B (3:23)

Yeah, I think there's. I have three speaker highlights, but the startup session was awesome.

A (3:27)

It was awesome.

B (3:28)

Like five great startups. So a little bit of a battle of the agents we had. The winner was this company, epaminds. Everybody should check out. It's very, very cool. Their agent is named Lucy. And then we had Gigi, which we've had on the pod Adam before. Their agent is called Gigi. So it was a little bit of Lucy versus Versus Gigi for the, for the crown, which was, which was indicative of, I think, the times.

A (3:54)

If I was doing this, if I was doing an AI startup, I would do a. A hyper, masculine, macho voice just to. Just to differentiate from all the pleasant female voice there. I would go like, you know, I would go like red state on it. Like, hey, what do you want, man? I'm here to help.

B (4:12)

Yeah, Chad. Chad is here.

A (4:16)

Chad. Yeah, that'd be perfect.

B (4:18)

Exactly, Exactly. Instead of, hey, Chad. Hey, Chad. Content wise, two. Two things stood out to me and, and very much in the. In the realm of like, doers and thinkers. The first was James Burrow and Nick Sharma talking about ctv. And James, James is like, he's pretty amazing. It was one of the few quotes that I think you posted on X, which was, you know, basically, you know, him saying, think of TV as a.

A (4:49)

Was it a vertical horizontal Instagram ad? Horrible horizontal Instagram ad.

B (4:55)

Yeah, but. But there was a deeper meaning behind it, which is think of TV as you think of Instagram and then other performance channels, you know.

A (5:06)

Yeah, yeah. I think he also put on Twitter. He didn't say this live, but I thought it was insightful, which is think of pixels as targeting, not as attribution. He said it.

B (5:15)

He said it there, but he said it very succinctly on X. And then. Yeah, the other thing is tv. Think of it as the second screen, not the first screen. Right. Like, so it is so impressive how, like, people like James are pushing the space forward and thinking differently. And then the other thing, so he's doer and. And then the thinker. I mean, Eric Seufert.

A (5:40)

Eric Seufert.

B (5:42)

I don't even know what to say. I'm still reeling. My brain is still trying to catch up. It was the craziest presentation I have ever seen.

A (5:52)

Yeah. So picture this. We're after lunch, and people are, you know, digesting their. Their nice L. They're hanging out in their ad tech tuxedos, blazers and jeans, and, you know, getting ready. Yeah. They're thinking to themselves, like, we're.

B (6:06)

We're.

A (6:07)

We're the intellectual, you know, leaders of advertising here. We know what a curation is. We know what a cookie is. And then Eric Suefer comes up, and he has, like, 40 slides and multiple, like, calculus slides that are talking about optimization of different ways to optimize things. And he's like, you know, it's obvious. You just take the derivative of this calculation to get the answer and then moves on to the next slide.

B (6:32)

No, Ari, it was not calculus. It was actual algorithms. Rich algorithms were on his slides. I. I mean, you know, you're more advanced than me on this stuff. I was like. But it was. It was amazing. And I learned so many new words. The satisficer's dilemma or the satisficer's regret.

A (6:53)

Satisficer's remorse is when you get what you want from a ROAS perspective as an advertiser. But you could have gotten more if platform wasn't taking a profit. So you. You're sort of, like, resentful at Meta for taking a profit on your roas, even though you're happy, you're satisfied with your roas, which was a great, great, like, insight. And I think what the end result was, everyone saying, well, if I ever have to build an ad platform, I know who to call.

B (7:22)

Exactly. Exactly.

A (7:24)

Right.

B (7:24)

You basically take Eric and James and, you know, you set them loose with some real budgets. You're good to go. What was your highlight?

A (7:32)

I really liked Olivia Corey from House. She went through the results of, I think 600, or I may be misquoting him, 600 Facebook campaigns, and compared whether Meta's ad system was better when it was on autopilot, AI only, or manual, and how they were different. And it was pretty interesting because it wasn't an obvious answer. It was like 40% better AI, 60% better manually. But the ways they were better were different. Effectively, the AI for Meta got to the answer a lot faster and harder, which is good if you have, like, an immediate conversion, like click to buy something, but not that good. If you have more consideration and, you know, multiple ways of getting to the conversion. The manual was better. So that was pretty fascinating. I'd never seen stats like that. And it's worth noting that we're going to over the next couple of weeks on Monday. You know, we normally on Monday publish like a vendor interview. We're going to have some of the content from the event kind of trickling out on Monday. So the next couple of weeks. So on the main stage, you should probably be able to hear some of these if you subscribe to our podcast feed.

B (8:41)

Awesome. Okay, great. And then, I mean the real highlight was the end with you doing the Fire Sabbath.

C (8:47)

Hurry.

B (8:47)

We're going to launch into that after we, after we did the news roundup. That was wild and great. And he broke some news like there was. There was a bunch there.

A (8:57)

Yeah, we'll queue that up after the news. Let's, let's jump into the news because it's a big week for earnings and other things.

