Episode 98: Marc Guldimann explains everything about attention

Summary

Marketecture Podcast Episode 98: Marc Guldimann Explains Everything About Attention

Hosted by: Ari Pipero and Eric Franci
Guest: Marc Guldimann, CEO of Adelaide Metrics
Release Date: November 15, 2024
Podcast Link: marketecture.tv

---

Introduction

In Episode 98 of the Marketecture Podcast, hosts Ari Pipero and Eric Franci engage in an enlightening conversation with Marc Guldimann, the CEO of Adelaide Metrics. The primary focus of this episode centers on "attention" within the advertising and marketing industries—a topic gaining significant traction in 2024. Despite Adelaide Metrics being a vendor in the space, the hosts sought a comprehensive 101 tutorial from Marc to demystify attention and its metrics.

---

1. The Rising Importance of Attention Metrics

A Shift from Traditional Metrics: Attention metrics have surged in prominence as industry professionals seek more reliable indicators of media quality. Marc emphasizes that not all media is created equal, and existing metrics like viewability have been gamed, leading to compromised quality.

Marc Guldimann [04:00]: "A lot of the metrics that are used by the market today have been fully gamed, like viewability. If you optimize just like whole hog towards the most viewability and the lowest CPM, you're going to get some pretty low-quality media."

Attention vs. Viewability: While viewability measures whether an ad is seen, it doesn't account for the quality of that view. Attention metrics aim to fill this gap by providing a deeper understanding of how audiences engage with media.

Marc Guldimann [04:06]: "If the metric is transparent or if the weights on the different sort of measures are more opaque, that's definitely something that we have a strong opinion on."

---

2. The Four Waves of Attention Metrics

Marc outlines four distinct waves in the evolution of attention metrics, each representing advancements in how attention is measured and utilized.

Wave 1: Viewability and Video Completion Rate

• Description: The initial phase where viewability and video completion rates served as proxies for attention.
• Insight: While these metrics provided a basic understanding, they lacked depth in assessing true engagement.

Wave 2: Viewability Plus

• Description: Extending viewability by tracking the duration an ad remains on the screen.
• Insight: Longer display times can indicate higher engagement but don't account for contextual quality.

Wave 3: Eye Tracking and Gaze Duration

• Description: Incorporating eye-tracking technology to predict gaze duration.
• Insight: Offers a more accurate measure of attention but solely focuses on gaze without linking to business outcomes.

Marc Guldimann [07:23]: "The dependent variable in the third wave of attention metrics is gaze duration and not business outcomes."

Wave 4: Probability of Attention

• Description: Shifting focus from maximizing attention to the probability of attention, ensuring higher-quality engagements.
• Insight: Aligns attention metrics with actual business outcomes, making them more actionable and reliable.

Marc Guldimann [09:42]: "The biggest risk an advertiser has is no attention, and that's what they need to mitigate."

---

3. Methodologies for Measuring Attention

Importance of Eye Tracking Data: Marc underscores the necessity of eye-tracking data to genuinely comprehend human attention in media consumption. Unlike methods that infer attention through indirect signals like mouse movements, eye tracking provides direct insights.

Marc Guldimann [13:11]: "Everybody should use eye tracking data. You need to understand human attention."

Outcome-Based Metrics: The ultimate goal is to tie attention metrics to business outcomes. This involves a rigorous process of research, data collection, modeling, and validation against real-world performance indicators.

Marc Guldimann [15:32]: "The job of advertising is not to drive attention, it's to drive actual business results."

Challenges with Alternative Methods: Methods relying solely on JavaScript or other inferential techniques lack the precision and reliability of eye-tracking data, potentially leading to inaccurate attention assessments.

---

4. The Competitive Landscape and Market Challenges

Proliferation of Attention Startups: As of the episode, there are over a dozen startups vying in the attention metrics space. Marc highlights the competitive nature as both a catalyst for innovation and a barrier due to the prevalence of subpar metrics.

Marc Guldimann [06:38]: "There's a few launching. It seems like every week. It's a crowded space."

Quality Control and Accreditation: The market's growth is threatened by "bad attention metrics." To combat this, Adelaide Metrics and peers strive for accreditation and transparency in their methodologies to build trust.

Marc Guldimann [17:10]: "We've had our product under review for accreditation... for everything from web to digital, out of home to CTV and cinema."

Market-Driven Standards: Adelaide Metrics is working with the IAB to establish industry standards, ensuring that attention metrics align with both research and tangible business outcomes.

---

5. The Future: Attention Metrics as a Currency

Market Adoption and Guarantees: Marc envisions attention metrics evolving into a currency-like standard within the advertising ecosystem. This would involve media placements being rated and guaranteed based on their attention metrics, facilitating trust and predictability.

Marc Guldimann [19:15]: "There are like A and E, for a while was making guarantees on TV data. So it's definitely starting to happen."

Case Study: Wall Street Journal: The Wall Street Journal's adoption of Adelaide Metrics' data to guarantee attention levels exemplifies this shift towards attention as a market-driven currency.

Marc Guldimann [20:45]: "The Wall Street Journal announced recently that they're making guarantees based on our data."

Long-Term Vision: As attention metrics become standardized and widely accepted, they will reshape how media buys are conducted, moving beyond traditional metrics to more nuanced and outcome-oriented measurements.

---

6. Proving the Link Between Attention and Outcomes

Evidence of Efficacy: Despite some claims to the contrary, Adelaide Metrics and competitors like Lumen possess substantial proof that attention-driven strategies lead to better business outcomes compared to traditional metrics.

