an issue that's come up multiple times.
- [23:24]
Brian Weiser
Totally. And I think that we have to beware. Certainly for any early stage company. You have to be thinking in terms of isolating the US from the rest of the world for most of these purposes. So back to agencies then. Rishada Bakawala always said it right. Agencies are cockroaches, not dinosaurs. They will adapt, they evolve, they thrive, they're human, they're organic. AI works in their favor. As long as they continue to invest, that is the biggest risk. If they stop investing in the right kind of people and products and processes, then they die.
- [23:55]
Ari
Are you seeing though any change in terms of the way the agency hold cos Talk about headcount and growth with respect to AI, besides just the platitudes, it's making us more efficient. What I expect to see is that sort of what we've seen with meta over the past year, which is headcount falls off a cliff while revenue stays the same.
- [24:15]
Brian Weiser
I think that's really different though. Agencies were already pretty lean and have to be run lean. If you're a meta, remember your contribution margin if you're a meta is like 90 something percent for every incremental dollar revenue. The incremental profit's massive. For agencies contribution margins closer to 30%. Right. Incremental dollar revenue on a fixed cost 30. You just don't have the latitude to have the extra people as the tech platforms had. So they'd never done that. And so you can't just cut like that. The big question for agencies is can they find ways to take a share of a client's wallet? The wallet essentially doesn't change as fast as technology. So the question for the agencies is are they finding new services to sell through Accenture Globent? Yes. Even though struggling right now Sapient. Those kinds of businesses are ways to capture that wallet that's basically allocated to the non media spend agencies. The traditional holding companies have not done as good a job as they should have on those sorts of services. But the money's there for the agencies to go pursue.
- [25:20]
Ari
Yeah, I mean I think the retail stuff is a very clear area where clients need help. So last topic I wanted to come back to which was what are these tail e tail wagging scenarios? Like what's your, what's your nightmare scenario?
- [25:33]
Brian Weiser
Well, I mean the problem is you can imagine a US that just cuts itself off from the rest of the world, which is the direction it's traveling on a lot of levels. And you can imagine really negative economic circumstances with no real checks and balances. That is kind of the disaster scenario we've all. I mean anyone live in New York kind of knows what Donald Trump's history is. The sequel is not as good as the first one when it came to business. Nor is the third, nor the fourth, nor the fifth, the sixth one or if I'm counting correctly, the Apprentice worked okay, but that was really Mark Burnett's thing.
- [26:08]
Ari
I don't think we need to go through Mr. Trump's long.
- [26:11]
Brian Weiser
No, but the point is that I think there's this optimism that Trump will make choices that are market friendly ultimately because he cares about the stock market. I don't think that's an accurate read. I think it's one read. But the problem is because people are willfully optimistic, they're ignoring meaningful economic damage domestically, which could cause something. I'm not saying global financial crisis happens, but I wouldn't rule it out. And so I think you have to be prepared for really negative economic scenarios. Just a very realistic scenario. When you eliminate banking regulation, when you eliminate all sorts of other government safeguards, you've all these possible consequences that play out that will have business consequences. That's not my base case, to be clear, but I'm saying those downside scenarios are way more realistic than they were.
- [26:59]
Ari
Six months ago, certainly. All right, let's come back with the refresh news of the week. Thanks for this, Brian. This podcast is brought to you by Innovid. Folks, we have our answer. There's been a lot of speculation about what the combination of Flash talking and Innovid would be called. Flashovid Innoflash. We've got our answer at Architecture Live. And it's Innovid. But not just any Innovid. It's a new Innovid innovation. Intelligence independence. The eyes have it. Check out new.innovid.com to learn more about how they're merging their ad tech stacks and shaking up the industry. That's new.innovid.com and as ever, this ad could have been an email.
- [27:43]
Eric
All right, welcome back everybody. We've got the refresh. So on the docket for this week, some crazy things happening in AI. The IAB is making people mad. Bunch of Google YouTube updates and then we got a couple more that we can hopefully slide in at the end. Let's start out with AI. So this week OpenAI introduced 4O image generation for ChatGPT. My question for you, Rembriant, have you messed around with image generation using 4.0 yet?