B (9:03)

Yes. So let's take it from the top with earnings 3 of the MAG 7, the so called, you know, top companies that are driving disproportionate value in the market reported yesterday. So I'm going to go through them and then I'll pick out a couple of things that I think are important and relevant and we'll talk about it. And also Critter reported. So spend some time talking about that. In no particular order, we've got Meta. Total revenue up 26% Year over year $51 billion Quarter ad impressions up 14% Price per ad up 10% Reels Shout out to Rich Greenfield, friend of the pod. He pointed this out. Reels is on a $50 billion run rate. $50 billion run rate, which means it will be larger than the US television market soon, which is a $65 billion market that is on a decline. So Meta is just crushing on so many fronts. However, if you looked at the stock last night or this morning and by the way, this is not financial advice, the stock was down significantly because their CapEx estimates where they're spending on AI, they're spending on companies, they're spending on infrastructure. We all of this stuff was upped as well to $70 billion for the year. And I think people are getting concerned about the level of spending that Meta is doing to achieve these results and bring them forward into the AI future. So good quarter out of Meta overall got dinged on their spending. Second is Microsoft. Microsoft total revenue up 18% cloud which is where a lot of these spending by companies by Meta goes to up 26% relevant to architecture, search and news. Ad revenue up 16% xtac so solid quarter for advertising. Google total revenue up 16% first hundred billion dollar quarter.

A (11:02)

A hundred billion dollar quarter. That's crazy.

B (11:05)

Crazy. Cloud up 34%. YouTube revenue is a $10 billion quarter for Google. So, you know, YouTube, if you extrapolate that on a run rate, probably close to reels, right? A $50 billion market network revenue, ads revenue down. It was down, I think a couple hundred million to 7.3 billion. So overall the big company is doing well. Cloud is where a lot of the growth is. Advertising is solid critio. Total revenue up 2%. Gross profit up 10%. Market loved that. And retail media revenue up 11%. So a solid quarter out of Critia. Take a step back. What does this all tell us? What does this all tell you, Ari?

A (11:56)

Well, first of all, we're not seeing evidence of a recession, Right. A lot of people are talking about a weak Q3, weak Q4 in the AD business, like people I know, but the big guys aren't seeing it, which tells me it's a sector issue, not a macroeconomic issue or an open web issue. Yeah, that's what I meant by sector. Open web versus closed and open web seems to be taking on the chin because of the AI stuff. Second thing I would say is that the convergence, cloud and AI are very correlated. Right. They're not separate line items for these companies. And ads also kind of peeks into that list a little bit because you're going to have cloud customers who are going to want AI and the AI is going to need content and that content's going to need to be licensed from places and it gets like pretty interesting if you're Microsoft or Google and you're looking at that combination. Whereas Meta doesn't have that appeal. Meta is just like a very different company. They just have very different assets going into this. And it's remarkable how they continue to thrive despite not having some of the structural advantages that the other guys do.

B (13:08)

Because they're so far ahead in AI, it's just so clear, you know?

A (13:12)

Yeah, well, they're, they're Meta's. When you say they're ahead, do you think Meta's ahead in AI compared to.

B (13:18)

The other guys in AI driven advertising?

A (13:21)

Yeah, AI driven advertising, right.

B (13:23)

So one think of everything as an ad.

A (13:25)

Yeah, everything's one little, one little snippet that came out was that Mark Zuckerberg said that the annual run rate of Meta's end to end AI tools, I don't know exactly what that includes, surpass 60 billion. So that was. Or 15 billion in the quarter. So that's like a third of their total Revenue is from what Mark describes as end to end AI tools. I would say that's the largest revenue producing AI implementation in the world.

B (13:53)

Exactly. Which is what I meant. This is like advantage plus, this is SMBs just doing the here's my URL, here is my cost per X target, here's my $50. Have at it. Like this is the manifestation of what he's been talking about all year. Right.

A (14:12)

But if you think, if you think about AI in general, they're not a leader at all. Right. In terms of the models, in terms of consumer adoption, in terms of implementations, I think they're actually in some real jeopardy in the big picture because basically what makes Meta thrive is consumer usage of their products. And if consumers start seeing benefits to advanced AIs and something like Sora takes off or TikTok continues to take off, that's a challenge to Meta.

B (14:43)

Yeah, agreed. Which is again why they're spending so much money and they're acquiring companies or they're acquiring people for billion dollar sums to, you know, keep up in this race. So this is fascinating but, but their ability to turn this into ad revenue is unprecedented. There's nobody better.

A (15:03)

Yeah, I think we have an obligatory mention of Google which is network revenue was down again, so that's 7.3 billion. So the vestigial part of Google that nobody wants yet they refuse to spin out remains, remains in the doldrums. The other news that came out, I think Yesterday was that YouTube's doing a reorg. So Neil Mohan, I'll call Neil Mohan, friend of the pod. He hasn't come on exactly, but Neil Mohan announced reorg, some cuts and some folks shuffling around. Christian Oslien, who's I used to work with at Google and was an investor in Beeswax, he is now in charge of really all the paid products at YouTube. So that's a pretty big promotion for him. He was in charge of YouTube TV and did a bang up job on that. So interesting that, you know, YouTube's growing like crazy and they're cutting and reorgang at the same time. So that's pretty aggressive move.

B (16:03)

Yeah, I want to talk about that and we'll talk about that and connect it to the Amazon news. And unfortunate, you know, job cuts that started earlier this week and you know, just the stuff that's going around on, on X with some of the large companies cutting roles. So we'll find the link and put it in the chat in the newsletter. Again, it was kind of a late one, but if I recall it was Couched as we are reorgang to invest in the AI future with the implication that AI is going to start making its way into the content that you see on YouTube. Is that a fair description?

A (16:42)

I don't know. I don't know what that means exactly. It could be that you just need fewer people maintaining the whole business, selling ads and monitoring content, building tech. Or it could just be that, you know, tighten the belt so that you can grow. I'm not really sure what's going on.