Marc Guldimann [16:29]: "There's a large body of work that proves that attention drives outcomes."

Addressing Skepticism: Marc acknowledges counterclaims, such as Kroger's unsuccessful test using DoubleVerify's metrics, attributing failures to the quality and data sources of the metrics employed rather than the concept itself.

Marc Guldimann [16:34]: "It depends on the vendor you use... I don't know how much eye tracking data they use."

---

Notable Quotes with Timestamps

• [04:00] Marc Guldimann: "If you optimize just like whole hog towards the most viewability and the lowest CPM, you're going to get some pretty low-quality media."

• [07:23] Marc Guldimann: "The dependent variable in the third wave of attention metrics is gaze duration and not business outcomes."

• [09:42] Marc Guldimann: "The biggest risk an advertiser has is no attention, and that's what they need to mitigate."

• [15:32] Marc Guldimann: "The job of advertising is not to drive attention, it's to drive actual business results."

• [19:15] Marc Guldimann: "There are like A and E, for a while was making guarantees on TV data. So it's definitely starting to happen."

• [20:45] Marc Guldimann: "The Wall Street Journal announced recently that they're making guarantees based on our data."

---

Conclusion

Episode 98 of the Marketecture Podcast offers a comprehensive exploration of attention metrics, their evolution, and their pivotal role in modern advertising. Through Marc Guldimann's expert insights, listeners gain a nuanced understanding of why attention metrics surpass traditional measures like viewability, the methodologies underpinning accurate attention assessment, and the future trajectory of these metrics as foundational standards within the industry. This episode serves as an invaluable resource for marketing and advertising professionals seeking to stay ahead in a rapidly evolving landscape.

---

Disclaimer:
This summary is crafted based on the provided transcript excerpts and may not encompass all discussions from the full episode. For an in-depth understanding, listeners are encouraged to tune into the full podcast episode.

Transcript

[Ari Pipero]

This podcast is brought to you by sao, a leading CTV advertising platform purpose built to help brands effectively reach, engage and validate streaming audiences. For over a decade, SAO's platform has combined deep industry expertise with innovative technology to meet the needs of an evolving media landscape ready to unlock the power of CTV advertising. Learn more at SAO WS Podcast. That's SAO WS Slash Podcast. This podcast is brought to you by the Current. Ever wanted to sit down to a candid conversation with the marketing leaders of the world's biggest brands? The Current podcast is your chance. On the current podcast you'll find exclusive interviews with the experts and trendsetters who are on the front lines of digital advertising. And they always leave the ad tech jargon at the door. Subscribe to the [e-mail] or anywhere you get your podcast today. Welcome to the Market Podcast. I'm Ari Pipero, I'm here with Eric Franci and we have a really exciting show today. So we have Mark Golderman who is the CEO of Adelaide. And so Mark is here because we were really interested in hearing more about attention and attention metrics. So even though he's vendor in the space, we were asking him to give us kind of a 101 tutorial about everything relating to attention, which I'm really interested in. Eric, what are you kind of looking forward to here?

[Eric Franci]

Yeah, I'm looking forward to talking to Mark about a few things. So number one, attention in 2024 has gotten super hot. So apparently there's like a couple dozen attention startups out there. Adelaide being I think one of the ones that's most well known and a leader. So I want to know like, you know, why attention matters all of a sudden.

[Ari Pipero]

Right.

[Eric Franci]

And then also into the methodologies because I was this being measured, you know, what are the different types of attention, what are the technologies used to like understand attention, how it connects to outcomes? Like I'm just looking forward to a deep dive with the guy and obviously as an investor I've spent a bunch of time and he thinks about this stuff in a really deep way.

[Ari Pipero]

Yeah. And there's also a lot of news this week. There's some earnings with Magnite talking about Netflix, how important they're going to be. We also a late breaking have a short seller going after Z Zeta and Vian acquired Iris tv. So we're going to have a lot to talk about. Quick Housekeeping. So this week's vendor interview was with ShopSense AI, which is a really interesting company. You're an Investor in that too, Eric, right?

[Eric Franci]

We are, yeah.

[Ari Pipero]

Okay, so ShopSense AI, they basically enable broadcasters to have second screen commerce. So you can basically, while you're watching TV, they have a website that you can buy stuff. It's pretty cool. They're doing really well. Also this week, AdTechGod does a podcast with the Vanderhook brothers to hear about the Viant Iris TV acquisition. So AdTechGod is stepping on my toes. He's starting to break news, not just doing his fun meme stuff, so good for him, but it's definitely worth listening to. All right, let's get into it. We'll have Mark on the show. All right, Mark Goldman, the CEO of Adelaide Metrics, thank you so much for being here.

[Mark Golderman]

Thanks for having me.

[Ari Pipero]

Ari Adelaide has been a pretty frequent advertiser on architecture, but this is not paid content. We really wanted him on here to give us his perspective on attention because attention is kind of a buzzw. So why do people care about attention?

[Mark Golderman]

So I think that people care about attention, at least in media, because all media isn't equal. And the metrics at their disposal today don't really do a good job of helping them suss out what's high quality and what's low quality. In fact, a lot of the metrics that are used by the market today have been fully gamed, like viewability. And if you optimize just like whole hog towards the most viewability and the lowest cpm, you're going to get some pretty low quality media.

[Ari Pipero]

Right. So you think attention is not gamed, that's different and is more indicative of quality media?

[Mark Golderman]

Well, it depends. I mean, some attention metrics could definitely be gamed and others, you know, like, I think the gaming aspect of it comes down to if the metric is transparent or if the weights on the different sort of measures that are used in the metric are more opaque. That's definitely something that we have a strong opinion on, which is like, in order for a metric to be trusted by both sides of the market, there needs to be a little bit of opacity.

