Google Had Its Day In Court. Now, It's Amazon's Turn

0:08Welcome to The Big Story, a roundtable featuring members of the AdExchanger editorial team. Every week, we bring you an in-depth discussion of key developments in digital marketing and media.

0:23Today's episode is sponsored by Basis, the leading intelligent operating system for autonomous advertising. Its enterprise AI solution transforms campaign briefs into strategies and media plans that integrate directly into omni-channel activation. You can learn more at basis.com.

0:46Hello, listeners, and welcome to this week's episode of The Big Story. I'm Alyssa Boyle, Senior Editor of Ad Exchanger and Synopsis, covering for our fearless leader, Sarah Sluice. So it's never a slow week in ad tech. This week, we're going to do a bit of a roundup of some of the biggest news developments, plus some of the bigger news that we've covered over the last week. So in a nutshell, we're going to talk about how big tech companies have been facing quite a few legal battles. This week, news broke that a federal judge rejected the Department of Justice's request to force Google to divest parts of its ad business after Google was found to have a monopoly over online ad auctions.

1:22The other major news headline of the week is that the FTC is suing Amazon for allegedly making advertisers pay more than they needed to for ads on the e-commerce site. Yikes. In the second part of today's episode, we'll be a little bit more, a little less womp womp, a little more AI. Yay. We're going to talk about some of the major AI developments our team has been covering in the last week. Number one, OpenAI has been bragging about its annualized ad revenue and the Trade Desk touts fancy new AI upgrades to its KoKai platform. Okay, that's a lot of words. But before we dive in, shameless plug for Programmatic I.O.

1:58New York, taking place this month, September 28th and 29th, media and advertising thought leaders will be gathering at the Marriott Marquis in New York City to discuss the biggest growth opportunities in the industry today from AI tech to the state of the open web. So if you haven't registered yet, secure your seat today and you can use the discount code POD10 for 10 % off. Okay, now let's dive into some of these top headlines that I was yapping about. With me today, I have Ad Exchanger's Managing Editor, Anthony Vargas. Hey, everybody. Oh, that has a nice ring to it. Managing Editor, I love it.

2:33All right. Congrats again. And then our associate editor, Joanna Gerber. Also a nice ring to it, if you ask me.

2:44All right. So let's let's start with the let's start with the big kahuna here as far as the news headlines of the week. So, of course, we've been covering Google's antitrust trials ever since a federal judge determined it has a monopoly over and online ad auctions, not just online search. And so many industry experts and, of course, Google competitors have been trying to get a slice of the online ad market from Google, hoping that it gets forced into structural remedies. But unfortunately, depending on who you ask, a judge determined that Google will not have to break up parts of its ad business.

3:21So let's talk about this. Anthony, you've been on top of what's going on with Google for a while. What's your read? Does it make sense to you that the same judge who ruled Google an online ad monopolist is then going to go easy on Google by not making them start divesting part of their ads business? Yeah, so I'll say it's not really unusual because we've seen this once already. So in Judge Emmett Mehta's Google search antitrust trial, a lot of stakeholders were hoping that Google would be made to divest its Chrome browser. and Judge Meta declined to impose that as a remedy, even though Google owning such a popular web browser was helping it protect its monopoly over online search.

3:57So now in the AdTech antitrust trial, Judge Leoni Brinkema had previously ruled, like you pointed out, Alyssa, that Google operated a monopoly in the SSP and ad server markets, largely because it bundled its SSP, which is Google AdX, and it's double-clicked for publisher's ad server into one product, namely Google Ad Manager. So it seemed like there could be a chance that Google would be made to split up GAM into different products. And in fact, the DOJ, the Department of Justice, recommended exactly that. But now that officially will not happen. So if you predicted that, you know, Google's not going to be made to break up its ad business, pat yourself on the back.

4:37um so yeah this is the second time that the federal government has found google guilty of operating a monopoly but declined to take any meaningful action to break up that monopoly so you know i've said this before it's basically like the government ruling okay yeah you're a monopoly and that's just fine uh to be clear uh judge brinkema is going to recommend that google change some of its anti-competitive practices as part of the final decision on remedies in the ad tech trial. It's too soon to say what exactly Google's going to have to change. And, you know, you better believe we're going to be staying on top of the story and covering that going forward.

