← Marketecture: Get Smart. Fast.
Loading summary
Transcript137 lines
- [00:00]
A
This podcast is brought to you by Playwire. What if the real competitive advantage isn't choosing between human judgment and AI, but knowing when to use each? Playwire is launching a self service platform with rules based control and machine learning optimization. No more one size fits all monetization. At Marketecture Live we'll reveal performance data including how one publisher saw 168% higher CPMs by choosing where to use human oversight versus automation. See the platform@playwire.com that's playwire.com.
- [00:39]
B
A word from our sponsors Imagine if a CMO could unify marketing and advertising on one platform to grow revenue, not waste it. No more silos, just accelerated growth. The Zeta Marketing platform is the only AI powered solution that unites owned and paid channels together in a single system, giving CMOs and CFOs a measurable path to predictable, profitable customers. See how to do it@zetaglobal.com ATG Again, you can see it yourself at zetaglobal.comATG.
- [01:27]
A
Welcome to the Mark Conjecture Podcast coming to you straight from sunny Alexandria, Virginia. This is Ari Paparo and I am here today with two special co hosts. We have Eric Frangi and Alan Chappell and we're here to talk about everyone's favorite subject, the antitrust goings on. Eric and Alan, thank you for being here.
- [01:45]
C
Thanks for having me.
- [01:46]
B
How you doing over there? It's been now a week and a half.
- [01:50]
A
Yeah, I didn't even go home for the weekends. I'm on like 11 days in a hotel room and it's getting a little challenging to maintain insanity at this point. I'm eating multiple meals at a Starbucks every day.
- [02:03]
B
Okay, what's the laundry situation?
- [02:05]
A
Not good. The laundry situation is dire. I would say that's a good word for dire. I, I also the seasons changed. The first week of the trial was summer, the second week of the trial is fall. It's really challenging, but I'm getting through it and my fellow journalists only need to smell me for one more day. And also journalists in general don't have a reputation for good smell, so I feel like I'm fitting in a little bit better. It's like I've become one of the tribe. But anyway, we'll get to that in a moment. Let's. We have a lot to talk about.
- [02:36]
C
You've managed to capture the experience of being like a mid level touring musician. That's what it is. It's hotel room. If you're lucky, a hotel room.
- [02:44]
A
That's true. And I've been Recognized once or twice in the elevator, and that boosts my ego. So it's all worth it. That's what it's there for. Before we get started, we have exciting news. So, as everyone knows, MarketExtralive is coming up in a couple of weeks and we have announced the finalists for the Startup Showcase that Eric is hosting. So, Eric, you want to talk to us about what's going on?
- [03:03]
B
Yeah, absolutely. As we said before, we were inundated with submissions. So first thing, it was just awesome. We had over 60 qualified submissions, so it was really tough to pick five that are going to be presenting on stage and. And duke it out for the. The title of Market Champion. But we have our five and I'm excited because I think it does a good job of kind of like spanning all of the interesting areas of AI that we've been talking about, you know, for God knows how long on the pod. Everything from, you know, companies building, like, agentic workflow. So we've got gg, Eliza and Epiminds. We've got companies that are, like, deep in kind of like enabling the infrastructure.
- [03:45]
C
For all this stuff.
- [03:46]
B
MadConnect. And then we've got companies or a company that's like bleeding edge creative AI, this business called Springboard. So those are the five. It's going to be fun. We're going to make sure that they are presenting, you know, great demos and, you know, some standardization, and we're going to name a winner, so make sure you're there.
- [04:08]
A
We're actually going to name two winners. There'll be an audience prize and the professional prize, People's Choice. That's right, People's Choice. I'm very excited about it. Tickets are pretty scarce at this point. I think we're going to be cutting off new registrations in probably a week or two, so I'd suggest signing up if you plan on coming@MarketLive.com. alan, you're speaking at Market after Live. What? What is it? What are you talking about?
- [04:33]
C
I'm talking about First Party Data, which I think is interesting, given that we're going to be talking a little bit about that today. But the title is First Party Data is bullshit and we'll just leave it at that.
- [04:42]
A
All right. I like you bringing it. Bring the heat on the title. It's all about the title of these things. The challenge is how to have a title for your presentation. Doesn't have too many colons in it like First Party Data colon. The fight for the rights colon. But you beat that rap. I will be Giving an intro and I haven't written it yet, so I'm sure it'll be awesome. But anyway, let's talk about the Google remedies trial, which I've been dedicating the last two weeks to covering. If you're not reading, I'm publishing a daily newsletter at Monopoly Report. Alan is taking a quick break or just letting me drive the car for a couple weeks and then he'll return to, is normally scheduled writing next week. So I'll start because I'm here and I'll, I'll kind of go through what, what I, how I think it's going, and then Alan, you know, feel free to tell me I'm wrong about everything, but. Okay, so let's start with the basics. This is a remedies trial, so they've already been found to be a monopoly. We're not debating Google's behavior, we're debating what to do about Google's behavior. Both sides, the DOJ and the Google, have presented their proposed final remedies, which are long documents which say, here's what I think the judge should rule and, and the debate and all the witnesses are saying that this remedy is right, this remedy is wrong, or Google's wrong, or the plaintiffs are wrong, et cetera, et cetera. So that's the nature of what's going on. There is a mixture of fact witnesses like, you know, folks like Andrew Casale and others in the industry, and there's also expert witnesses who, you know, if there's a greater misnomer than expert witness, I don't know what it is. But you know, they, they come up there with their PhDs and give opinions about stuff they don't understand and keeps happening. So the, to summarize, and this is super high level, I mean, you should read these things. Google wants mostly behavioral remedies. Google's proposed remedies are effectively to stop doing the bad stuff they were doing, to stop doing first look, last look, ubr. And all those things that you may or may not know what they are, that gave them advantages in the, in the ad exchange and probably most relevantly, to start bidding into pre bid, which means from a publisher's perspective, if you have addicts going into pre bid, instead of going directly into your ad server, you can switch ad servers. Right now you really can't switch ad servers. And so that's probably the biggest give that Google's offering. And I would say that most of the DOJ witnesses have poked holes left, right, center in those remedies that they're partial, they won't work, etc. They're not sufficient, blah, blah, blah. Then the DOJ's remedies are really comprehensive, long and complicated, but they boil down to more aggressive behavioral remedies. Like they also want pre bid. They want Google to not be able to differentiate different supplies, different, you know, supply sources based on who owns the ad server. But probably the headline is that they want to spin out Adex into a separate company. They want the, what they call the final auction logic of DFP to be open sourced. So there'll be a new open source project that's like the ad server logic and that would be able to be used by competitors. But also potentially you'd have like a custom ad server logic. You could deploy yourself in the cloud somehow, it's not clear how. And then they would have the right within three years to demand that DFP itself get spun out. Also, although it feels like their heart's not really in that last remedy, they're sort of like, that's like just there for the sake of completeness and no one really wants to do that. So that's the situation and we've heard a lot of testimony that all the options are bad, basically and the judge has to make sense of this and it's really complicated and really technical. So I'll leave it. This podcast is brought to you by Adelaide. Media verification and measurement are undergoing major disruption. Legacy players are pivoting to performance. Advertising AI is reshaping ground safety and attention is replacing viewability. Adelaide is leading the shift with au, a new way to assess media quality that scores placements based on their potential to drive attention and outcomes. Before your ads run, think of it like a credit score for media. Finally, a clear view of quality. Before you buy, take the guesswork out of your investment strategy and try Adelaide AU on your next campaign there, Eric, go for it.