[Ari Pipero]

Why? Why does there need to be opacity?

[Mark Golderman]

Because otherwise it's gamed.

[Ari Pipero]

Right?

[Mark Golderman]

Like so take your FICO score for example, or take the Google search engine results page, for example. You know, on your FICO score, people have a general idea of how it's used to judge their creditworthiness. But the actual algorithm isn't public. And if it was public, people would behave in a different way that wasn't really authentic. And the more a metric is game the less useful it is to the marketplace.

[Ari Pipero]

Right. I remember the human people used to always say that about their bot tracking, which makes a little more sense because bot is obviously an adversarial relationship. Are you applying that sort of adversarial with media companies that they're going to try to make things more attention grabbing if they understood how it was being scored?

[Mark Golderman]

Well, yeah, I think in general most markets are adversarial and once the seller understands the way that they're being measured in a market, they're going to try to deliver the cheapest product, which clears that hurdle. It's just like it's an inherent side effect of a free market.

[Ari Pipero]

Right.

[Mark Golderman]

Which is why after viewability came into vogue, all of the ads became viewable and they all shipped it to the top of the page.

[Ari Pipero]

But that was good. That wasn't cheating. That was the intention.

[Mark Golderman]

I don't think the intention was to litter the top of every web page with 50 different placements and 25 autoplaying videos or create this sort of the substrate for MFA sites to become popular.

[Ari Pipero]

Well, yeah, I think that a lot of it was the buyer taking viewability as a be all and end all when it was just a attribute.

[Mark Golderman]

Yeah, I mean it should have been like a sanitation metric. I don't know if I'm using the right word. It should have been just like sort of.

[Ari Pipero]

Yeah, cleanliness metric. So someone once told me there are 24 major attention vendors currently in the market. Is that accurate?

[Mark Golderman]

That sounds a little low. You might need to update that number. There's a few launching. It seems like every week. It's a crowded space.

[Ari Pipero]

Is that good or bad?

[Mark Golderman]

Oh, I think competition is great. I think that in some ways competition is awesome. There's no question that having other people in the market have forced us to make our product better. The one downside is that I think the biggest threat to the widespread adoption of attention metrics today is bad attention metrics. We have a harder time selling into people who have tested an attention metrics vendor and had it not work than people who are sort of greenfield and are currently using viewability without turning this.

[Ari Pipero]

Into a commercial for your company. Can you give us a overview of the different methodologies for attention that are out there in the marketplace?

[Mark Golderman]

Yeah, sure. And I will try to make this as non commercial as possible, please. Thank you. We view attention in four waves. The first wave is viewability and video completion rate. A lot of people don't think of those as attention metrics, but since the job of Media is to deliver attention to creative. And those are proxies for the quality of media. They are de facto attention metrics. That was the first wave of attention metrics. The second wave was using like Viewability plus. So instead of stopping the counter at one second or two seconds, you just keep it running and you figure out how long was the ad on page for. It's pretty predictive, especially if you have what we call like a politely interruptive format where it captures the user's full attention from out of time they control. But obviously it doesn't take into account all the context. Like you can have a little tiny crappy banner ad that's on screen for a long time and it's not necessarily a good placement or a good impression. Third wave is the idea of using viewability metrics and eye tracking to predict gaze duration. This was definitely an advancement over Viewability Plus. We actually ran an ad network where we were selling on predicted gaze duration called parsec. It works really, really well. The big downfall of this is that the dependent variable in the gaze duration approach is gaze duration and not business outcomes. The other big problem is when you optimize towards gaze duration, you have some unintended effects on audience. There's creative incentives that are a little bit weird. Like you don't want the creative that captures the most attention. But there's not a lot of like, unless we get generative AI into the DSPs. Like optimizing the creative to get the uncanny Valley and weirding people out and that attracting attention.

[Ari Pipero]

Coincidentally, gaze duration is my drag queen name.

[Mark Golderman]

I don't have one of those. Ari, maybe was it like the street I grew up in and the name of my first startup, Roosevelt Sponge Cell? I guess.

[Eric Franci]

That's not bad.

[Ari Pipero]

That's not bad. All right, what's the fourth?

[Mark Golderman]

Here's the thing on the audience side. If you optimize your campaigns to the longest duration, you get old, drunk people. Old people pay more attention than young people. Sober people pay less attention than intoxicated people.

[Ari Pipero]

Wait, wait, are intoxicated people just staring at banner ads? Is that what the kids do nowadays?

[Mark Golderman]

They're just moving their screens really slowly? Right? Like, people that are intoxicated take 33% longer to read than people who are sober. And again, and the more intoxicated you get, the longer you take to read things.

[Ari Pipero]

Is this all drugs or just alcohol? Like crystal meth probably has a very low impact on attention.

[Mark Golderman]

I was painting with a broad brush.

[Eric Franci]

It's alcohol, but Hemingway was it. Read sober, write drunk no, no, it.

[Ari Pipero]

Was write drunk, edit sober.

[Mark Golderman]

Yeah, I mean, maybe, you know, like, weed is legal everywhere now, so maybe that's also.

[Ari Pipero]

So old drunk people are responsible for the attention economy.