B (17:02)

Yeah, yeah. And then Amazon, so this one, you know, we will put in the newsletter. They announced that they're beginning to cut 30,000 corporate roles. So these are people that you know, work for Amazon proper, not necessarily kind of in warehouses or other areas or 10% of the corporate workforce I think started with 14, 18% this week. And there was, you know, apparently, you know, you always hear this stuff but like pretty bad form where People got a 3am text message not to come in. You know, you're going to lose access. I've seen two explanations for this. The first is preparing for a leaner AI driven future and need for less people and you know, maybe a less extreme example of Elon coming into X, cutting 80% and things continue to operate. Or it's these companies just overhired in the ZURP era, in the COVID 2020-2022 era. And this is just rationalizing. Do you have a POV on this?

A (18:08)

Well, I listened to the Hard Fork podcast on Friday, last Friday and they had a lot of discussion about how serious Amazon is about what they called bending the curve on their workforce. Their workforce, especially in the warehouses, was expanding linearly with their sales and they didn't want that to continue. So they're using robotics and things like that. And I don't think that's exactly what's going on here because we, as you mentioned, it's corporate reductions. But philosophically they're very tightly run ship. They run like a retailer and retailers can have no fat. And you know, to the extent that AIs can write those six page memos, maybe you don't need the middle management layer. I'm joking a little bit. But like it's not an easy place to work. No one will say that. Anyone who you talk to in our industry or otherwise who's worked at Amazon will say, you know, it's a tough place to work. And I think they see employees as a resource that needs to be optimized.

B (19:04)

Yeah, yeah, I get it. Makes sense. Okay, well it's unfortunate for all of these people that are now looking for new roles in a seasonally difficult time to do it, I hope lots of people land on their feet. All right. Speaking of people, we had two big CRO changes or new CRO additions at some of the big public ad tech companies. The first was ttd. So Jed Dederick, who I don't know, but has a fantastic reputation, he's been there for over a decade and really built the revenue organization and engine. He is stepping down as CRO at ttd and a new CRO, Anders Mortensen, who comes out of Google, is stepping in. Do you know Anders?

A (19:50)

I don't.

B (19:52)

Neither do I. Yeah.

A (19:55)

And he seems like he's more from the search side than display, but he had a lot to do with agencies. His role involved agencies. He was just recently made a VP at Google. So that's a pretty big position.

B (20:09)

Big.

A (20:10)

It's a big deal.

B (20:11)

Yeah.

A (20:12)

No, it's not highly paid, etc. But we don't know why. But he, you know, the CRO role at a major public company is a pretty attractive job and Jed had been doing it for a while. He wasn't. Jed wasn't the first CRO. There have been multiple CROs at Trade Desk through the years. So, you know, good luck to Anders and good luck to Jed, I think. What do you think? Do you think this is a negative on the upcoming earnings release? I mean, that would be the easy call, right? The easy conventional wisdom is CRO leaves. It means you're going to have a bad quarter.

B (20:42)

Yeah. And I think that companies are smarter than that. So again, I have no knowledge. This is not financial advice. I don't know. I'm always trying to think differently than what the conventional thinking is. I don't know. I will say that bringing in new talent across the C level is what Jeff is doing. There's a new head of Corp Dev, Michael Gupton, who recently started. He came out of Walmart. Yeah. Yeah. He seems to be nearing the company for the next phase. So I wasn't surprised to see this. I was surprised, maybe not surprised to see a new chief customer officer for Criteo, Edward Dinshardt. Sorry, Edward. If I butchered the name, most recently I think was CRO at Triple Lift, but a strong Amazon background. So this seems to be a really good get for Critio as Michael builds out his senior team.

A (21:47)

Yeah, interesting. I don't know this fellow at all. Definitely the challenge of working with retailers and sustaining that momentum in that market is there big companies they have a lot of options for how to work. It's a SaaS sale. It's a long term relationship as opposed to Critia's primary business retargeting, which is a little more transactional. So it definitely seems like a challenge.

B (22:14)

Yeah, absolutely. Okay, where do we go from here? You want to talk about the Google civil trials?

A (22:20)

Yeah, just a quick update, but a meaningful update, which is in Google's antitrust cases, there's the Department of Justice which has won a monopoly and we're waiting for remedies. But there's also, I think at least six major civil trials on the same grounds as the ad tech DOJ trial. They're led by. The big one is in New York, it's led by Gannett, Daily Mail and News Corp. They're all consolidated into this big case and the judge gave summary judgment basically accepting all of the facts and judgments from the Virginia case into the New York case. So effectively they've been found guilty to have a monopoly in the civil case. There's. They're not going to be, they're not going to be debating whether what Google did was monopolistic or not. That's a just decided law. So that basically means the case immediately moves into the remedies phase, which in this case is a dollar amount, a very big dollar amount. And further, it's because it's antitrust, it's what's called treble. The damages, triple damages. So could be a big number, it could be in the billions. It probably not a number that will hurt Google's profitability, but definitely a number that could hurt, that could help those companies with their journalism and their generally financial health.

B (23:42)

All right, great. Then let's close it out with one more. So you had Dr. Mark Grether from PayPal ads on ST Architecture was an awesome conversation, but he didn't talk about this one, so we should spend a sec talking about this. PayPal is partnering with OpenAI for payments on ChatGPT. So OpenAI is making a bunch of announcements around commerce. Most recently they had Walmart going to be providing some shopping results in ChatGPT and now they're going to be partnering with PayPal to facilitate payments. This feels big.