5:14So, you know, watch Ad Exchanger for more updates there. But the reason that Judge Brinkham had declined to go the breakup route appears to be, you know, she had previously expressed some concerns that, you know, breaking up Google would trigger a year's long appeal process. And that would then potentially delay any relief for publishers and ad tech companies that had been harmed by Google's anti-competitive practices. So in her mind, she's kind of going the easier route here rather than what the federal government had projected. And the DOJ, I think, estimated could be like a decade long, maybe as long as like 14 or 15 years long appeal process pursuing a breakup.

5:54But basically, if you're looking at this objectively, it sucks to say, but antitrust enforcement pretty much no longer exists in America when it comes to big tech companies. As Alan Chappell, who's a lawyer who advises ad tech companies, told our James Hersher earlier this year, if divestiture fails here in a case where Google systematically dismantled two ad tech markets and maintained a 20 % take rate, it admitted it had no plans to reduce, then when exactly does structural separation become appropriate? It's an excellent question and And apparently the answer is never. Or as one former media buyer I spoke to said, if the trust buster Teddy Roosevelt somehow was transported to today, he'd be flipping out at our antitrust regulators and wondering where their balls went.

6:43Is that what people mean when they say I was born in the wrong generation? Like we need we need some Teddy with us right now. Do you think do you think any ad tech nerds made a bunch of money on Kalshi or something for betting that this would happen? Or do you think they lost? I didn't check if Kalshi was taking bets on this, but I wouldn't be surprised. Probably not. Of course they were. What do you mean Kalshi takes bets on everything? That's kind of a whole thing. I don't use it, so I wouldn't know. But I have to imagine. Oh, man. Yeah. Well, it sounds like, you know, anti-competitive practices recommending that Google changes that.

7:14That is a very vague, elusive, and strange checkbox to have on our watch list. But that's so be it. That's the state of things. um one more question about google so does this essentially mean that we just have to trust that google's going to do the right thing because how far is that going to get us yeah i mean like i wouldn't say necessarily like it's just like trust google because i would imagine like you know the federal courts are going to exercise some oversight over whatever remedies end up getting put into place and they'll be you know checking to make sure that google complies with whatever kind of fixes are recommended here.

7:48But yeah, I like what this kind of boils down to, I guess, because like, since we won't be getting the DOJ's proposed remedy, which would have been a breakup of the ad business, it seems more likely that Judge Brinkham is going to mandate some of Google's own proposed remedies instead, which included Google promising to make real-time bid amounts for open web display ads from our ad exchange, AdEx, available to all rival publisher ad servers. Deprecate unified pricing rules for open web display ads, giving publishers the option to set different price floors for different bidders when using Google Ad Manager, and that Google would commit to not using first look and last look for open web display ads, although Google adds that these auction dynamics were phased out of Google Ad Manager years ago.

8:37So basically, the two big likely fixes are those first things where other ad servers are going to get more real-time information on what's going on with bid amounts and unified pricing rules will be deprecated and publishers can have more control over their pricing rules. Don't take that to the bank, though, because all of this remains to be seen. And like I said, we're going to be following up on this, so stay tuned. Yeah. So we don't want we don't want ad tech companies to create their own homework, but perhaps they can make their own remedies, question mark. I guess we'll find out. Let's move on now to to Amazon, the other big tech behemoth in the hot seat.

9:14The FTC and 22 state attorneys general sued Amazon, accusing it of making marketers pay more than necessary for retail media ads purchased on Amazon's second price auctions. Let's break that down into English. So, Anthony, what does this accusation reveal about how Amazon's second price auction works? And does it seem a little bit shady? Yes, I can say yes to that. And you didn't even ask me. well you know assuming like you know all this is exactly as the ftc alleges it would it would definitely be it would hint at like the same kind of like uh apparent auction manipulation that google had previously gotten into a lot of hot water for which i'll you know get into in a little in a little bit but just to kind of catch people up who may not have been following the news here on the the ftc lawsuit against amazon um so the take i'm seeing a lot is that the ftc case pretty much centers on a feature that Amazon introduced around 2019 called the soft reserve price.

