← AdExchanger
AdExchanger
Why Media Quality Should Be The Center Of Attention
00:47:49
Online advertising’s privacy problem isn’t just about bad actors – it’s about bad metrics, says Marc Guldiman, CEO and founder of attention startup Adelaide. “I think a lot of the invasive behaviors in the ad tech space can be traced...
Loading summary
Transcript85 lines
- [00:00]
A
Foreign welcome to Ad Exchanger Talks, the podcast devoted to examining the issues and trends in advertising and marketing technology that matter most to you. Foreign I'm Allison Schiff and you are listening to Ad Exchanger Talks. Thanks for giving us some of your precious time and attention. And speaking of attention, that's what we'll be talking about on this episode with my guest, Mark Gouldeman, CEO and founder of Adelaide, a startup that helps advertisers measure and evaluate media based on attention rather than proxy metrics like viewability or video completion rate. We'll talk about why it's critical that marketers understand what quality media really is, why attention is necessary for advertising to work, but it's not a good currency on its own, and why attention is a more appropriate proxy for quality than viewability because it creates better incentives for advertisers and publishers. But before we kick things off, you've still got a minute to get this hot ticket. And I speak of screen shift. The lines are blurring between streaming linear CTV and digital video, and the future of consumer engagement is being written right now. That's why leaders from brands, agencies, platforms and tech are heading to ScreenShift Connect on October 14th. ScreenShift Connect is where content meets commerce, where platforms evolve, and where audience attention drives strategy. Podcast listeners get 20% off their pass when they use the code POD20. That's podcast 20. See you there. Hey, Mark, welcome to the podcast.
- [01:58]
B
Oh, thanks for having me, Allison. I'm psyched to be here.
- [02:01]
A
All right, so lay it on me. What's one thing about you that not a lot of other people already know?
- [02:07]
B
So one thing about me, I guess that a lot of work people don't know is I used to have dreadlocks and follow fish around the country. I think I've seen like 100 or 120 shows at this point.
- [02:19]
A
That is an interesting fun fact because my previous guest, who is an economist, we were talking about the Google Attic antitrust trial on that episode, and his fun fact is that he went to 85 fish shows. He didn't mention dreadlocks. I'm trying to picture you with dreadlocks right now and I'm having a little trouble.
- [02:40]
B
I think I had a little bit of a different hairline and a little bit more volume back then. I think there's like a low key ad tech fish head scene. There's a bunch of folks that go to the MSG shows every year, I guess.
- [02:54]
A
So. I actually have a closet full of fish T shirts because my boyfriend used to follow fish around the country a few years ago and I borrow them all the time. They're great.
- [03:07]
B
Do you get some weird comments on the street from people?
- [03:10]
A
Not weird comments, but. But it's not that I'm not a fan, I'm just not as. I'm not familiar really. But I love the T shirts. And so I have had people come up to me in coffee shops and start talking fish at me and I can't really participate because I don't know anything. And so then I have to say, oh, I'm just borrowing this T shirt. And they lose interest pretty quickly.
- [03:32]
B
So if someone says like, oh, you missed. Do you remember that sick bathtub gin from Hartford 94? And you say like, what are you talking about?
- [03:40]
A
I was 11, but yes. Well, I'm going to take us back to not 1994, but to August 20, 2012, because I was curious, what was the first time that you were quoted or mentioned in the pages of Ad Exchanger? Because we've been writing about different companies you've been involved with for a long time. So I did a search and I found Q and a from 2012 between you and Zach Rogers, who was our executive editor. And it was about your startup called Enlighten, which basically let people manage their own data online. They could pick what personal data they wanted to share with advertisers and then they'd get something in return, like points or digital content, or they could support charities or stuff like that. And it got acquired at some point down the line and then you went and you did something else. And I know this is a, this is an old idea and you've moved on, but I do love the idea of putting people in charge of their own data. And it sounds so good on paper, but getting real world traction is really, really hard because, yeah, like, people want perks, but they're lazy and they're busy. And I speak for myself. So how much do you still like that idea? It does feel more relevant than ever, but I think people are also still busy and also are still lazy, I think.