A (24:16)

It does. Dr. Mark was on a Monday and they announced this on Tuesday. So my fault for doing a conference on a Monday. Yeah, so. So PayPal is just huge on the merchant side. It tends to be a little more long tail and it tends to be like a little European, but it's like, it's huge. I think he Said they had some 50, 30 million merchants on stage or something like that. So, so it's, it's also a lot of things that maybe aren't products that are available at like a Walmart like the SKUs are. It's a long tail of SK used as well. And so, you know, whereas Walmart gives them the feeds, gives them the shopping, this is a different set. It's also a checkout. So you could potentially be using PayPal to buy things that maybe PayPal doesn't necessarily rep. So I think this is a really, it's a good step for both of them. We're yet to see any volume of like real purchasing going on, but it seems pretty clear that that's an option if you give the consumers the ability to find the product.

B (25:15)

Agreed. The thing about OpenAI and if you just think through what some of the use cases might be for transaction, transactional answers. Right. Or you know, getting a product, getting an answer and then buying a product directly, it might benefit the long tail more than other parts.

A (25:34)

Right.

B (25:35)

Because you're just looking for, you're looking for a certain part, you're looking for a certain, you know, niche product. Like, you know, oftentimes that's where the answer engines thrive. So you can imagine this is a big one.

A (25:47)

Yeah, like, you know, find me a T shirt that has a funny saying about, you know, some event or something like, you know, those kind of weird things that you might normally go to Etsy for, you know, might be really good use case travel.

B (26:00)

Diy. DIY is incredible. Like I just take pictures of a thing I'm looking to fix in the house and you know, I typically get a link to Home Depot and it could be just like a 10 cent washer, but I would never be able to find this without that. So I can imagine the possibilities are like pretty substantial.

A (26:18)

I think he just repeated the exact plot of the Home Depot TV ads. Like that is what they do in the TV ads. They're like, they show them using the Home Depot app and be like, I don't know how to fix this. And then boom, it shows up.

B (26:30)

Home Depot, I know you have an rmn, come talk to us.

A (26:35)

Exactly, exactly. Cool. Well, should we wrap and introduce the Terry segment?

B (26:43)

Let's do it. We should not wait any longer for the violence that is about to come.

A (26:49)

So let's try to set this up a little bit. If you're listening to this, you probably know who Terry Kawaja is. He's probably the leading investment bank in our space. He is Known for his antics, his presentations, his songs, his comedy, etc. He came on. He was the final event of Mark Dutcher Live. It was supposed to be. It was billed as a podcast recording with me. And that's not really what it turned into. You're not going to hear my voice very much for the next 20 minutes. I don't know what got into Terry, but, man, did he go off. The audience could not believe what they were saying. They were laughing, they're crying. He just goes on this monologue that is really worth putting on, at least off Broadway. What was your reaction to what you saw?

B (27:39)

Yeah, I think that was a good summary. The big thing I took away from this, and I think the listener is going to as well, is it feels like Terry has been waiting for this moment that is in front of us right now with AI. You know, Luma. Luma helped architect much of Programmatic through deals like they were behind the Admiral and invite media acquisitions for Google. They were part of, like, you know, a lot of the foundational deals for Programmatic. My sense from Terry is that the chessboard is going to be completely reset. He's got a AI Lumascape that's going to be coming out over the coming weeks, and he believes this is going to be a time of not just company formation, which you and I are very focused on with our work, but company acquisition. So feels like he's getting ready to. And he even said it. He thinks he's going to have a big 2026. He thinks the space is going to have a big 2026 driven by M and A. So feels like he's just like, getting the. Getting the engine revved up.

A (28:50)

Yeah, exactly. And he revved up our audience quite a bit. People during the cocktail were like, that was the best way to end this. So we hope it translates to podcasting. It's also on YouTube if you want to watch his animated expressions and everything. It's a high energy. You will not hear me very much. You'll hear me, like, say, excuse me, I need something to say. Excuse me, let me jump in here. But I do not get many words in. We hope you enjoy it. So without further ado, let's go to Terry Kwaja's talk at Architecture Live and we will see you next week.

D (29:23)

Adlib is the DSP agnostic buying platform that helps agencies and marketers scale Programmatic, smarter, faster and more profitably run premium campaigns across CTV Audio, Geo Fencing and YouTube. All under one roof with zero trafficking headaches, no bloated fees, no complexity, just enterprise grade performance. Performance built for growth. And ad lib isn't just a platform. They are your partner. Their team helps you win new business, expand budgets, improve ROI that keeps clients coming back. Unlock programmatic and getadlib.com G-E-T a D L I B dot com.

A (30:10)

All right, so we booked this session as no bosses, no filter. Right. So that should be fun.

C (30:18)

Neither of us have a boss in. Neither of us have a filter. So good luck.

A (30:22)

That is true. And you're better at social media than me, so I bring you on to amplify. All right, so start with some quick questions. You gave a thumbs up. Buy, sell, or hold trade desk stock right now.

C (30:37)

Oh, so because I'm a registered, I can't answer that question. Nope.

A (30:44)

Okay. That's a good answer. All right, second question. Other than Roku, Take Roku off the table.

C (30:49)

Yep.

A (30:50)

One company they could buy. Give me a name. You'd advise Jeff to buy right now?

C (30:57)

Sure, why not? So what's the whole theme of this conference? Outcomes. Right. And. Yeah. Quality. Good luck. Yes. By the way. Yes. I would never say no to quality. You're 100% right. Right. It's an and not. Or. Here's the thing. I believe AI is going to accelerate the whole move to outcomes. I also think we need to reframe what we mean by outcomes. We do not necessarily mean, you know, direct response, bottom of the funnel, you know, final step before conversion. No, Outcomes is a full. As Lou Pascalis could tell you, it's a full funnel phenomena. And I think that they are missing a big opportunity in performance. So what company should they buy?