- [28:13]
Ari
I have not messed around, but I've seen the outputs. My whole Twitter feed is Studio Ghibli versions of stuff. Someone did the Hakatuha girl in Studio Ghibli and it's really adorable.
- [28:24]
Brian Weiser
You know what we're trying to do is put a few hundred thousand words into create text based things with but not video based products. Oh, this is really cool.
- [28:33]
Eric
I did it for you. So one of the most popular prompts right now is taking over the Internet is basically Ghibli me. So I went ahead and found the pictures on the Internet that are most associated with you, Ari and you, Brian, and I Ghibli'd you. And I'm gonna send this to Brian and Ari. You guys can post these as you want. First up is Ari. This is your classic headshot, and I Ghibli do. What's your reaction?
- [28:58]
Ari
I like it. It's, it's flattering. It doesn't have the beard, but my headshot doesn't either. And it gives me more of a full head of hair than in reality. So I'm pretty happy with it.
- [29:07]
Eric
It's pretty aesthetic.
- [29:08]
Ari
It's pretty nice aesthetically.
- [29:10]
Brian Weiser
And then here's Brian. Brian.
- [29:14]
Ari
What?
- [29:14]
Eric
What do you see? Describe it for the audience.
- [29:16]
Brian Weiser
I, I, it looks like I'm in one of the movies, like Where's Hell's Moving Castle? Like, this is, this is great. There should be like some dragons or something flying behind. I'm not sure what it's. It's really good.
- [29:27]
Eric
Yeah. All right, I will send this to you guys and you can use them as your new PFPs should you, should you want on the Internet.
- [29:34]
Ari
You didn't do yourself.
- [29:35]
Eric
Oh, I've done about 50 of myself. I changed my profile pic on X. You should check it out. I used anime rather than Ghibli. All right, so A, this is taking over the Internet, but B, we'll put it in the newsletter in the show notes. The capabilities that OpenAI is showing us with Image Generator is wild. So it's not only, you know, just, you know, put some interesting filters and effects on existing images and have, you know, limitless dials to make them more ridiculous or funny. It's text to image. It is, you know, high quality. If you have a high quality prompt of exactly what you want to see, it will create a photo realistic version of that. It is wild and people are already messing around with it for ads. There's this person on X. I don't know him. Shout out to him. Jacob Postell. He's been making, I don't think he slept. He's been making ads 247 for random brands, for existing brands showing some of the capabilities. It's pretty wild. And I think this is probably like, you know, a step forward of significance in AI being able to be utilized for a lot of the work that creatives do.
- [30:47]
Ari
I agree with you. I've seen some examples online. It seems like a big difference between, like the six fingers with the sort of what people call like the boomer AI vibe imagery. This is much more controlled. It's much more thoughtful in the way it uses surfaces. And I've seen some AI generated ads that are basically product shots with little text. The text is actually English, not garbled. Text. It seems like a pretty big breakthrough for kind of low hanging fruit ad creation.
- [31:14]
Brian Weiser
You know, to the extent I've tried to focus on this topic in terms of its impact on the overall ad market, it's through the products of like Advantage plus, pmax, et cetera from the digital platforms. And we published a piece, it was commissioned by Google. But ultimately it's worked that I, you know, genuinely believe in. I think that these platforms will be ubiquitous and that the, the elimination of friction between creative and media, let alone across media becomes a thing. And it's not just Meta and Google. I think it's. The real interesting question is what happens if you're say a Snap, a Pinterest, let alone a Paramount, an nbcu. You've got to be able to play in this space where you can do both the automated video or other visual asset creation paired with media allocation. This is a five year view I think. But this stuff is so important through that lens of what the structure of the industry will be.