- [05:09]
B
So you must get this pitch every six months, right? Like, I have this new startup that's going to help people take back control of their data and stick it to the man, you know, like, so I get an email once a year from people saying like, oh, I'm starting a similar startup. Give me advice. And my advice is typically like, don't do it. The incentives that are created around paying people for their data are really weird. And like, when you, when you finally, like get down to the deployment of the thing. You'll, you'll start to realize that you end up just creating incentives for the wrong kind of data and the wrong kind of audience. When I was at enlightenment, I didn't really understand how advertising worked and I wanted to build, we wanted to build something around audiences rather than around insights or like, you know, brand lift studies or kind of things that you might get from a panel instead of a, a big sort of pool of audience data. But we, so we, we tried a couple of different approaches. The first one was donate your data to a nonprofit. And that didn't work because the nonprofits typically get the vast majority of their donations from a few people. And if one of those two or three people got, you know, got the idea that the nonprofit was spying on people, then they might pull their donation and the whole thing is, is negative. The second idea was use your data to get around a paywall, which sounds like a good idea, like get a couple extra articles in exchange for giving this publisher a bunch of your shopping and search and social data. That didn't work because when we got to the paywall mechanics, people, they're like, you're absolutely not going to screw up my finely tuned economic model that gets in just the right number of subscribers. And then the second to last idea was get free shipping from an E commerce provider if you give, from an E commerce shop, if you give them your data. So give your data to Walmart and get free shipping for your Amazon data type of thing. That didn't work because the lawyers got involved in the very end and they said, oh, this is a little bit too sketchy for the upside. The last thing that we tried and we never, like, we were sort of at our fourth pivot and my co founder and I were at each other's necks and we never really, never really got out the door was something called Buck Hunter or Hodomatic, which was like a value exchange for your data. So you would install this plugin that would take your search data, your shopping data, your content consumption data, send it up to us, and we would then also index all of the web that was on sale. So like all, you know, everything, all the sale pages from all the, all the online shops, and we would, we would create a webpage for you that was specially curated based on your interest, everything that was on sale for you. So people actually liked that. But we never, we never got it launched. So I ended up shutting that down and taking the last, like remaining 50, 50,000 bucks and putting it into the next company which has actually since become Adelaide. So it's been sort of like this 10 year journey through a couple of different pivots here as well. The funny thing though is about a year and a half ago we acquired a company called Reta. And Reta did this idea, but with a different approach. They used GDPR part 15, which is the right to access your own data, hence the name reta. And it allows people to join a panel and give their data from major platforms as part of that panel without a lot of work. So you basically you join the Reta panel, you connect a couple of different platforms data and that makes, that creates the data for the panel. It's a lot more lightweight and the data is a lot denser. So we're actually using the data from the RETA panel to build, to sort of inform the algorithms at Adelaide. So yeah, it's come full circle. It was over a decade since I tried that idea and we acquired reta. I do think, I think their approach is a lot, a lot more efficient than the ones that we, that we tried previously.
- [09:10]
A
But some version of that idea is clearly your destiny.
- [09:15]
B
I'm, you know, I'm a big, I'm a big privacy advocate. I think that a lot of the, the, the behavior, I think a lot of the invasive behaviors that we do as in the ad tech space can be traced back to a lack of shared understanding of quality. So at Adelaide, we're trying to fix the privacy problem from a different angle. If you think about it this way, if you went to a gas station and they didn't have gallons and octane, like they didn't have a shared understanding of quality or quantity, you would very quickly build an attribution system for your car to try to figure out what did that bucket of gas or whatever it was that you bought did towards your business goal of transportation, right? Like where your windows up or your windows down? Do you have new spark plugs? What is your tire pressure? Was the humidity, all of the various factors. And you try to build this model to understand what was the value of this thing that you bought that was totally unmeasurable. We do the same thing in advertising, right? We just, we don't have a good measure of the quality or quantity of advertising we bought. So we build this big attribution system, except that instead of being a weirdo that measures all the different metrics in your car, we, we just wholesale invade the privacy of 400 million people. So I think that it's sort of the, it's human nature when you don't understand quality and quantity to get fixated on outcomes. And so I think that at Adelaide, one of the. We're solving the core problem that creates the incentive or one of the core incentives to track people.
- [10:45]
A
So you have made that argument in a bunch of places, including in the pages of Ad Exchanger actually, that transacting on business outcomes is a bad idea. It sounds great, but it does create big risks because brand brands lose control over their marketing and their product quality. And then media companies are just chasing like cheap low value results. Like that's I think pretty core to your argument. And you also need like tracking for that model, which isn't great for privacy and doesn't really account for long term brand impact, all of that. But, but I guess just to play devil's advocate, like what would you say to everyone out there that's been just full throatedly advocating for outcomes as our savior and like the cure to all that ails us in the ad industry, from waste to inefficiency to vanity metrics that don't really say anything like outcomes have been positioned as the answer.