A (31:51)

You get paid on a transaction, not on a word. Right.

C (31:58)

Maloco.

A (32:00)

Maloco. All right, that is a. Wow. That's a curve ball, but it resonates. Okay, now I want to hear why. Because it's about performance. Right. It's about outcomes.

C (32:13)

A company that has done well in performance. And by the way, you know, if you'd asked me this question five years ago, I'd have said applovin.

A (32:25)

All right?

C (32:25)

They are now at $220 billion market cap. Here's the thing. It is easy. You know the old biblical expression, it's easier for camel to pass through the eye of a needle than a rich man to get into heaven. It's easier for a performance ad network to address the general CPM marketing world than it is for the other way around.

A (32:48)

Yeah, I think that's accurate. Side note, some people think that's a mistranslation because it makes A lot more sense that it should be a camel hair, as in a thread to get through the eye of a needle than a camel.

C (32:59)

Except a camel thread. It's easy to get through a knife and needle and a camel is impossible.

A (33:02)

I don't know. All right, so let's talk about performance. So do you think if let's put M and A aside, I know that's not. That's against your religion, but ain't gonna happen. Do you think someone like the trade desk could build a performance engine on open web?

C (33:18)

I think it's very difficult. I think the people who built. If you look at all the scaled performance companies which I would list as, you know, like Applovin and Facebook are like 1 and 2 or 2 and 1. Right. In terms of scale, where did they learn it? It goes back to gaming and performance advertising. It has to be in your blood and I don't know, it would take a sea change. Look, when I've pitched MA to Jeff and Jeff has been very transparent about the. He said, you know, culturally he worries about cultural fit of acquisitions, which is fair, that can be a downside. But he also says we're not ready for it. I don't think we have the ability to integrate. And that's all fair. And by the way, I have said four years, dude, you guys have done so well organically. They've defied the law of large numbers, which is to say the bigger they got, the faster they grew. Which is kind of like not usually how it works, right? Usually you get this sort of like declining curve slope asymptotically. And so I think the culture of a performance outfit, intermediary is probably different. So I think it would be hard.

A (34:35)

Yeah, I agree. Culture is a big issue here. The customers don't get transparency, they don't get as much control. And the customers of the trade desk ask for those things all the time.

C (34:45)

Well, it's interesting you say transparency because, you know, I have attended events, whether it be IAB events or like the ANA where all the big marketers go, and for a decade all I have heard is big CMOs standing up there, Mark Pritchard, a whole list of them all saying transparency, transparency. Well, it turns out that if you look at Google, they have a product called Performance max which is 100% opaque. One could not get a less transparent solution in the marketplace that they offer their E commerce clients. How many of their E commerce clients have taken them up and utilized Performance Max instead of the manual stuff? 82%. Their clients have voted with their wallets. They don't care about transparency.

A (35:45)

Yeah, the call for transparency, in a sense is a symptom of not being happy with the performance.

C (35:51)

Correct? Well, and remember, in some cases, and remember, there's other things too, because transparent people want transparency. Because, you know, if you're Mark Pritchard and you are the shepherd of an incredibly valuable brand like P and G, you want transparency because you don't want to know if your ads are running on some screwy site, some MFA site. And so I get that part of it. But here's an interesting observation that I said 10 years ago and I don't think anybody groked, which is when you move from, like cpa, when you stop charging on impressions and charge only on outcomes, not only is it a better business model for the intermediary, it leads to better results for the advertiser, but it also mitigates all the issues that people talk about, fraud and viewability. Well, last time I checked, a bot doesn't make purchases. A bot doesn't have a wallet. So who cares if your ads are poorly viewed or there's bots viewing them that are getting measured in someone's vastly outdated measurement system? If you sell on a CPA or a cpi, none of that matters.

A (37:07)

Yeah, the word transparency is also used like a cudgel because it does mean things around brand safety, which is really.

C (37:12)

Important, which is 100% legit.

A (37:14)

But when they apply it to pricing or take rates, it's pretty disingenuous because they don't demand it from channels of the that produced results and they don't.

C (37:26)

Demand it, by the way, in like. So we're talking about how a big corporate is insisting on transparency for their media buys. I got news. Proctor and Gamble. I don't know what the percentage is that Procter and Gamble spends on. I don't know why I'm taking that as an example, but I don't know what percentage advertising spends of their total expenses of their corporation. But I'm going to go ahead and suggest that it's a really, really small percentage. And I don't think Procter and Gamble's procurement people for, I don't know, office cleaning, where they demand transparency of their janitors. Like, it's just not the same.

A (38:04)

It'd be pretty bad if your janitors weren't transparent, though. So let's talk about AI. So you put out some content about AI. I don't remember exactly what it was called, but it was a calling. Talking about how many companies are laggards in AI among this community, the ad tech community. Remind me what, what that was titled.