- [09:12]
B
Yeah, I just like prepping the docket for, for the show. Like a half hour ago there was an article on Bloomberg that had, presumably you were in the room for this that said Google actually explored the sale of ADEX in 2020, had identified potential buyers. The list of potential buyers is actually somewhat humorous because they are not the buyer set of 2025. It's like Oracle, Salesforce, so on and so far, so and so forth. But it's, it's interesting that this is, you know, coming up now.
- [09:45]
A
Yeah. So a couple of things have been revealed. There was a project called Project Sunday which was effectively to sell adex and to possibly turn DFP into a cloud service. So basically anyone could run their own dfp. And there was also something called Project Monday. I forget what that was. But they were sort of tied together, it definitely seems, and the timing is important. 2020, that was before most of these cases were filed. So they were starting to feel the heat from regulations and governments, but they weren't yet in the crosshairs. And they hired Lazard, the investment bank, to create a deck. We saw a couple slides of the deck in court today and they didn't do it. I don't know why they didn't do it. They probably should have done it. And then we also found out that in 2024, right before the trial started last year, they had a pretty big undertaking to estimate how much it would take to divest AddX again. And we think that was in response to regulators in some sort of settlement talks that they, they were probably Europe. I think there's been some back and forth with the court on that. So Addicts has been sort of in the crosshairs as a not vitally important part of Google and a pain in the ass for a long time.
- [11:06]
C
Yeah, I mean, I think that this is one of the big takeaways is like I don't think Google can really, you know, straight face now argue that divestment of ADEX is impossible. It's clearly possible. Now they can still continue to play around with, you know. Well, yeah, but it's still really complex. But the second thing that maybe people aren't are talking about now is like, isn't that sort of an admission that maybe Google isn't going to be in the network business in a couple of years? I mean, they've clearly thought about this and isn't it maybe just a matter of, you know, when they decide to ultimately pull the trigger? And that has almost nothing to do with this particular exercise.
- [11:44]
A
I think a lot of Google's arguments in the case so far, a lot of their effort, has been around saying that the ADEX spin out's impossible. Like it's literally technically impossible is what they're saying. And then they, and then the evidence comes out that they considered it multiple times. So I feel as though that line of argument is not very productive and disingenuous. But let's talk about the network business. So it's not new news to listeners of this podcast that the network business has been stagnant for multiple years while YouTube has been skyrocketing. We got a little more detail about it through various exhibits that have been shown in this Case that I published, I think on Wednesday's, Wednesday on Tuesday or I don't even remember what day it is, I published this, a bunch of charts and what the chart showed was that AdWords is largely just buying YouTube and that DV360 is largely just buying YouTube and Display and App are both tiny compared to YouTube in both of those platforms. And that's the buy side which is kind of the more strategically important part of the business. All right, on the sell side we know that adex is like 70 something percent of adex is just google adwords buying adx. So adex is exposure to the open web outside of Google or to other demand is pretty small too. So these are all kind of like loss leaders so that they could get advertisers to spend more money on YouTube it seems, I don't know what that says for like outside businesses. What is, on the one hand you say, well how are Magna and Pubmatic so big if they're these tiny, tiny market share slices of the market that's declining, I mean obviously they're putting their money on video. But on the other hand maybe the business, the market's so big that they have all these like public companies and successful companies that are in single digit market share.
- [13:45]
B
That's probably the case. You know, they're innovating, they're like attacking, you know, this market. They're trying to, you know, kind of keep the open web sustainable. And Google is I think properly just like focused on their own core assets, their O and O, which you know are far more profitable and they, they have the control. So I think it makes sense. What do you think, Al?
- [14:07]
C
Yeah, I've been saying for five years that the, the hidden thing here is that Google's just, Google's going to jettison the network business as soon as they can. And I, I do think that is starting to bear fruit or that that is starting to come to some fruition.
- [14:20]
A
You say Google's going to jettison but like why don't they do it already? Like why are they fighting this in court? Alan? So what I've heard is sort of a version of if they give up, if they give in or give up, it'll be the deluge. Like everyone and their brother will hire an attorney and come after them. You're a lawyer. So what do you think of that from a, from a strategic perspective?
- [14:42]
C
Oh, completely, yeah. They want to get the liability off the books as best they can and that that may be an unattainable goal but they want to get as close to that as they, as they absolutely can. And they want to keep their, their chits for as long as they can because they've got multiple jurisdictions who are going to want to have their say as, as, as this and the search case, you know, start to get closer to the finish line.
- [15:06]
A
So, but what happens? Let's, let's just follow through on that question. Suppose they get a very beneficial remedies here where they get a judge meta like remedy where they're let off pretty much scot free, which I think is possible. The how does that help them? They're still getting sued by everyone and their brother.