- [11:42]
B
Yeah, well, so let's first define what we're talking about when we say outcomes. There advertising is about driving business outcomes. For advertisers. There's no question if there would be no reason to advertise if it wasn't going to actually move the business forward and create incremental outcomes. But I think the problem that the new sort of outcomes era people face is that they're not talking about optimizing towards outcomes or measuring outcomes. They're talking about transacting on outcomes. And I wrote an article 10 years ago in Ad Exchanger talking about why trading on outcomes is a bad idea. But it's also human nature, right? So like go back to the gas station example. If you didn't have gallons and octane, you didn't want to build this big attribution system and sort of track what the. To, to calculate the value of that bucket of gas you bought, you might say to the gas station attendant, well, let me just pay you a hundred bucks and give me enough gas to get to Vermont, right? That sounds like a fair deal to you. But you would immediately put your foot to the floor, your air conditioning on, your windows down, and like your incentives are totally misaligned, right? And they would, they would probably give you the lowest quality gas possible that would ruin your engine over the long run. So in a lot of ways, if you're, if you're an advertiser and you are Buying business outcomes from a platform, that platform is going to give you the lowest quality outcome they can. And then they're going to get really, really good at figuring out how to deliver those outcomes, because you're going to have to give them a bunch of data in the form of pixels or conversion data, and they're. They're Right, justifiably so, going to want to get more and more efficient at driving that outcome. Now, what happens when your competitor shows up and says, hey, I'm going to outbid Allison for that outcome. Right. You know, I'm willing to offer 10% more. They will 100% sell that outcome to your competitor. And then what happens is just a game of margin compression. Right? You're just. You're just going back and forth to see who will pay more for that business outcome. And, you know, this happened 10 years ago with D2C, and they became sort of addicted to the outcomes from. From the walled gardens. I think that at a surface level, you know, the idea of trading outcomes is really good, but once you sort of go one or two, you know, or one or two levels down, it becomes really obvious that the incentives that are created are not good. And, you know, as a marketer, you're, you're. The purpose of being a marketer is to turn resources into demand for a product, or one of the main purposes of being a marketer is to do that. And I think that outsourcing that, that sort of job and asking somebody else to take that risk for you, in many ways turns being a marketer into being a fulfillment arm for the demand that somebody else has created. And that's not sustainable in the long run.
- [14:28]
A
I feel like so often in this industry, people somehow forget, and it's wild, because the whole point is, as you've said, and it's so obvious to sell the thing, see the ad, watch the ad, experience the experience, whatever, but buy the thing like that is what people want to happen in the end. And there's so much talk, there are middlemen, there are so many machinations, and it does end up being like this weird spiral to the bottom.
- [14:59]
B
Yeah. If you outsource that selling of the thing to somebody else, then you're outsourcing your job to someone else, and they're eventually going to start trading that outcome with somebody else.
- [15:10]
A
So I don't know why. I'm kind of in the mood to be a devil's advocate today. So I'm going to do it. I'm going to do it again. But why should attention be treated as a currency in advertising. So something that you can buy, sell, and trade on, like impressions or clicks. Because there is the argument that it's way more complex than those other metrics. It's so tied to context and intent and creative and overall quality and in a way that just doesn't really fit into a simple transactional model, which is why CPMs and stuff like that were used for transactions. So basically, I'm going to lean back, you know, convince me otherwise.
- [15:48]
B
Okay, so attention, in its sort of clinical definition of the amount of time that you choose to focus on something is not a good currency. And in fact, my. The company that turned into Adelaide was called Parsec. And we sold media on a cost per second where advertisers would literally pay us for the amount of time people spent paying attention to their ads. It worked really, really well until we started to optimize every single impression to the maximum amount of attention. Because just to your point earlier, that had an impact on audience and an impact on creative, which wasn't necessarily positive for the advertiser. You don't want the creative that captures the most attention. That would just be puppies and kittens with no branding at all. And you don't want the audience that pays the most attention because, fun fact, it takes drunk people 33% longer to read than sober people and they remember half as much. And there's a whole bunch of other audience incentives that are created around optimizing to the maximum duration of attention. However, attention is fundamentally necessary in order for advertising to work. And if, when you think about, like, what are publishers and advertisers trading? It's. It's really like they're loaning attention back and forth. So a publisher is giving an advertiser the opportunity to capture attention with their creative, and then that creative that the advertiser has made captures attention for a certain amount of time and ideally changes the way that you think or behave. If an advertiser is buying the opportunity for attention from a publisher, the best way to rate an opportunity is probability. So you don't want to trade on the amount of attention. Rather, we believe that you want to trade on the opportunity for attention and that we've created a metric called au, which is the probability of attention by any person to any creative in a placement. And I do think that that makes for a better currency than viewability. We're not trying to replace impressions with attention. We're trying to replace viewability with a better proxy for quality. And we think that that proxy for quality should be the probability of attention. And it creates a lot of really great incentives because today the incentives are as many viewable impressions as possible, instead of the incentive being the highest probability for attention on one advertiser's ad.