C (38:26)

So, so I. We're preparing to release, to debut the AI Lumascape. By the way, I swore I'd never do an AI Lumascape. And people said, why not? I said, well, because there's no Lumiscape for electricity or cloud or any like. In other words, AI is going to be intel inside la. Like isn't everyone going to have AI? I don't understand but you know, give the people what they want. So we're doing the freaking impossible task of creating an AI Lumascape. Now riddle me this, Batman. How do I exclude a company on the AI Lumascape? Every single company on every advertising Lumascape is going to insist on appearing on the AI Lumascape. So wait, are we just talking logo sizes here? So we have a challenge and I have a crew that's been working on this. We'll have had four months of intense work. And so what I did, like any lumascape, here's how it works. You do the research, primary and secondary. And by the way, shout out to the unbelievably great trade publications we have in this industry. Can I get a hell yeah for that? I mean we dissect every aspect and it's. And including marketing and needed in the marketplace. But so it's. I'm aided in this, in this research, but we're out there trying to figure out these solutions, blah, blah, blah. And of course I'm not a technologist, so you know, I don't know, could probably use your help. But here's the thing to do it. We usually get about 80% of the way there and then we publish it with the caveat that we. There's one thing I can guarantee about every Lumascape. When we publish it, it's wrong.

A (40:15)

Complaints.

C (40:16)

Well, no, but we say it's wrong and by definition we do not wait till it's 100% because that would be dumb. Be a bad use of our time. So we publish when we get to 80, 85% and then the marketplace, the ecosystem will tell us in about 30 minutes. It's amazing how my inbox fills up. You missed us, asshole. That's something some people actually are stupid enough to say that.

A (40:44)

Okay, fine, whatever do you take finally to that?

C (40:47)

I don't care. I don't care. You could listen. I've been called everything.

A (40:50)

How did Brian Okelley get his own box on the original Lumascape?

C (40:54)

Because that was me with him.

A (40:56)

It was reverse engineering.

C (40:58)

It Was reverse engineering. So most of what I do is. Has some relation to comedy or tweaking people. I tweak you all the time.

A (41:10)

Easy, easy mark, right?

C (41:11)

You're low mark. I mean, oh, Kelly is so easy. Anyway, so no, he complained when I put him in a box. He complained, said I put him in another box. He complained. I'm like, you, I'll put you in your own box. And then you go figure it out. And I've heard him talk about this. I was like, yeah, Terry put me in my own box. And that cost a lot. In therapy for years.

A (41:31)

Oh, I thought he was proud of his own box.

C (41:33)

Sorry.

A (41:33)

I thought he was proud of.

C (41:34)

He was proud and not proud. He's like. I said, he. He went into therapy.

A (41:38)

Let's get back to AI.

C (41:39)

So, so, so I. I called, I said, look, there's. Near as I can tell, and the formulation of this AI Love Escape, there's three types of companies. There's AI First, Right. We saw a few of them up here. There's fast adapters. That is to say, companies that either managed to find quality developers, built the product, got traction, and are making revenue from it. So for all of you going, we're a fast adapter. Yeah. Well, show me your financials. Or made an acquisition of an AI first company. And then there's everybody else, which I labeled AI posers and I, you know, intentionally derogatory tweaking. And what it did was it created reverse inquiry, as they always do. And so now we have every single company in ad tech reaching out to us to try and demonstrate their bona fides in AI. And here's the funny thing. They all come in with a big smile on their face. You can't wait to show you this. They do a demo, and half the time now we've seen like a thousand demos in the last four months. And most of them, I'm sitting there going, that it?

A (42:42)

Yeah. I mean, we were sitting. Terry, you and I were sitting right over there when we saw the startups present, and we're like, this is a lot better than the hundred million revenue companies we deal with every day.

C (42:53)

I leaned over and I said, I like this Lucy thing. This works. And I turned to him and I said. I introduced myself and said, I see an exit in your future, young man.

A (43:04)

All right, let's talk about that. So. So do you think there'll be a wave of like, let's buy the AI talent, let's buy the early companies?

C (43:11)

Yeah, we're already seeing it. So I think what you're going to see as a consequence of this sort of re steepness in the learning curve, in the innovation curve is you're going to see companies get acquired much earlier that will have raised less capital and less people. My bhagwat is to sell a single digit person company for a billion dollars. It may not be achievable, but if I come close, you know, a 24 person company for $800 million.

A (43:41)

Oh well, it feels as though, though I'm going to push back a little bit on that. It feels as though if you're selling to other companies on the Lumascape, there's a limit to the valuations. There's not a lot of crazy numbers. On the other hand, if you got Sam Altman wanting to build an ad tech company, there might be some opportunity there, there.

C (44:00)

So Sam Altman is building the biggest ad company in the history of mankind. He has an extremely highly qualified.

A (44:12)

Team.

C (44:13)

From Facebook Team Fiji led by Fiji Simo, who is the Sheryl Sandberg of OpenAI and I've known Fiji for 12, 13 years, very capable. She's actually hiring a person too because she's all applications. So advertising is just part of her purview. She truly is the number two there. And the job wreck that she's got out is the hottest job in the history of ad tech.

A (44:43)

All right, I'll have to apply later but okay. Same question I asked you about Trade desk. If you got an audience audience with Fiji, what's the one company you recommend they buy?