- [32:03]
Ari
I think there could be interesting effects in the open web advertising world. I've written pretty extensively about why I feel like creative as a standalone product is not a great business model. But if the cost of creation goes down and down and down, you might finally have scenarios where media companies can bolt on creative services in a scalable way where they haven't really wanted to in the past.
- [32:27]
Brian Weiser
Totally. And the really important point I've tried to hammer home is like the vision and the dream of AOL Time Warner from early 2000, let alone Viacom when it was created. The combination of whatever cbs, Infinity and MTV networks at the time. If you go back to the press releases or hear the comments of what they were talking about at the time, like none of that was possible then. Right? The idea of like cross Media Buy, whatever. No, it was a bunch of separate things. But all of those possibilities are really here now. And I think that, I mean large brands are not really ready to embrace this stuff for all sorts of reasons. They will, but I think it has all sorts of implications for how the overall industry is structured.
- [33:07]
Ari
Yeah, if you're a media buyer or an advertiser, direct and you go to a sizable publisher you want to work with and you say, well what do you have Creatively it's usually some sort of rich media templated, very high CPM unit that maybe makes up 10% of the revenue for the publisher and they sell it on the homepage, maybe special units. And that is a business that's existed for 20 years and works. But if you come in and say, well, I need some 300 by 250s to run across your site, can you help me with the creative. It's not worth it. They're not going to do it. It's not part of the tech stack and AI may enable it to be part of the tech stack and it could be a, you know, part of something that's self service like oh yeah, here's a URL. Just create your ads for us as needed. It also applies to DSPs. As a former DSP CEO, I got pitched every week by some creative company that's like, hey, just why don't you iframe our UI into your UI so people can create creatives. And the answer was like, nobody wants that. Right? It's like the media and creative are separate. But suddenly if you could create thousands of creatives very quickly and use the data from the dsp, it might break open that use case that has traditionally been frowned upon.
- [34:21]
Eric
Totally agree. Because creative then becomes not something that is like costs. Oh my God, I gotta either devote some dollars, peel off some dollars to get creative made for this campaign or have a lag time. But a actual growth driver because you can use it as a big lever of optimization. So I would imagine this becomes ubiquitous everywhere.
- [34:44]
Ari
And right now the action is in bringing advertisers who aren't traditional into the advertising world through companies mostly in video. Right. People want CTV buyers, they don't have ctv. So AI plays the role. But as it gets better and cheaper, it should expand into performance advertisers and folks who could create their own ads, but maybe don't want to.
- [35:02]
Brian Weiser
Keep in mind for larger advertisers, I think that they'll still spend money on various creative services. It's not that they're gonna stop spending, but it's different functions. The prompt engineer becomes this new function for a creative agency. The lawyers become continually employed and again the account strategy function becomes ever more important.
- [35:21]
Ari
Yeah, it doesn't replace the filming on the salt flats, but you take those assets and you resize it a million times and change it in different ways.
- [35:30]
Eric
Or if you're a small brand, you actually could do this stuff completely using generative AI and you know, build a business on top of it. So it's exciting times. All right, here's the next one, then this one. Brian, you in particular, I'd like to get your POV on. So there was a article in adweek non paywalled shout out to Trishla Oswald. So her last week this is really good. So I'll give you the title. AI agents at agency Jellyfish cut campaign launch times by 65%. So Jellyfish, which is an agency under the brand tech holding company, are they considered a holding company?
- [36:05]
Brian Weiser
Brian? Call them David Jones Might not.
- [36:08]
Eric
Okay.
- [36:08]
Ari
They're clearly a holding company.
- [36:10]
Eric
Yeah, yeah, I think they're holding company. They have a bunch of assets.
- [36:12]
Brian Weiser
They have a portfolio of agencies inside of one entity. It's a holding company.