10:11And the FTC alleges that this feature basically turned Amazon second price retail media auctions into something closer to a first price auction. So to kind of explain that in a typical second price auction, the highest bidder is only required to pay one cent more than the second highest bid. So if I bid$20 and extra highest bidder bid$10 for an ad impression, I would only have to pay $10.01. But in a first price auction, if you put forth a bid and that's the highest bid, you have to pay that full amount. So I would be paying that full$20. Now the FTC alleges that Amazon had this soft reserve price feature, which basically functioned as a fake bidder in their ad auctions that would bid closer to what the winning bidder's price was.

10:58That way, the winning bidder would end up paying pretty much its full bid rather than just one more than the next highest bidder. And according to the FTC, by 2024, winning bidders in Amazon's auctions were paying nearly their entire bid price 80 % of the time, which would have equaled advertisers spending tens of billions of dollars more than they should have for Amazon's ads. In a LinkedIn post, Tim Vanderhoek, who's the CEO of Viant, which has a DSP that competes with Amazon, described this as Amazon copying Google's playbook. He pointed out that Google did what it called reserve price optimization, which he described as an artificial bid pushing your price up.

11:35And yeah, that sounds exactly what the FTC is saying Amazon is not doing. Of course, Amazon denies that it did anything wrong and it claims its auction dynamics have actually resulted in the average winning bid for its retail media ads dropping by 50 % between 2019 and 2024. But yeah, this is not a great look for Amazon and I'm hearing a lot of outrage from my buy side sources uh but to be honest the public criticism seems to be a little bit lacking at least as far as like what my buy my my buyer sources are saying um for example i heard from david nurenberg at intermedia advertising and he said he's surprised that it seems the industry reaction here has been muted especially when it seems like everyone goes after smaller but still big independent players like the trade desk for every little thing uh when it comes to like a lack of transparency in their auctions.

12:23And I also saw an interesting LinkedIn take from Sarah Caputo, who's also been a big trade desk critic and has a buy side consultancy where she advises buyers on different tech platforms. She's been pointing to the Amazon DSP as an example for the rest of the buy side platforms to emulate. And after the FTC lawsuit against Amazon dropped earlier this week and these revelations came out, she posted that it was like an et to Brute moment where it's like, who can you really trust in the marketplace? You know, I don't know if this is a testament to the ad tech industry or a sign that we all need to go touch grass, but some of my friends outside the industry were recently talking to me and they were like, you know, Joanna, I feel like you always have just like interesting LinkedIn drama from work.

13:07Like people in your industry are always just getting into it on LinkedIn. So if that's what our industry is known for, yeah, I think we need to just think about what that means. Yeah, LinkedIn is always the place to go for the hot takes after anything kind of breaks like this. Apparently, and I have I just feel like I have a few questions. You know, if Amazon's claim is to be believed, then I'm wondering, like, how much were people paying for ads before 2019? Then my other thought is that I'm in the office and there's like this trophy here. This is the ad exchanger man. And I keep seeing myself reflected in this thing like a Snapchat filter.

13:38And it's hysterical. shout out to ad exchanger man um for those who are watching uh on video today um okay so let's let's wrap up the segment with one more question but um i'm gonna leave this to both of you either of you share share your thoughts and ambitions here uh what does this lawsuit and allegation say about how advertisers and the government expect big tech to behave uh yeah so i would say this this amazon ftc dust up is it was a perfect illustration of why advertisers call for more auction transparency from ad platforms um another another media buyer i i heard from about the uh the amazon ftc uh sued uh told me uh first of all we're talking about like search not open web programmatic so are auction logs even available you know meaning to like check like the transparency of like what was going on in these Amazon auctions.

14:36And they said, second, you'd need to review your own win logs as well as have the associated bids for that auction that include losing bids, not just your winning bids. So all of this stuff they were saying is basically impossible to get. And the only ones that would have it are Amazon and now the FTC through legal subpoenas. And they added that all of these recent legal actions against ad tech platforms have revealed a serious need for auction standards and regulation. They said, everything's still alleged, of course, but if ad tech companies, advertisers, and consumers can all be harmed as a result of how our market functions, then our market needs basic regulation to control auction mechanics.

15:20I think I was also just thinking about, I mean, obviously, you know, better late than never. But if this has been going on for years, why did it take the FTC this long to realize what was going on? And it unfortunately sort of begs the question, what other sort of sketchy operations are going on in these big tech companies that due to, you know, the sort of black box technology, we don't know what's happening in them. So how many more scams might be happening right under our noses that might never be uncovered or might take another five years? Yeah, we should we should say on that deeply optimistic note.