C (44:54)

Ah, I'm actually having that conversation with her. So I'm gonna actually, I'm gonna. If you don't mind, I'm gonna, I'm gonna divert. Divert that one. Here's the point. It won't be what I think there's less of a rationale for them buying existing companies, maybe others. It's a few capabilities here or there. But I made an observation. So. So, so I think we're entering the fifth phase of ad tech. Okay. This is like a new epoch and by the way, I've been the one I take pride in not just getting calling things, whether it's commerce media or the democratization of tv, all these sort of the fate of agencies back in 2015. Done a lot of those things and got them right. That's good. That's calling strikes and getting a strike strike. But I'm particularly proud of calling balls and then watching balls happen. And by balls I mean 5G is going to change everything Iot because we need fricking smart Toasters, you know, virtual reality. Because everyone's going to walk around with a stupid headset. I think Scott Galloway nailed it. He's like, that is. That's one way to make sure you don't get laid. Or the big daddy of them all. Wait, Web three. Oh, my God, what a stinking pile of dog poop that was. And yet, undisciplined observers got on the bandwagon because we love the shiny new object in marketing. Everyone, not everyone, but a lot of people were like, yeah, Web3, we're building a Web3 solution. I stood up at our digital media summit and said, this is bunk. Not the cryptos bunk. Not the blockchain's bunk. Those are real technologies. But their application to advertising, hell no. And I made a glossary where I stood up in front of 3,000 marketers at the ANA and I said, by all means, invest in Web3 if you want to fly around to fancy conferences and hear yourself chin wag. But let me tell you something. 0 was the definition of. 0 was the amount of money you will return, you will get on every dollar invested in Web3. It's categorical. So anyway, so with that as a backdrop, I do think AI is the big one. What the fuck was the original question?

A (47:10)

I don't recall. I don't recall for so long. I just remember you're the expert in balls.

C (47:14)

Goddamn. Oh, right. So, right. No, no, no. Fifth phase. Fifth phase of ad tech. So I'm gonna do this really fast. 2005 to 2010, I call that the ad network era. And then 2010, 2015, that was early programmatic where we went from what were bootstrapped companies doing arbitrage with an ad network model. We went now to a venture funded, either demand or supply, getting fees on take rates. Then in 2015, we actually the Lumascape hit peak fragmentation. That was the end of the VC era and the start of the private equity era. And so from 2000, where am I? 2015 to 2020, I call that sector headwinds. So that's phase three, sector headwinds, walled garden dominance. And we had the rise of CTV and Commerce media. But then 2015 to 2020, I call that scale and PE, because we now started to see really big companies and a lot of private equity investing in the sector. And yet, yes, we did see growth of commerce Media and CTV and mobile app, however. And we had the cookie circle jerk, I call it. I hope that's not offensive. Someone said that was offensive. I'm like, no, no. What happened was Offensive. So if you tie my comment about it as offensive, that's appropriate I think. Anyway, you know, Google fucked us over with the cookie thing.

A (48:48)

Sandbox was the sponsor of last market sector live. Spill out one for them.

C (48:52)

You know, I still have sand in my shoes from the last time I spoke to Alex.

B (48:56)

He's here.

A (48:56)

Alex is here.

C (48:57)

Yeah, Alex is here. Alex. Great guy.

A (49:01)

Unemployed if you need.

C (49:02)

And we had an M and A and an ipo. Boom and then a bust. But here's the thing. Over the last 10 years, I would make the following observation about our industry that we love. Innovation definitely waned. It was not like the early years. We were coming out with all kinds of new solutions. And every year when I had to do the state of digital, I was like, there's less and less things to talk about that are new. In other words, yes, more ctv, yes, more commerce, media proliferation, blah blah, blah. But it was basically a continuation of pre existing trends. That was a problem for me. It was a problem for me because Luma is a little company that's got highly specialized knowledge that is trying to compete against the big investment banks. Well, when you're talking about companies in the hundreds of millions of dollars in oodles and oodles of ebitda, the only thing you got to do when you're selling that company is differentiate between getting nine times EBITDA and eight and a half times ebitda. So my comparative advantage went away. And so where I thrive is when there's steep learning curves and lots of tech innovation. And I can help simplify what is otherwise a complex story. And for an acquirer to have the gumpt to make a scaled acquisition and risk not only their company's balance sheet, but their careers, they need comfort that this is real. And so I point to my unblemished track record of every bullshit stupid idea that's come up. I have said this is a bullshit stupid idea in real time. And then it manifested into a bullshit stupid idea. And the ones that are so that gives you, me, my North Star, which is objective credibility. So I'm so excited for fifth phase of ad tech. By the way, you're going to see all of this. I have a presentation. I came to the realization that I needed to come up with the ultimate AI and advertising presentation. Because this is existential for Luma, by the way, I think this is existential for my business. Every single person in this room should feel exactly the same, because it is. Dave Eisenberg can tell you I walked him through some of the Slides. This is it. This is the big one, which is why we're big hosting a big event. We're gonna debut the AI Lumascape. We're gonna come out, you're gonna be seeing me talk about this deck ad nauseam. You're gonna be sick of me. But what we're gonna do is after November 17, when we hold our Futurescape event, we are gonna be piecemealing it out. I'll also be sending it out. You'll see videos, you're gonna see shorts, you're gonna see TV TikTok. You're gonna see God damn everything. Why? Because I have to end up as the default advisor in the intersection of AI and advertising. That's my commercial.

A (51:46)

Praise Jesus. All right. Okay, so what if you were calling a ball right now? What's the nonsense that's going on? What's not a factor in AI?

C (52:08)

Just right now? So I find this is an industry that does a little bit too much navel gazing. So, like, you know this transaction ID thing, by the way, that's as much as I learned about it. There's a fight. I don't know what transaction IDs are. There's very smart people in this room that know exactly what it is. I don't care. So we do too much.

A (52:29)

That's not a ball.

C (52:30)

That's just like, no, no, that's bad focus. We spend a lot of time going back and forth on all these small ball issues. I'm like, guys, really? And gals, I'll tell you.

A (52:39)

Do you think consumers are going to shop using AI, like agents that will actually buy things for them?