- [15:23]
C
Well, firstly, I agree with you. I actually think Judge Brinkham has, has thrown out a couple of tells that indicate that divestment is not, is not the slam dunk that a whole bunch of people seem to think it is. You know, the whole discussion around what it is to be a recidivist monopolist, you know, just because you've been adjudicated as a monopolist three times is not the same thing as a recidivist monopolist. And the clear indication there is that Google needs to be given a chance to screw everybody over with respect to behavioral remedies. I mean that's sort of the net.
- [15:58]
A
It does feel like the case, right?
- [16:00]
C
Yeah.
- [16:01]
A
So the judge has said a bunch of times and the judge doesn't talk that often. So the judge is usually just, you know, coordinating the witnesses and whatnot. So when she talks, you really want to pay attention to what she says. And, and like I think two or three times she had said something along the lines of isn't this a matter of trust? If there's a court order, yeah. Won't they follow it? And people like me and Jason Kent and Ariel in the audience, like, we roll our eyes so much we almost get brain damage because like no, you cannot. I mean I, and I love the people at Google, but you can't trust them at all. Right? They like not even a little bit, but, but the judge doesn't seem to feel that way.
- [16:40]
C
Well, and this is where I think the DOJ may be dropping the ball a little bit. And I think they did in the search case. Google was adjudicated. As with all those evidence spoilation issues, that goes directly to the issue of trust. And if I were the doj, I would be literally holding that in Judge Bergman's face constantly. They didn't do it in the search case and I think that came back to bite them. I really would hope that they're going to do it here if they want to have any chance at a divestment.
- [17:10]
A
As an option to explain evidence spoilization. Spoilation.
- [17:13]
C
Well, okay, so there was a former, There was a formal policy within Google to push almost any discussion on, onto a certain type of chat mechanism and that that chat mechanism was set to shut off in what, 48 hours, which rendered the ability to figure out what was really going on over there over the last ten years nearly impossible.
- [17:37]
B
Corporate Snapchat, kinda.
- [17:41]
A
That's a good headline. Yeah. There was a whole thing where the guy who created the policy, Kent Walker, there are people trying to disbar him in California and he was actually at the trial the first day. He was sitting there all day. I think Jason Kent wanted to jump across the pew and strangle him.
- [17:57]
C
Both times you had the judge saying, oh my goodness, I have never seen such a blatant abuse of the attorney client privilege. So where, where is that now? Like, you know, that was a year ago. You have to make sure that the judge is thinking about that as they're, as they're trying to address core issues around whether you can trust Google.
- [18:17]
A
So, Alan, I sort of agree with you. I also think that's sort of a legalism, a legal technicality more than operational. And what I would love to see what I think the mistake here is, and you could defend yourself after this, but I think the mistake here is, is that the DOJ didn't bring into evidence a lot of the documented abuses that the state case has. And those names that you probably have heard of, like Bernanke and Poirot and Dynamic Rev Share and all the many, many ways that Google cheated over the years. Because that's really relevant here, because what Google is saying is if we just, you know, take away the tie between the Ad Server and Ad Exchange, that's good. And obviously we couldn't cheat after that. It would be impossible. Right. And the judge seems to be buying that and saying like, okay, like as long as we like have a court order, they'll be good boys. And we just know that, like it is just too easy to come up with some mechanism that may cause cheating or just a slanted marketplace, even if it's not overt cheating. And you know what was really interesting was Rajiv goal, the CEO of PubMatic gave his testimony yesterday. If you didn't read my newsletter, what he said was, right now I can't get access to inventory from one of my publishers. And because there's some bug that Google refuses to fix, that makes DV360 unable to bid through my path even though it works on all the other paths and I'm just sitting here not making any money and it's been happening for eight months. That's a great example because it may not be malicious but it has the same effect.
- [20:06]
C
All I could think of during that exchange was like how I feel when I'm dealing with my, with my isp. Like yeah, we'll get to that, we'll get to that in a couple, in a couple of weeks. And like I don't know, I don't know how you, I don't know how. I don't know how you quantify that from a perspective of antitrust law. And to your earlier question, I want to let the DOJ off the hook a little bit. There's, there's resource constraints here. There's only so many issues you can raise. You could certainly argue that they should have raised a lot more of the buy side issues because heck, even if you get what you want here and you know, even if you did divest DFP and Addicts, I don't know if that really addresses the, the, the, the core issue which is tying of the buy side.
- [20:52]
A
So Addicts Addict. I wrote this yesterday but the ADEX spin out feels like it's the big issue and a lot of the other issues. I wrote that I think this open sourcing of the ad server logic has been debunked effectively in court. I think that the DOJ has lost that issue. It's really complicated. Everyone said it's complicated. The Google witnesses who went into technical depth that was believable about why it was so complicated. I can't imagine she, she accepts that remedy. But the ADEX spin out is the big one that's still remaining where if you don't trust Google then Addict spin out is the right remedy. If you believe a court order behavioral can can rein them in and allow other people to compete then it. Then you don't need to spin out. That's kind of where it's at the I guess like what are you guys thoughts like Eric, like what do you think the market looks like to have a, another exchange out there that's maybe private equity owned? Like have you been kind of thinking about that at all?
- [21:58]
B
Yeah, they have to compete against I mean a much stronger index. Like that was the first thing that came to my mind. If you were to start an ad exchange today, the competitive set is not what it was right. Like 10 years ago or 15 years ago during the heyday of ADEX index is much further ahead. I don't know if there's others that come to mind to you that you know, are just like innovating, have like a ton of publishers, have all the relationships on the buy side. You have to deal with a TTV that is, you know, I mean, focused purely on, you know, the, the shortest path to inventory. So to me it's, it seems like this would be a, a pretty daunting task, quite frankly. Like what? Yeah, what do you have if not for the direct access to O and O Google inventory?
- [22:52]
A
Right. And the AdWords demand. Right, right. So one of the more interesting things that, it was kind of funny in a sense where one of the buy side Google witnesses, an engineer or product manager works on the buy side. He gave through the list of things that would get worse if there was an attic spin out. And it was funny because they're exactly all the challenges of running a DSP for anybody else besides Google, which are discrepancies and cookie matching. And he was, he was, he was putting these bullets out there like it was some brand new idea like, oh, our data won't match and there might be discrepancies and latency will be increased. And it's like, okay, welcome to the free market, guys.