15:54We should say we should say alleged scams, right? Yeah, just wanted to get that in. And yeah, on that deeply optimistic note, it feels like a good time for an ad break, right? Yeah, don't we all want to hear ads at a time like this? All right. So we're going to take a quick break. And when we're back, we're going to talk about some of the AI developments that our team has been covering over the last week from OpenAI to the Trade Desk's COCAI. I didn't even mean to rhyme, but here we are. All right. And stick with us.

16:31I'm Sarah Sluice, Editorial Director of Ad Exchanger, and I'm here with Mark McEachran, VP of Product at Basis, which is supporting our podcast this month. Hello, Mark. Hello. How are you doing? I'm good, especially because we're going to talk about three-letter acronyms, which I think is one of ad tech's favorite things, although this goes beyond just ad tech. So help explain this to our listeners. what is an MCP server and how are advertisers using it? I do love a good TLA. MCP stands for Model Context Protocol. Anthropic built it, open-source it in late 2024, and then handed it to the Linux Foundation.

17:14It's plumbing that connects the systems to agents. It lets an AI model reach into a piece of software, see what's available, and pull the right lever instead of hand developing of the spoke integration for every tool that the model touches. And over the past year, Google, Meta, Amazon, and others have rolled out their own ad platform and MCP servers, as have DSPs and SSPs. A buyer can now say, pull my campaign report, or build this audience in plain language and have that request land correctly inside whichever platform is listening without building a custom connector for every combination. And eventually, those instruction sets, they begin to look like standard adapters.

18:02Okay. So MCPs are taking off like wildfire, I would say, and with good reason because they're very useful. So tell us a little bit more about what advertisers are able to control when platforms have an MCP server. So control depends on who built the server. An agent can pull reporting, diagnose a stalled campaign, or explain a CPC spike. But in most cases, a human still has to make the change. Some folks like Meta have enabled agents to create audience, set budget, and launch a campaign all by itself. But where it's furthest along, advertisers can do all that. They can let the agent read performance data, build and activate audiences, adjust the budget, launch campaigns, and even generate creative variations all through natural language.

18:49letting one agent shift budget across platforms on its own is where the industry says it's heading. But that's not the standard practice yet. The industry direction is obvious, though. Fewer read-only endpoints, more agents with their hands actually on the wheel. And at that point, governance and guardrails become the real conversation, not the pipes themselves. So you're talking about a lot of new ways that people can connect and companies can connect to each other. So does that mean we're going to be solving for media fragmentation? Marketers run dozens of platforms to get to all that media.

19:28DSPs, walled gardens with the clean rooms, measurements, they're all bolted together with their own quirks and logins. Each of them historically needed their own bespoke API integration, its own authorization, and some team member who knew how it all fit together. there. MCP doesn't unify the stack, but it has the potential to unify the interface to the stack. One agent could sit on top of all of them. As an example at Basis, we have Compass, our proprietary strategy and planning agent. It has direct integrations with buying platforms, including our own DSP. Advertisers use it to upload briefs along with their research, performance history, which is then synthesized into a proposed media mix, audience personas, and creative strategy, initiating an omni-channel planning conversation through one interface.

20:22Okay. So if I'm thinking about MCP servers as a shorthand, it might help unify my interfaces if I'm a buyer, but the platforms themselves remain fragmented. Indeed. There's a lot of room for innovation under the hood. MCP is the unifying layer that is the command and control on top of it. Well, thank you, Mark, for helping explain MCP servers. I know I understand them a bit better after these few minutes. And thanks to Basis for supporting our podcast. Thank you.

20:54And we're back. So let's dive into what's going on with OpenAI. So Joanna, you've been on the AI beat since you joined AdExchanger and you've been covering OpenAI very closely, especially since they launched an ads business earlier this year. so tell us more about development of my career I know I remember like trying to be in chat GPT like show me an ad please I need to see what it looks like so now I'll occasionally see like you know something basic you know I'll be like you know asking a questions like what's a good gift to get for somebody and it'll show me an ad for flowers so we're seeing the early signs of an ad business anyway Joanna so tell us about the latest update about with OpenAI's annualized ad revenue and what that means in English.