- [36:15]
Eric
Yeah, so. So what they were talking about was pretty dramatic. So I'll give you just a couple of notes and quotes so we can go from there. So effectively they have an AI agent that mimics the job of a junior level media buyer for every single client that they have. Two quotes. Clients launching a camp new campaign today can have a reporting dashboard set up the same day with AI agents instead of waiting up to 40 days, said Jeff Matzoff, partner at Jellyfish. Quote1 quote2 the agents are replacing the work done by humans. Matz said in house marketers that don't want to have dozens of junior employees on their books now have these in digital employees deployed to do exactly the same thing. That takes teams of 30 or 40 down to a team of four.
- [37:00]
Brian Weiser
Brian, reaction so it's definitely possible to automate a lot of work. I think you end up with a higher probability of a lot of errors depending on what the client is doing or client needs. But I think that at the end of the day what happens is the client budget doesn't fall that much, so the new services end up replacing what would have gone towards humans. Because the marketer is going to say, well wait, what am I getting for this? Like, okay, so it's, is it absolutely the same quality, same predictability, same error avoidance, all that as what we had before? No, maybe not. Let's try it a little bit. And it all becomes very incremental. So maybe you slice off a tenth of your account and you try it for a certain brand or a certain campaign and then you try it again the next time as you figure out how to make it work. And this plays out over a year, two years or whatever. Meanwhile, the budget didn't change. So the agency finds a new way to add a new service. Right. In the meantime. Or it's like, hey, you've got this budget, let me help you with this. Right? That is how I see this stuff playing out. I'm not saying that the thrust of where they're going isn't going to be widespread, but we've always seen automation agencies. We've always seen this ongoing pursuit of driving costs down for the like, for, like, service. That's the life of the agency.
- [38:14]
Ari
Yeah. I just want to know if the AI replaced a junior media buyer. Does the AI enjoy sitting in the box seats at the Knicks game? Does the AI go to the jeans party? How do they fit the jeans on top of the model?
- [38:25]
Brian Weiser
The other thing to keep in mind is I remember when I showed up in agency land knowing nothing about advertising 2003. Within a year or two, I was like, wait, why are all these people sitting in New York? We could be offshoring them in India. And the explanation at the time was, oh, you need the cultural relevance. You need people who understand the context around which stuff works. It's like, okay, now I would argue there was an overestimate of how important that was. But the relationship with the clients matters. The ability to socialize an idea is probably the single most important thing. But you have to keep in mind that even if you reduce the cost through AI, are you better off having people who are a little bit more expensive, but now they're offshored, but they're human? Still a more favorable trail for many brands.
- [39:12]
Eric
Yeah, that makes a lot of sense. I've started a prompt it hasn't generated yet for you, Ari of Ms. Clippy, wearing a pair of designer jeans. The output hasn't. Hasn't landed yet, but I'll send it to you afterwards.
- [39:25]
Ari
It's worth it. That's a good thing.
- [39:27]
Eric
Final point on this. They mentioned the article and this is what you said before, Brian, is a new skill set and a new role that they're talking about at Jellyfish, which is what they're calling a multimodal strategist. So this is a person that clearly can be working across all these tools and using it, trying to figure out exactly how this works across the organization.
- [39:47]
Brian Weiser
Another thing to keep in mind, jellyfish. I think they're in a really good position, but they are a. I would characterize them more as a managed service than a traditional agency. Right. And so to the extent that there's this blurry line between a seller of advertising and a managed service, maybe look a little more like a seller of advertising than an agency. We already know that the sellers of advertising can automate a lot of their work. So keep in mind there's a kind of a different function that a jellyfish has relative to, say, mindshare. Yeah.
- [40:15]
Ari
I think they also have the sort of attacker's advantage. They're not. They're the smallest of their competitors. If you're, if you have a very large position in the market and you start saying you're saving 30, 40% of your costs, your clients hear that and they call you and say, well, let's get lower fees. Whereas if you're the attacker and you're trying to spin those accounts, to you, the lower cost could be a beneficial thing.
- [40:38]
Brian Weiser
Absolutely.