- [18:15]
A
So that makes me want to talk about standards. And we will do that, but in the second half, because there's a lot to say there. But I, I wanted to go back down memory lane, which is where we started, because you mentioned Parsec and I'd forgotten that you sold part of Parsec to Cargo back in 2022 so they could build their own Attention metric product. We covered it. And then you spun Adelaide, which is your current thing, like out of Parsec. But before all of that, it was Sled mobile. I just, I was remembering all this stuff that I'd covered, like, years, years and years ago, and that, that made me think about how long we've been talking about attention and I guess how long it's taking for it to really take hold as a metric that people are convinced they need to use. You can correct me if I'm wrong, maybe this isn't fair, but I do feel like you're often in the position of, like, explaining it, defending it, saying, this is why it's great. Like, why don't people get that yet? Like, why is it taking so long?
- [19:23]
B
It's because of the incentives that have been created from 20 years of trading on CPMs. Right? So when we can speak to somebody at an agency or a media person at a brand and explain our value proposition to them, and they're like, that totally makes sense. I, I get it. We should be buy, we should be understanding quality and then buying on value instead of just based on cost. And so we say, all right, well, great, let's go. And they say, yeah, but my incentives are all around cost. I'm supposed to drive CPMs down 5% every year. I'm not held accountable to quality. And so the entire supply chain, from the CMO down through the agency, through the ad tech middleman, all the way to SSPs and publishers, is get those CPMs lower every year. So it takes a long time to unwind those incentives. And it's only been recently that we've been able to do that at scale through a partnership with mediasense. And so mediasense works with a lot of brands to set up their agency remuneration contracts, and we've been able to get those contracts. And actually, I was on stage with Phil Jackson at the SIM Conference recently talking about this Phil Jackson, the. Yeah, no, unfortunately not the Phil Jackson. He's my Phil Jackson. So he, he is my like. So Phil Jackson at Halion has actually rolled Au out across 45 markets and he's set up incentives with his agency around media quality. And that's created this sort of, this sea change in terms of how they look at media and, and for the better. Right? So they're not just looking at how do I drive the lowest cost possible, but they're looking at how do I. How do I buy the highest quality at the best value? So how do I go find bargains for my, for the brands I'm working with rather than just like, how do I drive costs down?
- [21:11]
A
And there are some publishers that have fully embraced attention metrics. I remember one of the first stories that I wrote for Ad Exchanger that was like feature style was about attention. The headline. I remember it because I was really proud of it, even though it's a little cheesy. It was like, attention, exclamation mark, there's a new kid on the measurement block, whatever. But it was about four months after I joined Ad Exchanger, so it's like well over a decade ago. And it was about attention metrics as a growing trend. But there was a big focus on the FT Financial Times, which I believe works. I think they work with you guys, if I'm not mistaken. But there were quotes in it from John Slade, who was the commercial director of Global Digital Advertising and Insight at the time. And I looked him up recently. He's the CEO now of fti. I didn't realize that. So really interesting. He's been there for almost two decades.
- [22:10]
B
But attention might be a leading indicator there, right? I mean, you bet on the right horse.
- [22:16]
A
I don't think so you could make that connection. Um, but yeah, I was rereading that story in advance of chatting with you on this podcast and there was, there's a great quote from him. He was talking about CPMs and click through rate and the usual metrics. And he said, we're not saying they're entirely useless measures for brand worth, but we are saying that there's something more appropriate. And, and yeah, I mean, it's, it's interesting to see a publisher embrace something and then keep doing it, even though maybe there was some resistance from buyers that are used to doing something else. Because publishers often, maybe not the really big name ones, but they often do have to cave to what buyers push for. And buyers are fundamentally a little lazy with their money. I hate to say it.
- [23:05]
B
Well, yeah. And it's an unfair compromise that modern that today's metrics like viability and video completion rate have forced a publisher to make. It's like you can't really respect your audience and deliver against those metrics that the advertisers want. But to your point about the ft, there was. Brendan Spain was there doing a lot of that work and Jonah Goodhart was part of that and Chartbeat was part, like, it was sort of like this first wave attention.
- [23:32]
A
Chartbeat. Yeah, yeah, of course.