- [23:32]
B
The business challenges are what, you know, is important here. And I think it would be quite challenging.
- [23:38]
A
Yeah. The other, the other point which I think that guy made as well as an economist who went on today, was that even you could say it's good or bad, this tie between Google Ads and adx, but if you break it, probably demand will go down. Like if it gets more expensive to buy, then they'll buy less. And one of the mysteries, and I've actually spoken to some former Google employees about this question, this mystery, I always like to say that, you know, Google AdWords subsidizes AdEx. Like Google AdWords only buys through AdX and pays them 20%. So AdX only exists because of Google AdWords. But some former Google people have told me, I have it backwards, they say that Google Ads would never be able to buy display ads at scale if they didn't control the exchange too, because it's Google Ads needs certain signals and needs extra latency and needs all this hand holding to bid externally that only they get through adex. And if they, that's the reason they don't buy from other exchanges, that, that, that whole business, that whole demand source, which as I said earlier is something like 70% of ADEX would decline precipitously if Ad X was not part of.
- [24:54]
B
Google, have they ever tried to have a conversation with Index? Have they?
- [24:57]
A
Oh, no, they tried to test. They do. About 10% of AD of Google Ads bids through other exchanges like Index. It already exists. They've just never scaled it. They always say it's. They don't have to or maybe they can't or nobody really knows. That was one of the mysteries of the last trial that they didn't answer. So, Alan, let's change subjects for a second. So one of the interesting things about being courthouse is you talk to all the other antitrust reporters and people who follow and lawyers and stuff. And so I was asking about the other antitrust cases and I got here, here's what I heard. And you tell me what you think because you follow stuff more than I do. Facebook's a loser. Like, the government's going to lose straight up. And search is going to be appeal to all hell where like maybe even both sides appeal and it's not going to end anytime soon. What's your reaction to that?
- [25:49]
C
I agree wholeheartedly with both of those. I would add one other data point though, if this is any interest at all.
- [25:56]
A
Yeah.
- [25:56]
C
Keep in mind, in an antitrust suit, if the government settles that settlement, the terms of the settlement are reviewable pursuant to a law called the Tunney Act.
- [26:07]
A
Okay.
- [26:09]
C
Other settlements between the government and a company say the government suing the company for $25 million on some oddly trumped up charge, and the company pays, and everybody scratches their heads and wonders why the company pays. That settlement is not reviewable.
- [26:27]
A
Okay, let's talk about what Alan's talking about. So this morning, President Trump tweeted an AI image of him being given a very large, like, oversized novelty check by a character who looks suspiciously like Neil mohan with the YouTube logo and the check for $25 million. And as if my life hadn't gotten surreal enough, like, this stuff is just kind of swirling around. So Google basically just paid off Trump $25 million for, for removing his account after he, he tried to have a coup against the United States in January 6th. And. And Alan, you're saying that's a little bit untoward?
- [27:09]
C
Well, I know I'm saying directly that's untoward, but I'm also saying that that transaction is not something that's subject to the typical review. And so one could imagine a scenario where perhaps some of this antitrust cases were dropped or more loose or they decided not to appeal as a quid pro quo for a settlement in the other room. That is a subject to review.
- [27:38]
A
Hanging out in D.C. i'm just gonna end it with this. Cause I'm covering this sort of Hunter S. Thompson style. It's more of a lifestyle than journalism. But I'll tell you this. I'm walking around D.C. there are troops with machine guns guarding the metros. The government got shut down today. So there was talk about just postponing the trial, but they decided not to. The Jim Comey was indicted in the building I'm doing work in. And there were cameras outside waiting for Mr. Comey to come in in handcuffs. That didn't happen. Apparently it happened somewhere else. And then the. The head of the Eastern District of Virginia resigned during this case. And the beauty pageant winner who replaced him showed up in court today looking pretty fine, I will say, but you know, that's. That's what you get for that kind of thing. It's a little strange.
- [28:31]
C
What I'm hearing is Ari is making my point for me.
- [28:36]
A
It's just, you know, I didn't know I lived in a third world country. But, you know, justice will be served. All right, let's on that. Let's call it.
- [28:43]
C
Wait, wait, no, actually you said something before we came on that I thought was super interesting. Do you want to talk about Google's witness today?
- [28:51]
A
Okay, let's talk about Google's witness today. I forgot her name. I don't have it written down somewhere. So Google had only one witness in their witness list that was an actual customer and is a publisher, the CEO of a company called WikiHow. And I was very skeptical that this was gonna be meaningful because I just thought it would be a bunch of platitudes about how Google's a great partner, blah, blah blah, etc. I even showed up late, I didn't catch her intro. And I think it may have won the case for Google because this witness was just straight out of central casting as a mid size publisher who just wants to survive in an era of AI. And the stuff we're talking about in court is at best irrelevant to her and, and at worst a hazard to the future of her business. And she said as much. She was asked even. It didn't matter if she was being interviewed by Google or by the DOJ on cross. She would say things like, I don't care what take rate ADEX has. If it's 20%, it doesn't matter because the bids are net to me, which I think a lot of publishers listening to this understand. But no one in the court had Ever heard that sentiment before, that the price didn't matter? She said she was asked like, would an ADX spin out help you? And she said, no, of course not. I love my ADX account manager. Why would they switch that? I don't have time to do a new integration. I don't want anything to do with that. She was asked if she wanted to switch ad servers. She said, no, Google's great. I love Google. She was asked like, you know, she's one of the people on the free program. She's asked, well, you know, would you pay? What would happen if you didn't have free ad serving? She said, I'd probably go out of business. Right. The judge heard this stuff and every single one of them was a body blow to the DOJ's proposal. And maybe the judge, the judge has said previously in the case she was much more interested in the experts than the fact witnesses. But this I think absolutely delivered the judge a perspective that she wasn't getting from the more advanced like Steph Laser and you know, Daily Mail type people that had previously testified. And I think it put a dagger in, in the more aggressive remedies.
- [31:04]
B
Wow.
- [31:05]
C
Yeah, right.