- [40:39]
Eric
Totally. Okay. All right, good conversation. Let's move on. Ari, you put this one in the docket and this is what you said. IAB releases publisher server side thingy and makes a lot of people mad. Let me give you the headline. I'll let you break it down. IB TechLab introduces open Source solution Trusted Server to help publishers regain control of digital advertising amid browser lockdowns. What are they doing?
- [41:03]
Ari
That is a good question. And the Monopoly Report newsletter this week by Alan Chappelle went into it a little bit. We have Tony Katzer on the show, I think not next week, the week after, so we'll ask him directly. So the IAB positioned this as take back control over the browsers, which was kind of an interesting positioning. And the idea, the thesis behind that positioning was that browsers block IP addresses, they block cookies, they sometimes block ads. So instead of having the page, the publisher page with all the stuff in it, render in the browser, you put it server side, meaning in the cloud, and then you just deliver back to the end user a finished webpage with all the ads already there. And so by that token, the browsers would have less ability to do stuff to the page because they wouldn't see all these ad calls going back and forth and stuff like that. It'd be a lot harder to block ads, it'd be a lot harder to block IP addresses, et cetera, et cetera. That's like the reason why the IB positioned it that way. Why did it make people mad? A couple reasons. First, there's this feeling that it's highly duplicative with Prebid, the Prebid server. So the Prebid organization and the IAB Tech Lab are separate. And that has been a little controversial because the IB Tech Lab rejected ownership of pre bid. And some people allege that it was at Google's behest that it was made into a separate org. So they're friendly, but they're kind of rivals in a sense. And Prebid is incredibly widely deployed and already many, if not most publishers do pre bid server side. So one angle of attack here is like, what's the point of this? This is another thing. The other criticism this came from friend of the pod, Michael Sullivan, who's now part of Trade Desk, and he said, more or less, I'm radically simplifying. Like, having it in the browser gives transparencies. We know what's happening. We could see the IDs being shifted around. Go server side. You have no idea what you're getting as an advertiser. There's no trail. And, yeah, I would say that's kind of the current state of things. I personally don't have an opinion yet because I haven't spent enough time on it. So I'm going to spend some time on it before our Tony interview and maybe I'll have a better opinion.
- [43:18]
Eric
What if it's successful? Like, based on your understanding of this, if Trusted Server is successful and is implemented widely by publishers, what happens? What are the downstream effects?
- [43:29]
Ari
A couple of things. The web pages get faster, that's good. The ad blocking goes down. It's much harder to block ads if they're on certain server side. So that's good. Ad cookie syncing becomes harder or doesn't happen as much, and you end up kind of accelerating the transition to alternative IDs, I think, because cookies have to be client side. So if you have less client side, you have more server side, you end up benefiting, like DMPs and companies that have services for publishers that help them collect data. So that's already sort of underway. But as a publisher, you don't have to use any of those things. You could just be using cookies. And if you go server side, you kind of have to invest in some of that infrastructure technology. Those are things that come to mind right away. This is the sort of thing that is very inside baseball. This is like two baseball teams yelling at each other. If you're not in ad tech, if you're just like an ordinary human being listening to this podcast, which I think my wife's doctors are listening to this podcast for some reason, like, it doesn't make sense. You don't care. This doesn't matter.
- [44:29]
Brian Weiser
Can I bring it to the level that I tend to care about and focus on? I understood half of what you said, which is one of the reasons why I always listen to architecture, because it is the singularly best podcast.
- [44:40]
Ari
Half is all we ask for.
- [44:41]
Brian Weiser
Yeah, no, but I. I try to connect this to the money. That's all I focus on. The advertising money that does show does not. In a world where you have essentially no growth or limited growth for open web. And that. This was our view, by the way, before we brought our numbers down in December. Let Alone, the new March forecast. There's very limited growth for open Web in the framework we're looking at for the overall advertising business. If you have a world where expectations from publishers and participants and tech companies in this space are for, let's say high single or low double digit growth, because that's what they got used to. Even if they're wrong, do they stop investing? Do they reduce their investment? Do they need to invest? Is it. I see a shift of share between different tech companies and different sellers of advertising within a low or no growth market. Does any of that impact whether or not these technologies get deployed? That's what I'm asking or thinking about.