- [23:35]
B
Yeah, yeah, exactly, exactly. And Jonah used to have these attention IO conferences out in Montauk. And so that was. And Jonah actually was the reason why we went from sled to parsec. So we had Sled was a big like full screen mobile ad format and my co founder Zach introduced me to Jonah and Jonah was like, hey, I'm thinking about this attention stuff at Moat. Why don't we combine the Moat data with the sled ad format and you can start trading using attention data. And so that. You're right, that was, that was a decade ago. And there was, that was a moment where most folks in the attention space were focused on duration. Parsec was selling on cost per second. The Financial Times was selling on cost per hour. And to me, that's one of the most frustrating things about the attention space is there's a lot of vendors today that still talk about attention seconds or attentive time, and they're trying to make that into a currency. And you can't hold a publisher accountable for the amount of time that someone looks at your ad because that's largely derivative of the creative. So in those 10 years, I think the attention space for the. Or the, or the ourselves and a few other others in the attention space have evolved beyond just thinking about attention as duration into thinking about the probability of attention.
- [24:54]
A
All right, well, we're going to take a quick break and when we're back, we'll nerd out on standards because there's a lot to say there. So stick with us. All right, we're back. And I wanted to ask you before we dive into attention again, in Lichen, you tried that startup with four different iterations. Like you really, I feel like, need to have a, like a real love of entrepreneurship and love technology to really keep trying and keep trying. You can't be affected by, you know, certain failures and you kept going until you found your thing. What got you interested in tech and all of that when you were a kid in Maplewood? Because I remember that you grew up in Maplewood, New Jersey. We've, we've talked about this, but back when, I don't know, you had dreadlocks. Like, what was it that brought you from there on a path that led you to startup Founder.
- [26:09]
B
Must have been in middle school that my dad brought home a book about the Internet. And it was like TCPIP and Gopher and Usenet and like all these really old school things. And I got super into it as a. Geeked out, I guess. And then in high school, I started working in the summer for this company called Public Access Unix. Public Access Unix or Panix for short. It was the first Internet provider on the east Coast. It was like the east coast equivalent of the. Well, and ever since then, like I've been into into technology. I went to Carnegie Mellon in Pittsburgh because they had the largest wireless network like in the mid-90s. And I tried to do a couple of wireless startups when I was there. When I graduated, I got into wireless security. Eventually had wanted to start my own company and raised money from ipg. And so that's how, that's how I got into digital advertising. But it didn't study digital advertising. Have no formal studies in marketing or anything like that.
- [27:15]
A
But yeah, I love that your origin story was with a book, a physical book that brought home about the Internet.
- [27:22]
B
It was like an O'Reilly TCPIP. Like it was, it was really old school. It was fun.
- [27:28]
A
Amazing. Well, let's get a little philosophical about the concept of quality, which we talked a little bit about during the first half because like, the idea with attention metrics is that they try to go beyond basic impressions and clicks. You want to measure real engagement, like how much genuine focus a user gives to an ad, but not just based on time, based on quality. And it makes sense logically that people need to pay attention to something for it to have an impact. But there have been questions about how, you know, whether that attention really genuinely reflects value. And I ask this or do this little preamble because there's been research, like last year there were a couple of papers put out by Kroger and the ARF which found that things like viewable time and share of screen and video audibility, like different attention metrics, don't necessarily correlate with increased sales or improved brand outcomes. And I mean, it's obvious that consumers need to pay attention at least somewhat to an ad for it to have any kind of impact. But the theory, at least from those findings, from those papers, is that maybe attention metrics are better for planning rather than direct optimization goals. But what's your take on that? I mean this is like my gift to you. Go, go forth. Like I'm sure you have opinions here.