21:39Yeah. So earlier this week, OpenAI announced that it hit$1 billion in annualized revenue run rate. And annualized revenue run rate, we can't even abbreviate it because it would just be R and then we'd all sound like pirates. But what it means is basically, if you made$10 million last month, your annualized revenue run rate would be that multiplied by$12. So it's just a fancy way of saying if we keep earning money at the exact same rate we've been earning money very recently, this is how much we'd make in a year. So sometimes that's not a super useful metric, like specifically with companies that have big spikes in revenue at certain times of the year.

22:15Like if there's, you know, specific holiday pricing or people like to buy their products in the summer, especially an annualized revenue run rate might be incredibly misleading if you base it on, you know, August purchases for, I don't know, ice cream sales. But for a company like OpenAI that is actively growing and expanding right now, these past couple of weeks or months might actually be an extremely good indicator of growth that's to come. And it's maybe even going to continue growing. Exciting. All right. So, Joanna, you also are in close contact with many agencies and buyers. What do they make of this update and what this annualized ad revenue brag means about OpenAI's ad business?

22:56Yeah. So I spoke to a couple of agencies, Jellyfish, as well as MileMarker. And I also talked to LifeSite, which is a B2B measurement platform who interestingly does some measurement work within ChatGPT ads, but they're also an ad client of ChatGPT or of OpenAI. So they've been advertising directly within ChatGPT as well. And all of them seem to have, as you might expect at this stage in a very early ads business, some pros and some cons to talk about. So the positive feedback, according to Jellyfish, is that they're just excited to get insight into how to advertise in LLMs and what that looks like and what's working and, you know, what the ads should look like.

23:40Because, you know, obviously OpenAI wasn't the first major AI company to launch ads, but they were the biggest. Like, no offense to perplexity, but it just didn't resonate the same way. It wasn't part of the cultural zeitgeist, one might say. So that's just been very exciting for marketers to just understand what ad intent is going to look like in LLMs. However, the other thing that advertisers are very excited about, and I think OpenAI kind of made this their whole shtick early on, is access to purchase intent in real time. Because it doesn't get any more real time than someone saying, oh, I want to buy this exact type of product.

24:20I want to do this exact type of thing. I want it to be for curly hair and I want it to be moisturizing and I want it to be able to last for at least six hours when I go to this wedding. And I think advertisers are very excited about the idea of getting in front of potential customers at the moment they're ready to click the buy button. It's that point of sale that we keep talking about and I hate abbreviating it because my brain always reads it's wrong. Yeah, exactly. But you see POS and you think something else, get your mind out of the gutter. Yeah, exactly. I always have the same problem. You know, piece of poo.

24:51Yeah. I do too. I don't know if I'm allowed to curse. I always forget. So when it comes to like, when it comes to open AI's like platform itself and just kind of like the idea of like, you know, because it's like, this is kind of like my take on like open AI's like ad business because like open AI is kind of like the, what everybody kind of thinks of as far as like on the consumer basis, when they think of these like chatbot interfaces, like the most popular one in most consumers' minds is ChatGPT. And we're hearing all this talk about how these chatbot interfaces and alternate approaches to search and all kinds of different personal shopping assistants and all this kind of stuff that's going to be fueled by this technology is going to completely change the internet and completely change the way people find products and discover new brands and all these different things.

25:42So in my mind, you have to set the bar for OpenAI's ad business as high as you possibly can because they're setting the bar as high as they possibly can for themselves so i'm looking at like what is it going to take for open ai to become a 200 billion dollar advertising business like a meta and when you think about that it's like okay there's going to have to be a lot of development on the technology end uh to differentiate itself a bit and also you know it can't like keep coasting on the idea that like oh chat gpt ads are like the new shiny thing. So like the demand is just going to flow by virtue of that.

26:18If you want to 200x this annual run rate, there's going to have to be a lot of tech investment. And from what I'm hearing from advertisers is, you know, OpenAI has a very, you know, it has a basic self-serve buying platform that's similar to what a lot of other platforms like Meta offer. They, you know, we previously covered on an earlier episode of the big story, how OpenAI has been bringing in a lot of former, you know, ad people from meta to design its platform. They're pretty much kind of like, you know, doing a lot of like copying meta's self-serve features while also trying to like, you know, tailor it to the search chat interface.