- [45:34]
Ari
No, it's a very good point. If you're in an. If you're running a portfolio of different projects and websites and things like that, and some of them are very low growth, you don't want to invest in them unless that's a cost reduction. So I think that switching your entire infrastructure will probably be appealing to those who still make a lot on the web, but for those who don't, probably the investment level is going to go down over time.
- [45:56]
Eric
All right, let's move on. Let's talk about some Google news. So couple things here. So first, Google says news is worthless. So here's what Google did. They did a test in eight EU markets involving 1% of search results to see if revenue would be affected if they just turned off news from search results. And according to them, there was negligible to zero effects on revenue. This is a very creative use of that data for a headline, mind you. But I think it's interesting and this is obviously in response to the lawsuits that Google is facing in Europe.
- [46:34]
Ari
Yeah, so Google is doing the nice thing here. So Meta just turns off the news. They're like, you don't want news, we don't give you news. And as a result, you can't find out news on Facebook in Canada or Australia. Right. Because the laws require payment for clicks. Google is doing the nice thing, which is trying to show the data. That is obvious to any observer who's not trying to extract money from Google, which is that Google doesn't make money on news. They never did. They give away clicks to news sites and the news sites are dependent on them. There's no reason Google should ever be paying the news sites, but the governments want them to. And so there's link taxes and other ways of subsidizing news at the expense of the big tech companies.
- [47:18]
Brian Weiser
Yeah. And I think this also highlights a point I consistently been making for the last 20 years. I've been studying this stuff. Demand is independent of supply for the most part. There's this incorrect perception amongst so many people in the industry and yes, many governments that if you change the supply, it changes demand. And on the margins, absolutely not. The typical advertiser in the United States and around the world has a $10,000 a year budget, $800 a month. They're allocating money based on some kind of rule of thumb. They're splitting it up between say Google or Meta and whatever. Whether or not there's news or not doesn't impact that budget for the smaller advertiser for the largest ones, they also have a totally different process. There's going to be a small number of marketers that are highly performance based, where sure, change of supply does impact pricing and then therefore impacts performance, therefore impacts budget allocation. But it makes no difference to the vast majority of advertisers.
- [48:07]
Eric
Yeah, absolutely. All right, a couple other things. So YouTube, we were having a conversation about the podcast ad Tam and you know what you pay for podcast tools and you know, ratldr is, you know, none of this is really worth it. The sleeper in podcasting is YouTube. So YouTube recently announced that a billion users watched podcast content in a month on YouTube, which is ginormous. Again, I don't know what the consumption is vis a vis Spotify and Apple. But like, you know, YouTube is increasingly just moving into everything that that it can. And they had an announcement that they are testing dynamically inserted host red ads, which is the format of choice for Spotify and Apple when it comes to advertising. So, you know, throughout your podcast listening experience, you could get an ad read by the host. And as ads change, as new advertisers come on for your catalog, these platforms give you the capability to dynamically insert new ads. To date, YouTube hasn't had that capability, which has made it, you know, probably the least attractive platform for podcasters, both video and audio. They seem to be turning this on. This could end up being an interesting catalyst for the market. This could end up being a move by YouTube to bring podcasters onto YouTube because they've got some interesting capabilities. It's probably one to watch if you.
- [49:30]
Ari
Think, yeah, let me jump in here because I'm the adopts professional at architecture. I'm the person who uploads the ads at architecture, unfortunately. So I have experience here. So background is there used to be a product called Google podcasts and they retired it last year and moved it all to YouTube. YouTube creator interface has A separate tab for podcasts where you could import your RSS from another platform. So it works really well. We're starting to see listenership on it. Not as much as Spotify or Apple, but it's a clear number three for our podcast here at Architecture and is going up. The state of podcast content management and advertising is abysmal. There is simply no way to run a podcast business without duplicate work. Uploading things in multiple places, changing ads at different places. We use Spotify tools like Megaphone. And Megaphone does great ad insertion on Spotify. And then the ads aren't dynamic at all. When it goes to YouTube, basically everyone hears the same ad. So the second best thing, the best thing would be if it all worked perfectly, which is never gonna happen. The second best thing would be if each major listening platform had their own ad system and we could upload ads separately. So as a podcaster, I'm pretty excited about this. This is great. And I think YouTube is definitely the one to watch in podcasting.