[Mark Golderman]

No, no, no, no. Old drunk people are what you get when you optimize towards the maximum amount of attention, which is what you don't want. Like, you don't. You don't want the maximum attention. Like, you get that, you know, the Clockwork Orange joke, which I think you made at me on Twitter like five or six years ago. @ attention metrics, you don't want the maximum of attention. What you want is the highest probability of attention on your creative. And that's what the fourth wave of attention metrics are. Instead of saying we're going to maximize the amount of attention, we maximize the probability of attention. There's a great researcher named Duane Varen, and he says the biggest risk an advertiser has is no attention, and that's what they need to mitigate. And so that's where I think that the fourth wave of attention metric solves.

[Ari Pipero]

How correlated are these things to viewability? Like, I would think at some level they're correlated. Like if you have no viewability, if you don't know attention. That's right. If you have high viewability, you probably have high attention. So why is it so much better if the two metrics kind of march side by side?

[Mark Golderman]

For one thing, there are sub viewable impressions that get enough attention or to work. Anybody that buys on Instagram or measures Instagram knows this and the meta folks will talk about this all day.

[Ari Pipero]

So can you explain that? What do you mean sub viewable?

[Mark Golderman]

You know, 75% on screen for 0.75.

[Ari Pipero]

Seconds and they get attention.

[Mark Golderman]

You know, people can notice a brand. I wish I had some data in front of me on this, but you can definitely notice a brand in under a second.

[Ari Pipero]

So is that a flaw? It's like the opposite with viewers.

[Eric Franci]

It's like the opposite of a drunk effect.

[Mark Golderman]

Right.

[Eric Franci]

Where if you're locked in and you know, just like scrolling through Instagram or a platform like that, like, you're probably extremely attentive.

[Mark Golderman]

The meth effect, right?

[Ari Pipero]

Yeah. Except everyone under 25 has it without drugs.

[Mark Golderman]

So that's on the under viewability side. On the overviewability side, viewability treats all impressions that cross that threshold as the same. Right. And there's definitely a divergence in quality. Once you cross that viewability threshold. There's ads that are massive and take your full attention that might look less appealing than ads that are tiny and capture a very little bit of your attention if you're using viewability and especially if you're using video completion rate. Right.

[Ari Pipero]

You gave us a very good explanation of sort of the sort of waves of attention. I like that. But on the technical level, what are the methodologies that are out there? Because my understanding is like you have some attention companies that have opt in panels, like paid panels where they literally track people's eyes and pay them and then you have other folks who maybe are just using some JavaScript and using some inference about what the mouse cursor does and stuff like that. So can you lay out what the methodologies out there in the wild are?

[Mark Golderman]

Yeah. So I think everybody should use eye tracking data. Like, you need to understand human attention. You need to understand how humans engage with media in order to predict attention.

[Ari Pipero]

How much would you pay me to track my eyes? Like, what's the bounty?

[Mark Golderman]

Oh, let's see. I think it depends on your demographics, socioeconomic.

[Ari Pipero]

I'm an old drunk person. How about that?

[Mark Golderman]

Nothing $10 a month. We buy a lot of eye tracking data. We don't do our own eye track. And in fact, I think that's an advantage because like we get a lot more diverse eye tracking data.

[Ari Pipero]

Wait, are there marketplaces where I could buy eye tracking data?

[Mark Golderman]

You know, there used to be. I think there was a company called Tobi that was trying to set up something called Atex or Attention Exchange. I don't know how well that went, but we like it. Ironically, the people who are now in the attention space were eye tracking companies. Like about six years ago, we were licensing a lot of their data and they kept asking us what we were using it for and we showed them and then they pivoted. So take that for what it is. But in any case, you need eye tracking data in order to understand how humans engage with media. But the difference is what's the dependent variable? In the third wave of attention metrics, the dependent variable is the duration of gaze. In the fourth wave, the dependent variable is actual business outcomes. And I'll explain. The best analogy is credit ratings. If you want to rate the quality of something, there's a lot to be learned from either like FICO or Standard and Poor's or Moody's. If you're going to build a new credit rating that you'd do three things. You'd first work with researchers, understand what are the attributes of people that make them more or less likely to pay back debt. Then you would go out in the market and you would Gather all of that data about people and build a model. And then you would get historical debt repayment data and you would train your model so that it's informed by research, but predictive of actual historical data of people repaying debt. So the best approach to attention metrics is doing the same thing. You work with attention researchers to figure out what are the characteristics of placements that make them more or less likely to pay back debt, not to pay back debt, to drive attention. And then you go gather that data about every placement in the world. You build a model, but then you get outcome data and you train your models so that the metric is based in attention and informed by attention, but predictive of outcomes. Because the job of advertising is not to drive attention, it's to drive actual business results.

[Ari Pipero]

Okay, so that kind of tees up my next question, which is, what proof do we have that attention drives outcomes?

[Mark Golderman]

There's plenty of proof out there. There was an article, an unfortunate article, an adage a few weeks ago, maybe a month ago at this point that said attention doesn't move sales or other brand outcomes.

[Ari Pipero]

A great article.

[Mark Golderman]

It was a burner, Great headline. The headline could have just been like, some attention metrics work, other ones don't. Because what they found in the article, it literally said, Kroger did a test, it didn't work. Oh, but look, Adelaide has a ton of proof. And one of our competitors, Lumen, also has a ton of proof to show that attention drives outcomes. Every single client we work with, we make them go through a very rigorous test where we prove to them that attention drives more efficient incremental outcomes than viewability or video complete train, whatever else they're using. There's a large body of work that proves that attention drives outcomes.

[Ari Pipero]

Does it always work? Or I mean, Kroger, it didn't work for Kroger. Why do you think it may work for one company, not another?

[Mark Golderman]

Well, depends on the vendor you use, right? They were using Doubleverify. And I mean, I don't know DoubleVerify's products any better than anyone else, but I don't know how much eye tracking data they use. And like, you know, it doesn't surprise me that the product that comes from the makers of brand safety doesn't work very well.