26:59And, you know, there's been a rapid pace of development of introducing new features, but I haven't really heard anything about like major differentiators from OpenAI from an ad tech standpoint. And then in terms of just kind of like getting advertisers onto the platform in these kind of early days. Cause you know, to be fair, I said, we're setting the bar high. It's still just the first year of open AI as an advertising company. So, you know, one of the things I'm hearing is that in order to kind of juice adoption from the buy side, open AI has been offering buyers a lot of ad credits to get them to test the platform.

27:33And it's unclear if those credits are included in the calculations for this estimated run rate. but the fact that those credits are a thing shows that OpenAI is serious about really pushing aggressively buy-side adoption and how long it kind of takes them to get to the meta tier I guess that we'll have to keep an eye on but that's where we're setting the bar so if they don't get to that 200 million level I'm going to say you can say that I said it that they failed okay quoting Anthony well you heard it here Yeah, but so as far as what advertisers expect from OpenAI to invest, you know, more ad dollars and have more confidence in it, they want to see, you know, self-serve capabilities, transparency, proximity to point of sale, maybe some ad credits.

28:20That's one other thing I actually wanted to make sure we squeezed in there. So between the agencies and the brand I spoke to, all three of them immediately when I said, oh, you know, like, what do you think advertisers will need to be more comfortable spending with open AI? Better measurement capabilities. But specifically, one thing that came up multiple times was competitor data, because brands are apparently increasingly asking for this competitor data to see not just how they're performing but how they're performing compared to other brands selling similar things because this is a new channel so there aren't these kind of you know baseline metrics that everyone's already familiar with and i think until they know how they perform they're performing in comparison to the rest of a given industry they don't feel super confident necessarily giving a ton more money um they also said it's not super great in terms of measuring roi yet which is again to be expected.

29:12But as of May, as Anthony mentioned earlier, OpenAI launched a self-serve ads manager for ChatGPT ads, and it took away its prior minimum spending commitment. So obviously it's more accessible to a much wider array of advertisers now, but it sounds like they're still in the pretty early stages. And as far as self-serve ads managers go, it's not exactly boasting any super exciting, unique capabilities yet. And again, no shade they've only had a few months to develop it but i think they're going to need to to pull out the stops if they really want to impress advertisers because there's so many channels right now yeah that jives with what i'm hearing as far as as far as like the big kind of like obstacles to getting like advertisers more comfortable on here they want more measurement they want more benchmarking you know basically it's like an open question what it will even look like to measure the effectiveness of chatbot ads right now and no one really has any satisfying answers at the moment but as joanna said it's still very early days a lot of the measurement seems to be happening in-house right now like jellyfish is offering measurement capabilities to its clients um life site as i said is a measurement platform so obviously it's handling a lot of its own measurement but it would be nice if there was kind of yeah a baseline consensus of this is how we measure ads on this channel because again creating one's own homework not great Mm-hmm.

30:34Well, the to-do list is clear, so we'll keep an eye out for that. Okay, so sticking with AI, but zooming out from OpenAI. So, Anthony, let's talk a bit about the trade desk. You covered the trade desk pretty closely, including Kokai, which is the buyer interface it wishes everyone would use. Tell us a little bit about the update that you covered late last week. It's Kokai Zuma update, which I keep wanting to read as Zumba, which makes me just really want to start dancing you know i mean you could do i could yeah so there is a video element to this podcast so the the big thing that that the trade desk is adding here uh is you know speaking of chatbots a chatbot interface called co-assistant that will pull a bunch of different agentic solutions it's developing together into one kind of conversational interface uh for example the trade desk has a new AI agent for audience building.

31:28It also added new agents for ad frequency optimization, customer support, troubleshooting, building campaign templates from a media plan, pulling performance insights from campaigns mid-flight. All of this can now be pinged from just through one chatbot interface. And the other big thing is they're overhauling KOKAI's whole user interface. So the periodic table interface that the Trade Desk introduced in 2023 is no longer KOKAI's default view. And now there are visible campaign insights right on the campaign landing page. So advertisers have to click into fewer widgets to make and model their campaign adjustments.