- [31:06]
B
Alan, what do you have to say this?
- [31:08]
C
I. Google is brilliant and this is why they usually win. I'm almost wondering if they just like, why even bother have tomorrow. They should just say nothing. That's it, we're done. Get out. Like the George Costanza thing, just, you know, get out on a high note.
- [31:22]
B
Yeah. WikiHow is, according to their website, the most trusted how to site on the Internet. And I'm sure it's a great site. I'm sure it's got like awesome instructions. I'm sure it's got a large user base. But the most ironic thing is that with all of the AI overviews that Google is now automating with Chrome, Google is, with AI, an even bigger threat to that business.
- [31:47]
A
I'm going to read you a quote from my notes here. I really want to spend my time on all the problems we're having with AI and, and don't want to worry about ads right now.
- [31:58]
C
So is the DOJ going to have the opportunity to cross this person or have they already done that?
- [32:03]
A
They really did. And it got worse when they crossed her.
- [32:07]
C
Did the term overviews come up?
- [32:10]
A
Yeah, you know, they brought it up and she said it was a really big problem, full stop. I mean, what else do you want to say? Like, you know, making this, this argument that because Google has search and AI and was adjudicated A monopolist in in search of therefore the ad tech monopoly, blah, blah, blah, is way too complicated. I don't even understand it. Like, yeah, like shit happens. AI is happening regardless of how much ad serving costs. So I'm going to write it up tonight. Hopefully it'll get out tonight. I'm pretty exhausted. This is Wednesday night we're recording this. You'll definitely have it by Friday. All right. With that said, maybe we'll do a follow up where once everything is done we can predict the outcomes. But for now it's a work in progress. It's a good conversation. We're going to come back with a ton of guess what, AI news, trade desk news, etc. Etc. Let's do that.
- [33:05]
B
Hey everybody, we are back with the refresh. So we're recording this on Wednesday, October 1st. Welcome to Q1, ladies and gentlemen. So some of this news here is going to be as of midweek, there's rumors that there's more news coming out. Ad tech God was saying something about more news coming from ttv. So I don't know if we're going to get it tomorrow or Friday, but there's plenty to talk about today. So we have a bunch of stuff from OpenAI, a bunch of stuff from TTV, Amazon continuing to go directly for all the supply and maybe one or two more things if we got some time. Let's talk about OpenAI. So last week, end of week again, we kind of missed it with the, with the timelines. There was two interesting things that hit the X conversation. So first was rumors of ahead of monetization. Being interviewed by Fiji Simo, the head of applications at OpenAI. And you know, basically this is a person that would own the ads business that we all know is coming. And then number two, there was a job description. I think adweek saw that they were hiring for an in house growth marketing team. You do that because you believe in paid advertising. You do that because you want to start building your own. And presumably it could be part of a bigger push towards advertising. So as much as you know, Sam Altman is pushing back on this notion like it couldn't be clearer that ads are coming to open AI. What do you guys think of these two things? And are you connecting these dots like I am?
- [34:40]
A
Yeah, I think that Sam Altman is a smart guy and I don't, I don't put a lot of value into the things he says publicly about the stuff because he doesn't want to tip his hat. And you know, it's pretty clear that AI monetization will be somewhat paid and somewhat free and the free will be ad supported. I don't, I haven't heard any alternative points of view on that. Whether it's ad supported or commerce supported, they're kind of two sides of the same coin. And I think he's in, he's in execution mode. I don't think it's a coincidence that at this very time they're also running TV ads. Have you seen any of the TV ads for OpenAI that have been running?
- [35:18]
B
They're really nice and out of home ads and they are brand, pure brands. It's fast.
- [35:24]
A
They're really good.
- [35:25]
B
Yeah.
- [35:25]
A
Really good ads. And you know, that's not a coincidence. They're, they're looking at capturing share and monetizing it and, and getting that flywheel going so that they become either the number one or the number two. Because there's not really room for three or four and five, I don't think.
- [35:42]
C
Yeah, and there's the old, you know, pay very little attention to what they say and pay all your attention to what they do.
- [35:49]
B
Yeah, well said. And what are they doing? So they're shipping this week. Nod to what Ari just said. Ads, commerce. It's, you know, probably the same damn thing. They had an announcement with Stripe. So Stripe is powering instant checkout in ChatGPT and they released an agentic commerce protocol co developed with OpenAI. So you will be able to have products from first Etsy merchants and soon Shopify merchants be recommended and then via Stripe have the ability to have instant checkout. So A, you're right. Right. Maybe the beginning of this is commerce. B, it's interesting because it's starting with the long tail. Right. Like, you know what I do, I'm sure you guys do the same. You look for recommendations on sometimes like really niche things. All of a sudden you're able to buy that niche thing like right away and presumably the best thing. Right. So, you know, kind of smart. Right. So it's number one, kind of like empowering the long tail, which is exactly what Metta and Google have done to build their ad business. And then on the other side of it, Eric Schiffer, you know, said this, this is, I mean, the data that they're going to be collecting as part of this to fuel what will be ultimately another advertising business is pretty interesting. So that was announcement number one this week that I thought was super impressive.
- [37:14]
A
Yeah, I just want to throw a quick contrast which was that perplexity seemed to be going for the Head where when Taz Patel spoke at Market Live last spring, he really talked about a couple of tests at a couple of segments that had sufficient search volume. So I don't know what they were, maybe auto or something like that. And this is a little bit more data driven, a little more Facebook ish. And we know that Perplexity is currently revising their ad plans. They're not pursuing them as strongly, so this will be super interesting to watch.
- [37:47]
B
Agreed. And then on the other side, they released this Sora app. Did you get access to Sora, either of you guys?
- [37:58]
C
I've heard it's just breathtaking.
- [38:00]
B
Yeah, same thing. I saw the video. The video was well done. Incredible. Oren Hoffman, if you're listening, if you really do have invites, you know where to send them. You know my email address. Anyway, so it's this like, incredible. It's, it's like a TikTok, but just purely AI AI videos, you know, in a feed. It's powered by Sora 2. It looks awesome. What's going to become of this? I have no idea. But this one's going to be pretty interesting to watch too. And imagine, you know, if it takes off the AI, the ad AI capabilities.