- [28:51]
B
So that that article, I think it was an adage by Jack Neff that was about the Kroger Interesting. You know, come on. I do know that left a mark in my, my brain. So that article was about the Kroger RMN test of one of our competitors products, a legacy verification company who doesn't? I mean they're not as focused as we are in attention metrics. I, I, I'll push back a little bit on what you said about engagement. I think that engagement you should not optimize towards because you end up with the, the audience bias and creative you shouldn't optimize towards attention because you end up with you know, suboptimal creative. You want creative that like does the most work, not creative that captures the most attention. So you asked a lot around. Let's talk a little about standards. The I think standards are important but standards make for poor currencies because if the sell side of any market understands exactly how a metric is calculated, they will game it and then it won't be as useful to the buy side. Look at FICO score for example. Like if you knew exactly how your credit score was calculated, you would probably behave in a certain way which is inauthentic and doesn't really represent your credit worthiness. Or think about the Google search engine results page algorithm. You know publishers are trying to game that all the time and if you as a publisher at Ad Exchanger knew exactly what it took to get the number one article, if someone searches for attention you might write the write the article differently in an inauthentic way which didn't deliver the best value to the people who are using the search engine. So I think that standards are important but more unlike for protocols and more for the structural things in markets rather than for currencies. I think that currencies should be made in service of the buy side and it should in you should. The first step of creating a market driven currency is to, is to show the buy side that this new metric is a better reflection of quality and drives more incremental outcomes than the old metric. And at that point the buy side has asymmetrical information that they can go out into the market and use to find bargains. Right? They can go out and they can find high value things that are underpriced. The more buyers that have that information, right? So imagine if you And I were the only people that knew what high quality inventory was out there. We could go have, we would have a field day, right? We could go out and, and make a lot of, or create a lot of alpha, right? Create a lot of asymmetrical information that you and I could use to buy a lot of bargains. Then as more and more people know about, have that data and more and more people are also doing that arbitrage, those bargains get more and more expensive. So the alpha that's created by the new metric gets compressed and the price of goods approaches their value. At that point certain buyers say, you know what, this arbitrage stuff isn't worth it. They go to the sell side and they say, hey, will you guarantee me what I'm buying using this new metric instead of the old metric? And at that point you have a market driven currency instead of a sort of a standards driven currency. Standards driven currencies like viewability, like they got us the web of today. Like I think that they've, like we've had a shot with standards driven currencies and now we need more of a market driven currency that buyers have proven drives more incremental outcomes and sellers can trust. And I think, Let me, I'll talk about that first then because I think it's super important. One of the benefits of using attention as a placement quality score and sort of a probability of attention to any creative in a placement is that it's deterministic before an ad is ever served. So if a publisher knows that an AU rating of a placement is 45, every impression that's served out of that placement should get a 45 AU rating. Now our technology isn't perfect yet, but we're getting very, very close. And it's important because then as a publisher you can have much more confidence in what's going to happen before a campaign ever starts. You know, advertising is sort of a weird industry. It's one of only three industries in the world where someone will spend $10 million and then say, I wonder what I got. Let me go look at the reporting and see the other two industries are gambling and venture capital, right?
- [33:20]
A
Every other industry form of gambling.
- [33:23]
B
So I guess it's really just advertising and gambling, right? In every other industry like you go and you say I want this quality and this quantity and I'm going to pay you this price. And then if you don't get it like something bad happens to the person who sold it to you. Imagine you go to your stockbroker and you say, I want 100 shares of Apple. And they say, alison, great news, it's over. Benchmark. You got 95 shares of Apple, three shares of some weird fruit company, and two fraudulent shares that don't actually exist. Right. If that happened in the stock market, someone goes to Capital P prison. But in advertising, it's just like, oh, 10%, that's totally fine. So I think that this ability for attention to be derived into a placement quality score, it fixes a lot of the bad incentives that are created in the market by focusing on cost. And it creates this shared understanding of quality. But it also really opens publishers up to having much more certainty about how they're going to be measured, which they haven't had to date, even with something like Nielsen who tells them after the fact what their ratings are. In this way, it's a much more viable currency.
- [34:34]
A
This is totally unrelated, but in terms of information sort of asymmetry. My mother has an ebay shop and she buys and sells jewelry. She knows so much about what she buys and sells, and there are so many sellers that don't know what they have. And she has really no compunction about buying something that should be worth $200 for like 20 bucks and then flipping it for 250. She will do that all day, and she absolutely loves doing it. And she brags about it to me all the time. And. And yeah, I mean, there's also that too, right? Like, you need to know what you're selling, you need to know what it's worth and to stick to that.
- [35:18]
B
Yep. But that's a perfect example. Imagine like two or three of your mother's friends figured out her secrets and they were also bidding on the same ebay auction. She wouldn't get it for 20 bucks anymore. You know, maybe it would be 70 or $80 and the arbitrage wouldn't be as worth as much to her.
- [35:34]
A
Luckily, her friends don't know how to use ebay, so it's fine. What's the latest on industry standards for attention metrics? Because I know the IAB and the MRC have been working on guidelines and you guys have been very involved with that to create like a framework for standardizing how attention gets measured and reported. Super complex process. Because attention, like we were just saying, is not any one thing. I think there was a public comment period this summer, but I'm not sure where it is right now.
- [36:05]
B
Yeah, so the IAB and the MRC have been doing amazing work sort of herding the cats of the attention vendor space and trying to get people to agree on what attention is. What is attention measurement? How do you use it for audiences and creative and media? So that's going to be a very long process. And I think just having shared definitions and a shared understanding of what we're talking about will come out of that. And that's good. I don't think we're going to see currencies come out of that process though.