[Ari Pipero]

Burn. So that also tees up the next question, which is what is the current state of the big guys offering this? Because obviously your company's small. You mentioned Lumen, which is also a pretty small venture funded company. IaaS, DV, Nielsen, comScore, who offers an Attention metric. And who doesn't?

[Mark Golderman]

Yeah, so I mean the irony is like Moat, like Jonah at Moat, sort of made this whole category over a decade ago. So I think they would definitely be in the lead if they were still around. The DV folks made their own product called Authentic Attention. They licensed data from TVision for their CTV product, but as far as I know they don't have any eye tracking data in their sort of desktop mobile product. That product is MRC accredited though, just for desktop and mobile.

[Ari Pipero]

And the MRC has an attention standard?

[Mark Golderman]

No, they just accredited the Double Verify Authentic Attention metric as accredited. Outside of their they're working on a standard with the IAB at Adelaide. We've had our product in the under review for accreditation, which is what we just announced. And that's for Omnichannel. That's for everything from web to digital, out of home to CTV and cinema. IES has a product as well that uses Lumen data. So they've licensed Lumen eye tracking data to use in their web and mobile product.

[Ari Pipero]

Did you say cinema?

[Mark Golderman]

Yeah, cinema. It gets a lot of attention.

[Ari Pipero]

Yeah, sure. How do you measure it?

[Mark Golderman]

So there's actually companies that have set up eye tracking inside of cinemas and we've licensed that data. And actually the two biggest cinema companies, Screenvision and NCM in the States, both ran eye tracking studies. They shipped us the data, we processed the data and then we ran through our normal sort of three step process of using the research, gathering the placement data, then getting outcome data and training the model.

[Ari Pipero]

Is it 100% attention or people checking their phone?

[Mark Golderman]

It's pretty high up there and it depends on if it lights on or lights off. There's all sorts of factors that play into it.

[Ari Pipero]

People forget cinema, but it's literally the largest market in digital out of home or it was for a while. Okay, so let's talk about the future. So what do you think is going to happen with Attention metrics? Are they going to become a currency? Yes. No. And also just what do you expect to happen in the marketplace?

[Mark Golderman]

The nice thing about markets is that opinions don't really matter about currency. The market's going to decide if attention becomes a currency or not. And there's two ways that currencies are made. You can make a currency via Fiat, which is a bunch of people in a room that decide that 50% for one second is good enough to trade on. Or there's arbitrage driven currencies. This is not the typical arbitrage you have in adtech. This is a Normal free market arbitrage where the buy side has asymmetrical information. They use that to make more educated bids. And over time, as the information used for arbitrage becomes more diffuse, the alpha that's generated by that information gets compressed. Right. So the price of undervalued assets starts to approach its actual value. And at that point the buyers with the data, at some point they're like, I don't want to do all this arb stuff. I'm not an investment bank. I just want to go and buy media and have predictable, consistent outcomes. They show up with the new metric to a publisher and they say, hey, will you guarantee me this buy using this metric instead of using viewability? And then at that point it's a market driven currency. And we've already started to see that happen. So the Wall Street Journal announced recently that they're making guarantees based on our data. There are like A and E, for a while was making guarantees on TV data. So it's definitely starting to happen and it's really, really interesting because it's market driven and not driven by a bunch of old white dudes in a room.

[Ari Pipero]

Driven by a bunch of old white dudes. They're just in different rooms. So when you say guarantees, so that's like basically there's a currency like impressions on an IO with the Wall Street Journal, let's say. But they're also putting a provision that is going to be at least X percent attention based on your metrics.

[Mark Golderman]

Yeah, well I, well, I wouldn't say we don't deal in percentages, we deal in ratings. We rate media placements. And in order for a guarantee to be met, the ads need to be served into a placement which reaches that minimum rating. In our industry, anything that's in a contract is considered a currency. And so it's sort of like it's not really using the word technically correctly, but I think that if a guarantee is in a contract, then it could be considered a currency.

[Ari Pipero]

Yeah, I think so. I think often in the TV business they talk about secondary currencies, Nielsen for ratings and then you have t vision for something else or something or video amp for something else. So that's kind of that characteristic. Awesome. Well, thank you so much for this great overview. I learned a lot about attention and about the whole market. I guess we should have disclosed that. Eric, you're an investor, right?

[Eric Franci]

Yeah, my fund invested in Adelaide this year.

[Ari Pipero]

Cool. I'm an investor through you, but I'm also an investor in Lumen through First Party Capital. I'M maxing up my attention portfolio. All right, we will be back with news of the week with a lot of earnings, some controversy around Zeta and Vayant buys Iris tv. So stand by. This podcast is brought to you by Adroll. Adroll is a name a lot of advertisers are probably familiar with. A leader in high performance Digital Advertising for 15 years, AdRoll is a brand you can trust to deliver strong ROI across display, native and paid social campaigns in one platform. Get in touch with an account expert or create a self Service [e-mail] Ari Today. That's adroll.com Ari Today.