- [38:35]
A
Yeah. I think it's also contrasted pretty well with Meta's announcement last week of Vibes. They're, they're sort of kind of what's called the slop feed, some people called it degradingly, doesn't look as good. And also I think Sora too, if I understand it correctly, is more creator friendly because it's allowing people to kind of insert themselves, their human selves inside of AI creation, which is very appropriate if you want creativity and things like that, as opposed to, you know, orienting yourself to a, you know, a video editor mindset. This is more of a creator mindset, which seems to be going the right direction. I think they're all both so early that, like, it's too early to make predictions at all. But we're a long way from just a year ago when these video tools started coming out and we were saying, well, oh, it's cool, but it's, you know, it has six fingers on every hand and you can't make it look like a movie unless you're a professional, know how to write the prompts perfectly. We're already past that and, and it's been amazing to watch.
- [39:37]
C
So let me, let me just throw the lawyer cold water on it a little bit. I mean, wasn't it just this week, wasn't it just this week that there Is this viral video of, of somebody who looked a lot like Sam Altman stealing from a target. Right. And so this is going to usher in a whole new era of who do you trust?
- [40:00]
A
I wouldn't put that past Sam Altman though. Got to get baby the diapers. But there was also the opt out thing. Alan, did you see that? They said, well, if content creators want to opt out of using the data in Sora 2, they can but not opt in.
- [40:16]
C
Yeah. Okay, so on the one hand, that's completely insane, but we are now at the point where governments all around the world are almost collectively saying, ho, ho, ho, wait, wait, we need to, we need to beat the Chinese. So we need to relax our approach to AI So in light of that, that their position here is significantly less insane. Now, whether that ultimately means they get sued by Disney or whomever, I do not know. I'm still waiting for Disney to sue somebody other than, I guess, Mid Journey because those are like the titans. You need the R. RIAA types, you know, to come in and get, get real litigation heavy with you.
- [40:57]
A
But you know, we haven't heard much from Mid Journey in a while. Do you think they're being held back by the lawsuits? Is that why they're being leapfrogged? It's a good question. We probably don't know.
- [41:07]
C
I don't know. But my point here is that, you know, they're pushing the limit here, but, but where we are on the pendulum is pretty far anti content creator right now. And so they might be pushing it a little further, but you know, that's the environment that we're living in. I'm not here saying it's a good thing, but. Yeah, but it might be a sound business strategy.
- [41:35]
B
Yeah, that makes sense. As an attorney, do you just see a wave of problems coming from like every angle that are just gonna pack the courts? Basically?
- [41:48]
C
Yeah, yeah. We're just starting to get into the, the copyright cases and there's only been a couple that have even made it, you know, made it to the point where it's sort of worth even talking about. So you're gonna see a ton of that stuff happen over the, over the next couple of years. By the way. That's assuming that like tier two and tier three copy, like content creators exist in two or three years. And that's sort of the, that that's, you know. Well, I don't know how you make, you know, again, I know on this podcast we've had a whole bunch of, you know. Well, when are publishers really going to Go extinct type conversations. And like, I think there's a similar one just with content creators. There's just no real incentive. And you know, there's great for guys like me who at this point in my career I'm a hobbyist, but, but you know, for somebody who needs to make a living, it's just going to be really difficult.
- [42:36]
B
Absolutely. Okay, we'll end on that note. Let's talk about ttd, TTD shipping too. I mean, good stuff, right? So first was they announced a major overhaul of their digital advertising data marketplace. So third party data, this is called Now Audience Unlimited and uses AI to find the right segments and leverage presumably lower cost for, for brands to use third party data. Sounds like there's kind of two ways of cutting this. There's performance mode where you know, you just kind of set it, forget it, and the AI will find the right segments for you. Price of data is baked in, you know, just optimizes based on your performance goals. Or there's control mode where traders are choosing everything from data to inventory. And then there's a, you know, incremental upcharge for data. Is this a good thing for the third party data marketplace or a bad thing or. Too early to tell.
- [43:33]
A
I think it's bad for third party data marketplace because if they're, yeah, if they're sort of commoditized and it's just like let the AI figure out what segment works. First of all, they can't charge a premium for branded segments. That's one of the dynamics here, is that in most DSPs, especially Trade Desk, you have branded segments and you have sort of white label segments and they're often the same data but at different prices. And, and they're also at different profit margins. For the dsp, the white labels are much higher, higher profit margins. So the AI isn't going to care about that. They will care. The AI probably will care about the margin, but won't care about the name. And so that's bad for them. The other thing is that we all know that a lot of data doesn't work and if the AI could turn it off faster and save you some money, that's good. The, I mean, I think the reaction from the, you know, the usual people online we, you know, who you are is being mixed. You know, I think that this is one of the consistent complaints about the Trade Desk is this feeling like they're taking margin in ways that you don't fully understand and that your, your actual take rate is higher than you thought because there's more data fees and this kind of feeds into that. It's another, it's a little mini black box that potentially, if you're skeptical or cynical, is a way to get more money into the trade desk coffers. But if get performance, it gets performance.
- [45:00]
C
Yeah, I think there's going to be a pretty big shakeout in the third party data space over the next 18 months. I mean even leaving this aside just in terms of like some of them are going to get figure out how to leverage AI in a way that makes sense and has value and others are going to pretend that they're doing that and that's going to create some challenges. And to Aries point, like there's just a lot of bad data out there and for 20 years if you had a really good sales team, you could correct for that to a certain extent because you had the relationships and it was harder to say, you know, that travel intender is not as good as this other travel intender. And so I think AI is going to, is going to force, is going to force some uncomfortable decisions within third party data companies.
- [45:46]
B
Doesn't sound like a bad thing from a marketer perspective.
- [45:51]
A
It doesn't seem bad to me.
- [45:52]
B
Yeah, overall market I think, you know, positive for them, you know, margin increases for, for TTD aside, I think it's a, it's a good thing for marketers.