- [36:40]
A
When are they going to be finished? Is it soon or are they done already?
- [36:44]
B
I don't think so. It's going to be a long process, but I'm not certain.
- [36:50]
A
So I keep quoting stuff that appeared in Ad Exchanger, but what the heck. We had a column from last year from someone at Cognitive with a very provocative headline, why Standardization is Wrong for Attention Metrics. And her argument is that standardizing attention metrics like viewability is just an oversimplification of something very complex, a very complex signal that's influenced by context and creative and a person's mindset. And you just, you can't put a score on that because of how much it varies by campaign goal and format. So you make a rigid, rigid standard and it could limit supply, it might increase costs. So what's your take on that?
- [37:31]
B
Yeah, I mean, I, I don't think they're wrong. I think the conclusion, I'm trying to remember that article, but I think the conclusion that they reached saying that it could not be a currency because it couldn't be a standard is incorrect. Like I said before, there's two ways to get to currency, right? There's the market driven way and then there's the standard way or the fiat way, which is just some people negotiate on a metric. You should never like metrics should be driven by the buy side. And if buyers trust a metric and they know that it's a proxy for the outcomes they want, then cool, they're going to use it no matter what anybody says. Like, that's one of the nicest things about markets is like people can have opinions about things and they can write articles about things, but it just doesn't matter. Like the market's going to do what the market's going to do. And I'm really proud that like more and more of the market is adopting AU as their standardized measure of media quality. And I think that it becomes more inevitable every day that we're going to start to see, see more people trading on AU and that will establish this sort of function of being a currency.
- [38:37]
A
So I'm going to change topics away from attention A little bit to get nerdy on some other interesting stuff that you said at this Luma event that I went to. We were both there, of course, earlier this year. You were not on stage. They just set up little like director's chairs basically and they're very interactive sessions. And I happened to be in the room for your session and we wrote a little story about it. And you sort of blasted the online advertising industry for like lacking any clear definition on what counts as quality. And of course like that just leads buyers to focus too much on cost. We, we've talked about that. And creates like a lemon market because then you just get like tons and tons like floods of low value ad supply. But then publishers keep adding more and more ad formats to meet supposed demand. And it's another spiral. How many death spirals are publishers engaged in right now? And there's also no way for buyers to hold sellers accountable on quality because there's no common language or standards for media opportunities, which is also what we were just talking about. I don't think you're wrong, but it does paint like a pretty dire picture. So like what has to happen so that publishers aren't compelled to participate in facilitating their own demise and publishers aren't like being pummeled with low quality crap inventory, albeit it's fair to point out that they're a little bit creating that situation. But why can't we have nice things? That's my point.
- [40:05]
B
I don't blame publishers at all. Sellers are going to create what buyers want, right? So if the buyers say we want the cheapest viewable impression possible, then it forces publishers into making this unfair compromise. Now the sort of more disingenuous long tail publishers will be happier to game those metrics and to make MFA websites than premium publishers like the New York Times or the Wall Street Journal will. And so it creates this really unfair compromise. That's the problem we're trying to fix. We're trying to give the buy side transparency into quality so they can justify paying higher prices for higher quality things, which then creates incentives for the sellers to make more high quality things. Like there is to your lemon market comment. There's this paper from 40 or 50 years ago by a guy named George Akerloff called the Market for Lemons, which was about the used car market in the US and it very clearly walks through what happens when there's opacity and quality. And that's the basically buyers bid at like below 50% of the expected value so they can make sure that over time they're not being ripped off. And the used car lemon market in the US was fixed by carfax. Right. Which was a tool that initially buyers used. So when I was in high school and I was buying my first car, I used a Carfax report to understand what was the quality of these cars. Over time as everybody was using Carfax, eventually buyers would go to sellers and say, hey, I expect this car to have a certain Carfax rating. And now when you go to a dealer, all the dealers have all the Carfax reports ready for you. So it effectively became a market. Carfax became a market driven currency that fix the lemon market.
- [41:47]
A
So you're saying we can have nice things, we just, I guess don't want them yet.
- [41:51]
B
Well, as long as you use Adelaide. Well, I shouldn't. Sorry.
- [41:55]
A
Oh, Mark. I'll allow it. Okay, so we're nearing the end of the episode and I'd like to do one of my favorite things on the podcast. If the guest is game a lightning round and I want excellent answers that are five seconds or less. Otherwise I'm going to mute you.
- [42:16]
B
Perfect.
- [42:17]
A
So let us embrace the spirit of the lightning round.