[Eric Franci]

All right, and we're back. So on the docket today we have earnings and some public market drama, some M and A, a Forrester survey that you're mad about, ads coming to perplexity and a bitcoin prediction from Mark G. So we have something for everybody today. Quick just kind of recap on public market stuff. So last week Ari, you were in uk, we did news of the week. We focused a bit of time on Applovid who is now the most valuable pure play ad tech company out there. TTD and Magnite subsequently both reported didn't have blowout earnings and stock going crazy like app level did. But I think there were a couple of interesting things on TTD and the Magnite front, both of which are around CTV. So Magnite CTV revenue up 23% year over year. Barrett actually had an interesting comment. You know, somewhat cagey details but expects Netflix to be one of their largest customers by 2025. Elsewhere Netflix reported that its ad tier is now up to 70 million users. The other thing interesting on the Magnite front is that curation is up 100% year over year. So tracking with everything we've been talking about TTV overall revenue up 27% year over year. They don't break out CTV like Magnite does, but Jeff said it's the largest channel now. Video, which encompasses CTV +OLV is almost 50% of total revenue. So next year TTV will be a video company.

[Ari Pipero]

It's exciting stuff. I like the Netflix point. That's really interesting. I think. Here's my perspective of Trade Desk vs Netflix AppLovin because they're pretty different approaches at different companies. I think historically the most revenue has accrued to companies that create marketplaces where you are on both sides buy and sell. And when you're only on one side you're always at the risk of being disintermediated from your supply or your demand depending on which side you are. And so in a sense it's kind of remarkable that the trade desk has done so well only being on one side. And if you look at their strategy of Openpath and UID2, they're effectively trying to cobble together the second side. It's like a federated approach instead of owning it. Whereas Applovin is on both sides and that's their secret sauce. And Eric Suffert's excellent newsletter covered this this week. So it's an interesting comparison. They're not competitors with each other, they're just interesting to look at side by side as two different approaches to the same problem.

[Mark Golderman]

I think it's interesting that I read a interview with the Applovin CEO where he talked about them taking the risk out of it for advertisers. So they're guaranteeing outcomes sort of like the principle based version of dsp, I guess, I don't know. But their value proposition is guaranteeing installs if I'm right, or doing some sort of lower funnel activity guarantee.

[Eric Franci]

That's been their business model, just driving downloads for mobile app developers across mobile apps. More recently they've started to open up that inventory to other performance marketers and they're selling it just in that same way, like a PMAX or Earn Advantage. Plus, my sense is that's not where a lot of the revenue growth has been today, but it could lead to a monster business in the future. That's actually back to you, Ari. That's an interesting point. You said this in a different way yesterday at the Bid Switch event. Invite only. You had a fireside with my partner Josie, who was really good. Basically the bull case for ad nets. It's because they're on both sides, right? And you can build a bigger business as a result. Am I right in thinking you're kind of connecting those two?

[Ari Pipero]

Yeah, I think the beauty of the Ad net is that it gives the customers exactly what they want and if they don't get what they want, they could go somewhere else. As opposed to a software product where you're potentially locked in for years and you have training. And Applovin's executing a really good ad network model. And people there are a lot of critics, you know, people chiming in on Twitter and whatnot. But I always like to remind people that the customers are choosing to work with them. So let's just assume the customers are smart instead of assuming the customers are stupid.

[Mark Golderman]

Well, the customers are following their short term incentives, which is to drive the cheapest possible business Outcomes, which is great for Applovin in the long run because they learn how to do that more efficiently than the customers do. And so the customers are sort of addicted to that.

[Ari Pipero]

Yeah. It's also relevant the way they position themselves. They don't say we're the best roas you're going to get in the universe. They say there's only so much you can get out of search and social. They're acknowledging that their roas is probably not as good as meta, which it probably isn't, but it's above the line where the cost of capital makes it worth investing to get new customers. The LTV CAC calculation still makes sense. It's a very solid business. One of the dark horses here is ctv, where they acquired Rural, which is a really interesting company and I think you'll see a lot more from them on the CTV front.

[Eric Franci]

Yeah, makes total sense. They're the one to watch.

[Mark Golderman]

All right.

[Eric Franci]

In some not so good public company news, Zeta Global, which we've been talking about just given their recent acquisition of Live Intent. And the business being valued quite largely is no longer in the public markets. Valued what it once was, the stock was down 37% yesterday because of some allegations over consent farming. Before we get into it, consent farming, new term to you guys or you have heard it before?

[Ari Pipero]

It's legion. It's legion, right? Legion's always been a pretty dirty part of the ad tech business.

[Eric Franci]

Yeah, it's the dirty part of lead gen. So consent farming is luring individuals to provide data under the guise of providing benefits and rewards. The data is then sold and the consumer often doesn't receive any of the promised benefits. Like give us all your data and you win an iPhone and the iPhone doesn't show up.

[Ari Pipero]

This is like some. Sometimes it's a search arbitrage. So you search for doctors in Detroit and they're like, we'll find you the best doctors. You give them the info and then they sell the information to a bunch of doctors. And it's legit. It's just sort of a search arbitrage because you'd be better off on the search having found the doctor site the first time instead of the intermediary who collects the data.

[Eric Franci]

All right, so how does this go back to Zeta Zeta? And again, these are allegations from a short seller. Short sellers benefit when stocks go down. Zeta, allegedly, according to this short seller, uses two way deals with consent farms to inflate revenue, raising round tripping concerns. And Zeta's own consent forms apparently have some O and o. Reportedly collect consumer data under false pretenses. And again, allegedly, according to a short seller, drives most of its growth and 50% of adjusted EBITDA, which risks FTC action. I don't know anything about this stuff. I had no idea this was such a big part of Zeta's business. But this is wild.