- [50:47]
Brian Weiser
Yeah, I, I and I think that for years it's been evident that YouTube has underplayed or underestimated its presence as the world's biggest audio platform. Forget podcasting. I just mean in general. And they don't, they haven't really gone to market that way, but I think it's important. Again, there's so many different dimensions in which you can analyze the industry and one of them is to include some chunk of YouTube as part of an audio budget. So there, that makes it, it's a small mark. If you look at podcasting by itself, it is tiny. But if you look at audio broadly, I think it's all one in the five year time horizon. It's one competitive market.
- [51:21]
Ari
Yeah. There are two big advantages of podcasts over audio. One is that it has context. So this podcast is a business podcast about advertising. So advertisers get that, whereas music has no context. And the second thing is it has no royalties. We're not paying any royalties for our music. Obviously we're trying to make money on the podcast, but that's different.
- [51:40]
Brian Weiser
But I'd separate the B2B part from the consumer part, meaning whether it's audio on Spotify, whether it's audio on a radio station or any form of audio. I mean, the fact is that if you're say, Hilton or some other brand that I get hit with when I listen to your podcast, they don't care about whether it's music or whether it's. Clearly they're not talking about B2B when they're getting.
- [52:00]
Ari
But in our case, Hilton is remnant. Right. We sell all of our premium ads to ad tech companies. If you want to be on this pod, please reach out. But when we don't have an ad, we throw in the Spotify network, which gives us like $15 CPMs from places like Hilton. And that's our remnant. So it's a pretty rich market.
- [52:19]
Eric
Hilton's a good advertiser, I would argue, for this audience.
- [52:22]
Ari
Yeah, right.
- [52:22]
Eric
You know, anyway, we have a couple more things where we gotta call it here. I just want to come back to one thing. The question of can an AI agent go to Jean's parties? I can answer that with a yes.
- [52:36]
Ari
Oh, my God. So we're looking at Clippy inside a pair of jeans. They're kind of dad jeans. They're not that stylish.
- [52:43]
Brian Weiser
Yeah, they would never get that from a salesperson.
- [52:46]
Eric
I haven't prompt engineered it, boys.
- [52:48]
Ari
Do you think if you gave some extra GPUs, gave them some credits, they would shift their media plan to your platform?
- [52:57]
Brian Weiser
I don't know.
- [52:58]
Eric
It's a good question. Anyway, we should call it here. This has been super fun. Thank you, Brian.
- [53:02]
Ari
Yeah. Brian, where can our listeners find you?
- [53:05]
Brian Weiser
Well, madisonwall.substack.com is where the free product is a weekly summary product. You get links to our podcast, the Madison Wall Podcast, and also agency business with Olivia Morley. You can get those on Spotify or wherever you get your podcast.
- [53:19]
Ari
Sounds good. All right, everyone, follow Brian Weiser and thank you for being here.
- [53:23]
Eric
Bye bye, everybody.
- [53:24]
Brian Weiser
Thank you for subscribing to Market.
- [53:27]
Eric
New interviews are added every week at.
- [53:29]
Brian Weiser
Marketexture TV and your favorite podcasting app.
- [53:40]
Ari
Thank you for listening to the marketecture podcast. New episodes come out every Friday and an insightful vendor interview is published each Monday. You can subscribe to our library of hundreds of executive interviews at marketecture tv. You can also sign up for free for our weekly newsletter with my original strategic insights on the week's news at News Market tv. And if you're feeling social, we operate a vibrant Slack community that you can apply to join@adtech God.com.