- [42:19]
B
I'm ready.
- [42:20]
A
Is the open web dead?
- [42:23]
B
It's not dead at all. Read Luke Lambert's comment about how Amazon looks at upper funnel media buying and they use the open web and he wants to buy as much of it as possible. This is in the DOJ thing that was this last week.
- [42:37]
A
Oh yeah, the remedy phase. He was on the stand. That is correct. Do you personally use AI search now? More than traditional search, I use search.
- [42:48]
B
For search, But I use ChatGPT multiple times a day for harder problems.
- [42:53]
A
Do you use an ad blocker?
- [42:55]
B
No. That's horrible. Who in our own.
- [42:58]
A
I know people.
- [43:00]
B
Who do whatpocracy.
- [43:03]
A
You know what it's like. It reminds me of all the big tech executives who don't let their kids use Facebook or play with devices. So kind of has a bad vibe. A little bit AI. More hype than help right now?
- [43:19]
B
No, it's tons of help. I think that people that say it's an auto completion tool or like the next word guessing tool are missing the big picture.
- [43:29]
A
What metric would you like to kill with?
- [43:32]
B
Fire, viewability.
- [43:35]
A
Cats or dogs?
- [43:36]
B
Dogs.
- [43:39]
A
Okay. I have a dog and a cat, but my cat predates my dog. My cat is sitting to my left and the dog's on the floor. Do you have. Do you have a dog?
- [43:48]
B
Yes, I have a husky. I've been in New York for two weeks. My Husky's been staying with my parents and we're going to go pick her up this afternoon. So looking forward to that.
- [43:56]
A
Exciting. The ad tech buzzword you wish you never had to hear again.
- [44:03]
B
Carbon.
- [44:05]
A
Carbon. I haven't heard that in a while. Actually. People seem to have forgotten about sustainability.
- [44:10]
B
Maybe my wish has come true.
- [44:14]
A
And last one, one word to describe the state of programmatic today. It can be hyphenated if you want.
- [44:21]
B
Inefficient.
- [44:22]
A
Okay, thank you. That was an efficient lightning round. People never take me seriously. They just answer all the questions as if they're regular questions. My guys, this is a lightning round, please, last question. This is not the lightning round. Putting aside the paradox of time travel, if you could travel to one specific point in history and make one strategic change that could alter the online advertising industry, ad tech for the better, when would you travel to and what would you change? And don't worry because you won't have a Marty McFly moment with your mother or anything. This is hypothetical.
- [45:03]
B
So, you know, I would say. I was going to say viewability again, but I think I would go back and I would help. I would have a conversation with Nandini when she was dragging the advertisers who ended up on Breitbart over the Kohl's about what are the long term effects of brand safety. So I think that, you know, the work that sleeping giants did is super important. But you know, it was the early work around like the advertisers showing up on Breitbart that created this market for the verification companies to make brand safety products that have just wreaked havoc on news publishers. Like I think personally, I think that brand safety could just be a list of websites that you should and shouldn't advertise on. But it's hard to make a venture backed much less so a public company off of a list of websites. And today some of the highest, the best bargains in digital media is news content on premium publishers. So in some ways the best value for verification data is just to fade it, right? Buy unsafe inventory, low viewability, high impact ads on news content at premium publishers. And it's going to be a huge bargain because brand safety has concentrated all the spend on like recipe sites and garbage like that.
- [46:21]
A
I like that answer because I've had a lot of people on and I've asked that question and so many people have said I'd go back and say that Google couldn't acquire DoubleClick. I'm like, all right, guys, yes, yes. But that's a thoughtful answer and I appreciate it.
- [46:38]
B
Let me be clear. Nandini and Clara are amazing, and Ariel's amazing. The work they're doing is super important. I just think that a little bit more nuance in the beginning would have made things a little bit less extreme today.
- [46:51]
A
And I hate to be a pedant. However, I know this because I had Nandini on the podcast and I mispronounced her name. Yes. Don't worry. No, she's used to it. She's very. She's chill about it. Don't worry.
- [47:03]
B
I mean, I said ari Paparo for 10 years right before he finally was like, Eri.
- [47:10]
A
I think he accepts Ari and Eri. I think he answers to both at this point. Well, Mark Gouldiman. Guldeman. How do you.
- [47:20]
B
It's like a ghost. Guldeman Gouldiman.
- [47:23]
A
See, because I thought it was Guldeman for the longest time. So there you go. We're all going to walk away and know how to say each other's names. And regards to your husky.
- [47:33]
B
Thanks, Allison. It was great, great being on the podcast.
- [47:41]
A
It.