[Ari Pipero]

I think Zeta. I have absolutely no basis for saying this short seller is right. If there's anything, I have no information, but I will say that Zeta hasn't done itself favors by having a very confusing discussion about what their actual business is. I think we've talked about it previously on this pod. We're all professional advertising tech people and we have no idea how Zeta makes money. It's one of the biggest public companies out there and you read their documents, it sort of sounds like they do email marketing, but do they. They have all this data, all these consumers profiles, sort of like, well, how does that actually work? I've never had any idea. So yeah, I guess it's kind of an easy target. If you find, maybe you find the worst websites, affiliate with Zeta and you show them and there's some other businesses that are totally legit. I have no idea. I have no idea. Yeah, the other thing in the allegation is they're alleging two way deals. So where they're paying for work, maybe they' paying some of these content farms for users and then they're also maybe driving traffic to them. So there's revenue going both ways, which is a not illegal thing to do, but it brings up a lot of accounting issues and it can be a way to manipulate earnings to some extent. So that's the other allegation from this company that we should mention. The company is called Culper Research, a short seller.

[Eric Franci]

Mark, got anything on this one?

[Mark Golderman]

Nope. No. I mean, I think, I feel like if short sellers dug into most of adtech they would find some sort of questionable things in a lot of places.

[Ari Pipero]

I will one other little point here. So a couple of weeks ago we talked about Open Web and their CEO. By the way, the CEO lost his lawsuit. So I believe he's being kicked out. Officially is out. Out. Yeah, we didn't have that in the news, but I saw it elsewhere. So the original CEO got replaced by new CEO, refused to go, brought them to Israeli court. He lost. In case none of that's relevant. What's relevant is we were skeptical about Open Web's business model because it's effectively the comments section of a bunch of websites. Well, Zeta's business model is the comment Section of a lot of websites. One of the main sources of data is Disqus, which they purchased many years ago. So there's a. There's sort of an analog there that this is kind of like very similar to Open web, let's say that.

[Eric Franci]

Yeah. Yeah, that's interesting. Okay, we'll move on. Some M and A news. Vyant acquired Iris tv. Iris TV was an investment by my fund and I believe a personal investment by you too as well, Ari. So plan is to keep Iris as an independent business, but I think also use that data to support Vyond's growing CTV business. That's your take?

[Ari Pipero]

Yeah, I think the Vanderhooks just love a nice, juicy, cheap asset. And I haven't listened yet, but ad tech God actually got the Vanderhooks on his pod today, so you should listen to that. You probably would learn a lot more than us speculating, but I think they have aspirations to use a lot of data and to be a bigger player in ctv. And their stock is skyrocketing, so. Good for them.

[Mark Golderman]

Did they disclose the price?

[Eric Franci]

No, the price wasn't disclosed and neither of us are at liberty. So back to the Vanderhooks, though. The vanderhook or the Vyent 2024 Ark. It's impressive. So they came out with the AI planning tool. They made their first acquisition as a public company and the market is rewarding it. Stock is up 136% year to date and TTD as a comparison is up 75% year to date. So these guys are doing well and they've been at this as long as anybody.

[Ari Pipero]

Yeah, I think the DSP business is unfairly maligned because obviously you've had flameouts like Media Math and others. But really, if you execute a DSP business well, it could be quite profitable. I mean, you have private companies like Basis and Simplify that are really successful companies. And Viant's public, so they have to do it in public. But there's no fundamental reason why the DSP business is a bad business.

[Eric Franci]

Yep, totally agree. All right, you're mad, Ari, about the Forrester SSP survey.

[Ari Pipero]

Okay, let me just. So I didn't read the whole thing. It was just the typical Forrester chart with the leaders. And I don't have it in front of me even, because I think they asked everyone to take it down. This is a joke. This is just an embarrassing joke. Let's just go through the jokes. First of all, it's an SSSP survey. No one uses the word SSP anymore. So they're already on their back foot by using a term that's totally outdated. Secondly, who is this for exactly? If I'm a buyer, a DSP or a buyer, I don't choose SSPs. I just buy whatever inventory I want. If I'm a seller, I don't choose S and P's. I use all of them. Have you ever heard of header bidding? There is no such thing as an SSP buyer. It's not a buying product. You don't buy it. There's no CIO somewhere evaluating SSPs to go. I think Magnite's a leader. No, that makes no sense. Next, there is no one SSP market. There's a video market and a banner market. They're totally different. Don't put them on the same chart. Last, the number one player by far the leader in the industry is Google. We have to acknowledge that access more than 50% market share, they are considered a laggard on this chart. It's a joke.

[Eric Franci]

Just more specifically, more specifically and maybe even funnier, Google is being termed a challenger.

[Ari Pipero]

A challenger? A challenger. They're a fucking monopoly. The actual government is going to declare that product a monopoly. And someone at Forrester is like, sounds like I've heard of this Google company. They seem really innovative. Maybe they'll challenge some. For God's sakes, just stop publishing this crap. It's not even worth criticizing. It's just an embarrassment to everyone involved. It's like the sort of thing that you're embarrassed by when you see it, like just as a reader, it's embarrassing. Forget the publisher.

[Mark Golderman]

Everybody knows this is pay for play, right?

[Ari Pipero]

Everyone says that. Is it? Is it not? I don't know, like, it's what the Pay for play. You don't get better rankings by paying. What you do is you manipulate the questions. So the way this thing works is the analyst comes up with a spreadsheet that's got like 10 tabs in it. And each tab has questions. And those questions are highly opinionated. And the analyst has a say, of course. But if you're a paying customer of the analyst, you might have a meeting and say, you know, what's really important is audio support. I think audio support is the future. And so you end up with a new tab that's like audio support. And if you might be a vendor that doesn't do audio support, suddenly you're really on your back foot. The other thing is qualifications. There's probably 60 SSPs out there in the world where do they draw the line about who's on the chart, who's not? They'll come up with some arbitrary number. They'll say you have to have a certain amount of gross billing, plus you need to have at least X number of named Fortune 500 customers who you put in case studies. That number will automatically disqualify some real challengers. I'm speaking a little bit out of personal because Beeswax has never listened to any of the DSP ones. Even though we're the fifth largest DSP in the country. We couldn't get on any of the Forrester reports because the qualifications were wrong. And actually they changed the qualifications. After we had filled out the survey and spent hours and hours on it, we got this little email like, hey, we changed the qualifications. You don't have enough Fortune 500 customers, you're out. So it's not exactly pay to play. It's more like influence. Pay to influence.