- [46:05]
C
Well and, and what doesn't get talked about a ton is that there's this whole new cadre of like AI infused contextual targeting plays that really weren't available a year ago that are kind of interesting and been pretty effective. And so that's coming in and competing with, you know, the traditional not to beat up the travel intender, you know, salespeople. But like, but it's almost like the envelope Salespeople from the DMA back when you were what, 1999? I'm not trying to disparage anybody but like, but you need to be really rethinking you're, you're offering at this point.
- [46:41]
A
If you were outside our industry and you saw the way traffickers and DSPs use data segments, you'd be horrified. Like there's a list of literally like 75,000 options and the trafficker has like a spreadsheet that tells them how to do ands and ors and put a parentheses around here and exclude this segment. But this segment doesn't work. My client didn't approve this segment. It's nuts, it's, it's error prone, it's not scientific and it's definitely an area that could be improved with AI.
- [47:13]
B
Totally agree. Okay, move on. But stay on ttd. Jeff Green penned an op ed in the current, which is a TTD publication titled The Open Internet 2025 Part 1, Framing a Year of Change. Did you guys have a chance to read it?
- [47:31]
A
I did. It has poetry in it. It's. It's a bit something. Right. But I get the sentiment.
- [47:39]
B
Yeah. Yeah. So just to summarize, he's just talking about the dynamics of the market today, how it's still opaque and unfair and unbalanced. And he goes super hard at the TID change by prebid.
- [47:55]
A
Super. Yes. Yeah. So I talked about this when, when the TID change happened. It happened live, on air, on market, three weeks ago, which was funny. The thing about the TID change is it was the straw that broke the camel's back. Because I'm going to try. You have to put yourself like emotionally into Jeff Green's psyche here. I'm going to try to do this. He is running a successful company. He is. His competition, he thinks is Meta Applovin. These big ass companies that just have all the power in the world, they have AI and all that stuff. And all he's trying to do is compete with them. And on the other side of his transactions are a gaggle of companies that sometimes help, sometimes hurt, sometimes compete. Publishers who send them duplicate bids, fraud, malware, all this stuff that's persistent and that's not being cleaned up fast enough. And then publishers complain they're not making enough money, don't want to be an open path because they don't want to give up control. And I think he hit a breaking point. This TID change. TID change was just like kind of like a final, not a straw, but like it was basically a middle finger to all the efforts to clean up the Internet. It's like, you know what we want Bid duplication. You shouldn't be able to unduplicate things. We think bid duplication is a good idea. And that was it. He was done. He lost it. And he's like, f you, F all of your exchanges, publishers, take my tags, give me the inventory and I'll pay you a fair price and stop complaining. And I don't think this is going to end well for anyone who's in the display business that's not being transparent. Let's talk about what Adtech God said today. So we're recording on Wednesday. And Adtech God had a genuine scoop, which is he got a screenshot of a presentation Trade desk was giving to its employees. That said, they're launching their own pre bid for tomorrow, meaning Thursday, the day before this comes out. So we'll see if he's right or not. But basically they pre bid. The organization is the one who made this change to get rid of tid and he is forking it and he's going to have his own version of header bidding, similar to way, you know, Amazon doesn't use header pre bid. They have their own tam. And you know, it'll be interesting to see how hard they go if they require it, for example, to get into OpenPath or to get certain spend levels. But it's another front of the war Jeff has on trying to tame the open web and make it work for him.
- [50:26]
B
Yeah, this is fascinating. So presumably if this gets announced tomorrow, Open Internet part two gets dropped tomorrow. Announcing this.
- [50:34]
A
Yeah. Now I'm speaking at the pre bid summit in two weeks on the keynote and I think Jeff is scheduled to speak too at the pre bid summit. So that's going to be a spicy day of content. If Jeff goes through with it and if they let him speak, it'll be after this announcement.
- [50:49]
B
Yeah, I find it fascinating that this is like the most important thing. Right. Like Jeff could be writing about anything. I mean, they just had this announcement we'll talk about in a second with Ventura, the operating system, which is actually pretty interesting for ctv. But like he thinks this is, you know, pretty damn important, that he's penning op EDS and you know, putting a lot of time into this stuff. I juxtapose this with Applovin. Do you guys know what the market cap for Applovin is today?
- [51:17]
A
Jesus, I can't even guess.
- [51:21]
B
240 billion. Yeah.
- [51:25]
A
Yeah.
- [51:25]
B
And you know, just like just wholly focused on outcomes and commerce and you know, just pedal to the metal on it. So you can't bet against Jeff. Right. We've said that numerous times on this pod. We will say it numerous times on this pod. His instincts over time have been right. I just think it's pretty fascinating to look at what they're doing, which is largely about just kind of cleaning up and rearchitecting the open web because they think that that's going to be the most important thing and the channel to focus on in the future.
- [51:58]
A
There are people betting against Jeff. I just want to say that there's starting to be more and more people who are betting against Jeff and I'm not betting against Jeff, but it's becoming a little bit of a, of a wave.
- [52:10]
C
Is this not also, I mean, can this not also be viewed as a attempt to exert more control over the open web? I mean really, both this and the previous thing, I mean, that's really, I think what we're going to look back at in a year or two, assuming he's successful at both of these endeavors, is like, you know, these were the moments when the trade desk, you know, I said they're not already big, but like that, that the level of control that they had over the, the open Internet reached the tipping point. Ari, you look skeptical.
- [52:41]
A
Well, no, I'm not skeptical. I'm skeptical it's a bad thing. I'm not skeptical it's happening. So I think we should talk about Applovin for a second because I think there's a misconception that Applovin is a DSP that just has a really good algorithm that focuses on in app. And that's not really what they are. They're a closed loop system that has their own supply, their own ad format and their own demand. And, and that's why it works. And their supply is not stupid little banner ads that you click by accident. Their supply is overwhelmingly rewarded video that you have to spend 15 or 30 seconds watching. So it's a, it's a very aggressive format. They own it, it monetizes really well, and then the algorithm is works too. So that's all important. Right? So imagine if, imagine if trade desk's new strategy was the same, was like we have, we only buy this ad format that, you know, we'll publish how to build it. But like it's our ad format. We only buy it through Open Path. We stop buying anything else and we're going to optimize it. I don't know if that would work commercially for them because a lot of customers wouldn't like it necessarily. But I can almost guarantee you it would get better performance.