32:05They can just see them all at a glance. So the trade desk has been pushing Kokai for some time. Anthony, do you have a sense of how buyers are or will be reacting to this update? Yeah. I reached out to a couple of my buy side sources and I definitely made sure to, again, ping David Nuremberg about this too because he's been an outspoken critic of kokai in a lot of ways while you know also giving it its flowers and you know some of the a optimization features uh that the trade desk has introduced uh through its koai tech um you know some of that stuff is producing a better roi but like you know there's so much annoyance out there about the actual user interface for kokai and just like the um the ease of accessibility for different insights and you know a lot of people just feel like the old Solomar interface was just much more intuitive and there were less of these just like weird kind of like design choices um and yeah David was telling me he seemed very impressed you know he he was very impressed with uh the new conversational interface he says that you know because he's at the executive level now at his ad agency he doesn't have the time that he used to to like be like pulling these like pivot tables for like individual campaigns and really digging into what's happening but like because he's leading more of like an oversight role he he needs to be aware of what's going on with all these like kind of different campaigns that his agency is running.

33:25So he thinks the chat function and the ability to drill into data on demand just by like querying a chat bot is really going to make senior ad agency executives a lot more efficient in how they oversee all these different pieces. So, you know, some good, some positive developments there. And overall, I think, you know, everybody that I've seen has been pretty positive about the UI adjustments um where like you know it's like oh they have like more data visible at a glance they don't have to click into so many different like individual like widgets to kind of see what's going on and model campaign changes and stuff and uh you know the the advertisers that i'm talking to want to see more of that they want even more information visible just like on those landing pages cut out you know some of those additional kind of like extra uh graphical elements that don't necessarily need to be there and just give me more data right there at a glance

34:18Okay, well, so we covered a lot of ground. Do you think there's anything else that sticks out to you as far as in terms of addressing pain points or needs that buyers have? Is there anything else that seems particularly unique about Kokaizuma? And I ask this because, you know, in your in your coverage, you mentioned that a lot of programmatic platforms are rolling out a lot of fancy AI and agendic upgrades to make buyers lives easier. But I'm wondering if there's anything else that this seems to really be addressing. yeah so you know to your point there like it feels like every buy side platform uh is introducing these kinds of like uh chatbot type like features and like audience building like you know agents where like you just put in a campaign prompt and it gives you all kinds of like deal ids to target audiences to like you know include and exclude from the campaign targeting uh different like data graphs that you could layer on to try and find like new audiences so like i feel like everybody's got that like audience building like piece um but then you know i was kind of struck by like well you know just to kind of put a finer point on that so like all like all the trade desks like major competitors like google dv 360 has like a like its own like kind of like audience building assistant amazon dsp yahoo dsp have introduced similar tools ssps are also doing the same thing um you know we're also hearing from like some like like data management platforms like Optible is like as it's like solutions for pulling publisher first party data and surfacing audiences that the publishers may not even have been aware or in demand aware or in demand meanwhile we're hearing from some ad agencies that they're building their own kind of like audience building tools and stuff like that but what struck me about the Trade Desk kind of new feature set it feels like they're really leaning into like the having the and a specialized AI agent for every kind of like pain point and I know this is something that I'll speak to like here be covering CTV Alyssa the frequency capping piece like the idea of having a dedicated agent that's just looking at ad frequency management like that struck me as unique I'm not sure if like other platforms have similar things like that but it's like okay the idea that like but it feels like the industry is kind of like coalescing around this idea that like okay we can we can solve everything with agentic AI now you know all of like those like those pain points are now solvable.

36:38Whether or not the AI tools are more effective at addressing those pain points, all that remains to be seen. And we'll be keeping an eye on that, you better believe. I'm going to quickly just respond to that, Anthony, by saying that something I've been seeing a lot more of is companies launching a suite of agentic agents that each have a very, very hyper specific task, but the marketer or whoever's engaging with the agent talks to a centralized agent that then deploys whichever one is best suited for the task. So the human doesn't actually have to balance 15 agents at once. They can just say to the central one, oh, I want to check out, you know, how to improve the number of clicks I'm getting on this one particular ad in this specific demographic.

37:24And the central agent will deploy the correct one from there, which I I do think makes it easier to balance, even when it sounds wildly complex to have so many different ones floating around. Yeah, for sure. All right. Well, we have a lot to watch out for and a lot of exciting and interesting and weird things happening in the ad industry, as always. So thanks for sticking with us. And we'll see you at Programmatic I of New York. Thanks for listening. Woo!

37:57for autonomous advertising. Its enterprise AI solution transforms campaign briefs into strategies and media plans that integrate directly into omnichannel activation. Learn more at basis.com.