[Eric Franci]

My question is, Ari, do you feel better?

[Ari Pipero]

I do feel a little better.

[Eric Franci]

Felt like you had to get that off your chest.

[Ari Pipero]

Thank you.

[Eric Franci]

You're welcome. Tune in next week for weekly therapy with Eric Fresche. Okay, well, let's keep going. We have an update on. Everything is an adnet trademark Eric Sufert. So ads. We've known this from some rumors and other kind of scuttlebutt that ads are coming to Perplexity. They put out an interesting post on. They're on their own site discussing this. I thought it was interesting from two standpoints. So number one, they're not just slapping some standard IB units on the page and calling it a day. They have created a very native basically sponsored follow up questions. So you ask Perplexity a question, you then get a list of follow up questions to dig in further on the topic. There'll be a sponsored one that's clearly labeled. It looks really good, quite frankly. Question is, you know, about, you know, how engaging it'll be. They've got some flagship brands already signed up, indeed, Whole Foods, a couple of agencies. But the other thing that was really good was the post itself. And you know, there's this always consternation in industry about, you know, users don't understand the value exchange and so on and so forth. And I thought they did a really good job of explaining like why they're doing this, why it's important, how it's going to fund the core product. So we'll put a link to it. But I thought it was well done. Did you guys take a look?

[Ari Pipero]

Yeah, I think it is interesting in that if you think about search as directing people to a website and that's not the goal anymore, the goal is an answer. You can ask a question like, how should I plan my vacation in Cancun? And then it's pretty obvious. You could have some really great opportunities to be like, is the Sandals resort all inclusive or not? That sort of thing. Which could be actually quite a bit more valuable than just clicking out to some web page.

[Eric Franci]

Yeah, no, that makes a lot of sense. I would assume these are high attention units, Mark.

[Mark Golderman]

We haven't studied them yet, but I would agree that they're probably pretty high attention, pretty engaged people.

[Eric Franci]

Yeah, we should keep an eye on this stuff because this ends up being probably the format of the future as LLMs start to become more of a consumer product. All right, Mark, we're going to clear the deck for you. You have the hottest take of the week and it's related to bitcoin.

[Mark Golderman]

So Bitcoin's up by 50% in the past month, which is great. Yes sir, I'm long myself. But the downside is it means the inevitable return of the ad tech blockchain scammers who want to take every part of the ad tech ecosystem and decentralize it. Like weird ideas like putting your ad serving logs on a blockchain or using the slowest database ever created to store all of your personal data. And so I like why we went through this. Like I think it was five years ago when Ethereum first came onto the scene and was popular. And so, you know, I think inevitably we're going to start to see people, maybe it's going to be like Gen AI had distributed like DCO stupid ideas. There's like one use case for blockchain and that's like allowing for trustless transactions of digital assets. And the irony is like one digital asset is money, but the only other, like real, truly only ephemeral digital asset is digital advertising. And so there is a use case for blockchain and digital advertising, but it's likely just like, you know, the identity like the digital asset being the opportunity to serve an ad to somebody. And so if you could tokenize that, like maybe that's an application of blockchain and digital advertising. But I don't think all this other stuff is just is such a waste.

[Ari Pipero]

As far as I know, I can only name three blockchain adjacent companies that are in advertising. By far the winner is the Brave browser, which has nothing to do with blockchain. They just raised all their money on A fake token. So now they're just a browser that has hundreds of millions of dollars in the bank. Two is this company called Permission IO that I really don't know anything about except they DM me on Twitter all the time. And three is Antonio Garcia Attribution for Spindle? Yeah. Do you know any others, Eric?

[Eric Franci]

No. There's been a lot of companies that have tried and ultimately ended up pivoting and a lot of companies that tried and failed. And I think that, yeah, Mark, you raise a couple of good points. I think the applications are going to be the next or the sort of like ultimate next value creator in blockchain. The real question is around its applicability to digital advertising. I think what Antonio's business is doing is interesting. It's attribution and using wallet data to measure it.

[Mark Golderman]

Right.

[Eric Franci]

So if everybody has a crypto wallet, you can then do some interesting attribution based on on chain activity. But the on chain activity has to happen.

[Ari Pipero]

Yeah. It only really works for marketers who are also on chain. Right. Where the conversion is on chain. All right, well, let's call it there. So I think there's a great conversation, a lot of news this week. So, Eric, thank you for news of the week. And Mark, thank you so much for telling us and teaching us about attention.

[Mark Golderman]

Thanks for having me.

[Eric Franci]

All right, everybody, we'll see you next week. Bye bye. Thank you for subscribing to Markitecture. New interviews are added every week at.

[Mark Golderman]

Markitecture TV in your favorite podcasting app.

[Ari Pipero]

Thank you for listening to the Markitecture 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 Markitecture tv. You can also sign up for free for our weekly newsletter with my original strategic insights on the weekend week's news at News Market tv. And if you're feeling social, we operate a vibrant Slack community that you can apply to [e-mail].