- [53:54]
B
It doesn't sound like a bad idea at all. And that actually might be how this thing all comes together.
- [53:59]
A
Wait, isn't that Undertone's business model? Is that what you guys did like 10 years ago?
- [54:03]
B
We did. It was a great business. Jeff, I'll let you. Boy, come on, I can give you some ideas. There's a playbook for this stuff. Yeah, right. Like it's a, you know, you find a performance format, you, you know, lock up like publisher inventory, you pay publishers well, you have an algorithm that drives performance. Like, I mean that's not just Applovin, that's meta, right? Like it's, it's yeah, their properties. But it's like a great format, you know, across Instagram and Facebook, so on and so forth. Incredible AI, the ability to, you know, kind of onboard advertisers and make it really easy to build creative. Like that's what the biggest platforms are right now.
- [54:45]
A
So I'd contrast that with Alan's statement that you know, he's trying to control it as sort of a bad thing. Like maybe, maybe that's proposal to only have one format and only go direct is unrealistic, extreme, too disruptive. But maybe, but that's the direction they're headed. Like that would be a better business than his current business. Right. Probably a better web.
- [55:07]
B
Yeah.
- [55:07]
A
And the open web publishers would make more money too. So. So maybe this like the way he's seeking more control, not total control, more control is a good thing.
- [55:19]
B
Alan, do you want to rebut before we move on?
- [55:21]
C
No, I think you may be right. But I reserve the right to completely contradict myself in a year and go, you know, that was the point where the tipping point was okay.
- [55:34]
A
And bringing it back. This is effectively Google's argument for why I should own adx. All the demand comes from them. It's their demand. Adex is just a, is a thin shell to getting Google Ads demand into publishers hands. Right. So they just shouldn't have done it in conjunction with the ad network with the ad server. The real solution to the Google problem is to sell the ad server, not to sell the ad exchange. Keep the ad exchange, sell the ad server. All right, sorry I'm a one trick pony here.
- [56:03]
B
Final point on ttv, they just announced this a little bit earlier today. They're partnering with DirecTV for Ventura operating systems. So they're going to be looking to do deals with OEMs to bring the DIRECTV. I think consumer experience along with TTD's monetization, I think this is a good idea. One that might have some legs. But we shall see. All right, that was a lot on ttd. Let's talk about Amazon. Amazon. I will review. Week before last or last week they announced that they were offering access to Netflix in the Amazon dsp. This week they are offering access to Spotify in the Amazon dsp. This is significant from two perspectives. So number one, one of the areas that TCD talks about as being under monetize, the best of the Internet, a growth area is premium audio. So Spotify is premium audio. Spotify is probably second to or first versus Amazon Music. So Amazon probably has access to the most premium audio and then Also, Spotify had that bad quarter and they blamed a lot of the bad quarter on execution on their ads business. So this gives them, I think, a good answer for, for that. So I'm impressed. Amazon is going for it. And by it I mean all the supply.
- [57:37]
A
Yeah, I mean I tweeted this, but I'll reiterate it for those who don't follow him on Twitter, which is basically access to supply is becoming a scale advantage to the biggest players. The Spotify and the Netflixes of the world are not opening themselves up to the top 10 DSPs globally. They're opening themselves up to 1, 2, 3 and then they say, maybe we'll do number four at some point. If you were going to start a new DSP today, you'd have a real hard time, especially if you're, if you're going after, you know, premium advertisers who care about this sort of stuff. These hedged gardens are being cautious and it's a big advantage to those big guys.
- [58:18]
B
Agreed. Dsp. You got me all fired up. Start another undertone.
- [58:25]
A
Yeah, we should totally. I'll call it overtone or something. The other story here, which is a longer. Has been going on for a long time, Spotify's diminished ambitions, like they, five years ago, four years ago, they were, they were trying to become like the Google of audio, where they were trying to run the platform play. They'd have an ad server, they'd have inventory, they have an ad network, all that stuff and it's sort of flattened out to, you know, monetizing their inventory and having tools to monetize their inventory and nothing wrong with that. But like, it's noteworthy that that strategy didn't work for one reason or another.
- [58:59]
B
Yeah, I can, it's super hard.
- [59:00]
C
I can just tell you as a user of their ad system over the last, you know, several years.
- [59:05]
A
It's so bad.
- [59:06]
C
Horrible. Yeah, clunky.
- [59:10]
A
I've talked on this podcast before about it. Okay, I've ranted.
- [59:15]
C
I'm with you, brother.
- [59:16]
B
All right, what? One more and then we'll, we'll call it a day. Just shout out to. To Vibe. Are you are on their board? Vibe raised a 50 million dollar Series B at a 410 million valuation, which is a monster. Vibe is going after the SMB CTV market, presumably doing really well. So shout out to them. And Arthur made a appearance on TBPN this week. So for those in the tech nerd circle, it doesn't get bigger than that. So shout out to him and shout out to Vob.
- [59:50]
A
Yeah. Huge accomplishment for Arthur and the team. They were executing like crazy. It's a hot space. I'm an investor and I was on their board. I do. I should probably announce that I'm not on the board anymore. As part of this financing, they reshuffle the board to be more financially oriented. But it was. I did my time, helped out, and my portfolio of angel investment is going really well. I'm pretty excited. Maybe I'll start a VC fund to compete with Eric.
- [60:18]
B
Awesome. That's just what I need. You're competing against me.
- [60:23]
A
If you know any LPs with extra funds, just send them my way.
- [60:26]
B
Absolutely. All right, I think we call it there. Guys, this has been awesome.
- [60:30]
A
This has been great. So, Alan, thank you so much for joining us. And we're probably going to have you back when we get more info about this whole thing. This is like an endless story, but thanks so much. And where can we find you, Alan, besides the Monopoly Report podcast?
- [60:43]
C
Find me on LinkedIn. You can find me on the Future Media podcast. I do with publisher guy named Ricky Sutton, which goes into some similar issues of the Monopoly Report, but from a slightly different vantage point. Lots of fun, great conversations.
- [60:57]
A
It is. It's a really good one. And Eric, thanks as always.
- [61:01]
B
Thank you, guys. See you next week, everybody.
- [61:16]
A
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@adtechgod.com.