C (52:46)

You know what? At the very beginning of this deck that I have about AI and advertising, I say something a lot. When I get asked questions about AI that I. I think anybody responsible should say with frequency about AI. Three words. I don't know. I think it's so damn early for people to make sort of declaratory statements. I'm making predictions. And then I caveat with but I really don't know. I don't know. They're like, well, you don't know. Yeah, how the hell do I know? And you know, when I frame AI's application to advertising, I do it using Maslow's hierarchy of needs, with the base layer being, you know, audience data and workflow and media buying optimization and measurement. Because that's the meat and potatoes of ad tech. And that's going to have a huge impact on that. And by the way, less headcount. The next step is ad creative and content and communication and I think that's like a higher order thing. Just like Maslow's original. The higher up you go, the more disruption, the more impact. That's going to be very interesting. Here's the thing for the last 25 years, you know what the navigation pattern we've decided was best is? We find someone wherever they are on the Internet and we drive them to a website so we can convert them. Seriously, like in this world that we're in, you take your message to people wherever they are. If they're doom scrolling on TikTok, leave them on TikTok, get your message in front of them and allow them to convert and stay on TikTok. Stop trying to shepherd consumers in the hundreds of millions and billions all around the Internet by going to what your www? How the. No fucking way. Like that's crazy.

A (54:28)

It's actually a controversial topic and our earlier speaker, Eric Sofer has written about.

C (54:32)

It a lot because I didn't understand anything.

A (54:34)

No, I understand, but his writing, don't worry about that.

C (54:37)

That dude. So seriously, quantity and quality of. I do not know how he does it. Eric Suford, you are the real ad tech guy.

A (54:46)

He's like me, if I cared or worked hard anymore.

C (54:51)

You might be a little generous there on the IQ points but.

A (54:58)

The point.

C (54:58)

I was making, I mean seriously, he is so high, his IQ is so high. Like, you know, has a little trouble sometimes interacting with humans, but that's what happens, right? He's so damn smart. But by the way, if you don't subscribe mobile dev memo subscribe everyone in the room. It's unbelievable.

A (55:14)

All right, last question. So tell us a little bit about Terry the person. What's your day like, what do you do all day? You get in what's like a day at Luma like.

C (55:29)

Yeah, I work from home until about 10. I like, I get up plenty early. 6, 36, 45. And I have a routine that I like. Routines. I work out every morning because Eric Franchi is my North Star. No, I work out every morning. I just respect him. And I like to have the breakfast that I like. So if you invite me to a business breakfast, I broke the wrong rule for this man because I love live ramp. But I don't do breakfast, business breakfast, generally speaking, because I like my routine. I work out, I have my breakfast. I probably will make like a meme or two, whatever comes to me. Nothing is planned. So you know, and most memes I produce I do under 60 seconds. That's my rule. If I can't get it done in 60 seconds. I got, I got work to do. Then like I'll answer email, blah blah blah, and then I'll go into the office is mid morning, by the way. You know, I live in the Upper east side, I take the subway to work and you know, just like a normal person. Yeah, but, but I do. So I'm 63, so I'm not a young man anymore. Which is not to say like, not physically. I'm saying I, I refuse to suffer the indignities of being, you know, sandwiched up like this in the subway. So that's why I do my mid morning thing and then I'm in the office and it's mostly meetings. We do work. We do, we, you know, we punch way above our weight like market and we pump out a lot of content and a lot of work product for a small mighty team. And that means either actively working on live deals and we've got a handful of them now, by the way. It's really ramping up. Watch 2026 Bueno as it relates to activity. And then there's a lot of like content. We host events, we publish content, we've got the quarterly report. We do, you know, financial content that's totally serious. I don't make any jokes there. And then the other stuff. And by the way, by the way, in case you were asking, you didn't. But I'll answer it anyways, you know, because this new AI deck, there's a lot of loose in it. There's a lot of animations and sound effects and graphics and I use a lot of humor. And people say, well, I guess it's because you're a ham. And yes, they would not be wrong. But there's a business purpose for why I do that. I believe every time you got one of these, a microphone and a crowd, you have a responsibility. But also I have two objectives. When I stand in front of you, those two objectives are I want to capture your attention and your retention. So there are plenty of phones, there's plenty of laptops. I've seen the speakers. Bright stage, dark room. Everyone's just getting email done, getting work done and occasionally looks up. Not for me. You know why? It's because I'll have some animation with a sound effect and everyone laughs. And if they're down on their, they're like, shit, I missed it. The second one, forget it. They are paying attention. So when I give this AI presentation, not a single phone is out, not a single laptop. So I got their attention. And then to get their retention. I don't use words. I use graphics. Why? Well, you just saw it with the trade desk. A graphic quite, quite literally opens the cabinet. First of all, comedy opens the capillaries of your mind. So great. It's the same as LSD. And a graphic is comprehended 60 times faster than the written or spoken word. 60 times. And as a bonus, you store it in a different part of the brain. So I'm in long term storage back here, whereas everyone else who talks, they're in ram in and out.

A (59:26)

Unfortunately, we're recording a podcast, which is entirely worse, so. So with that, we're out of time.

C (59:33)

Can you believe. You know what? I 100% forgot that.

A (59:37)

I know. It's really shocking. And the things you said were pretty shocking too. No. So, everyone, let's thank Terry for his amazing conversation.

B (59:51)

Thank you for subscribing to marketecture.

A (59:53)

New interviews are added every week at marketecture tv. And your favorite podcasting.

B (59:58)

Yeah.

A (60:07)

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.