← AdExchanger
AdExchanger
A Standard-Bearer For Standards
00:49:46
What do toasters have in common with transparency standards for online advertising? Quite a lot, actually, says Ben Hovaness, chief media officer at OMD Worldwide. When people can trust a set of standards – and companies conform to them – the less...
Loading summary
Transcript137 lines
- [00:00]
A
Foreign.
- [00:13]
B
Welcome to Ad Exchanger Talks, the podcast devoted to examining the issues and trends in advertising and marketing technology that matter most to you. This is Alison Schiff and welcome to Ad Exchanger Talks. Let's talk about some Ad Exchanger y things on the agenda today. Transparency, the need for it and the lack thereof. The pros and cons of principle based media buying, creating standards for ad auctions and retail media and the thorny challenge of how to evaluate AI powered advertising tools when they operate as black boxes. My guest this week and our guide through the thicket is Ben Hovines, Chief media Officer at Omnicom owned media Agency Network OMD worldwide. But first, please save the date for CTV Connect, taking place March 12th and 13th in New York City. And hey, don't just save the date, snag yourself a ticket. Ad Exchanger is joining forces again with synopsys Ad Monsters and chief marketer to host a can't miss summit on all the key issues and opportunities in Connected TV. Learn more and register@ctvconnect.com hey Ben. Welcome to the podcast.
- [01:31]
A
Hey, thanks for having me.
- [01:32]
B
All right, so lay it on me. What is one thing about you that not a lot of other people already know?
- [01:39]
A
I was actually a pretty serious violinist in my youth. I played for 10 years. I did music camp over the summers and everything.
- [01:47]
B
Do you still play?
- [01:48]
A
No. It's one of my few significant regrets is that I gave it up after college.
- [01:54]
B
If you picked up a violin today, would it just come back? Is it like riding a bike?
- [02:00]
A
There's a certain riding a bike element to it, but there would be a lot of rust on this particular bike.
- [02:05]
B
Okay. Well, I'm not musical at all, but I really appreciate music. And one of my great regrets is that my parents didn't really care if I did activities or not. Not in a neglectful way. They just let me do whatever I wanted. So I was never in piano lessons or. I don't know, I never really did anything like that. I never took a second language very seriously other than the sort of pathetic way that you learn it in high school. And now I'm. I'm without skills. I'm without talents in that regard. I can knit. I taught myself.
- [02:39]
A
That's a very practical skill.
- [02:41]
B
Well, I can't purl, though. I can only knit, so I can do the one stitch. But I'm quite good at it. They're very consistent, my knitting stitches. So you've been at Omnicom for a while. I'm just going to do a little rundown of your resume for our listeners. But you went from Associate Director and East Coast Social Lead at resolution Media in 2014, which Omnicom owns, to Executive Director of Biddable Investment and then Executive Director of Digital Activation at Hearts and Science, which is also an Omnicom shop. And then there were a few roles at Omnicom Media Group, SVP of Market Intelligence among them. And now you're the Chief Media Officer of OMD Worldwide for the past year. So my question is, having moved up the ranks, do you ever still get to do the hands on keyboards thing? And if you don't, do you miss it at all?
- [03:36]
A
I don't do the hands on keyboard thing anymore. What I miss about it is isn't so much the day to day like, okay, I need to log in and I need to re optimize this campaign or I need to launch something. I do miss the fact that it gives you an immediate currency with the state of a platform. Right? Like if you log in, you will see the current feature set. You don't need to go ask someone what's changed. So I do miss that sense of direct connection with the platforms. And so now in my current role, I try to make up for it by maintaining really tight relationships with our practitioner leaders, like our heads of Search and Social and programmatic and so on. But it's not quite the same. Of course.
- [04:19]
B
I was going to ask you how you keep yourself up to date and clued in. I guess you also read a little thing called Ad Exchanger.
- [04:26]
A
I do. I am an enthusiastic reader of Ad Exchanger and I have been for many years. But in terms of how I keep myself clued in, more broadly, it's a mixture of staying close with subject matter experts inside of the organization and also just maintaining tight relationships with our key partners.
- [04:45]
B
And it's the case in most jobs that when you rise through the ranks, you just spend less time in the trenches. I mean, it's so true in journalism. Also, if you're an executive editor, you're rarely writing unless you make it a priority because you just really want to do it because you love it and usually don't have time to do it. So you take on a story and then feel really guilty because you never end up filing it and then it becomes obsolete whatever you were writing about. So it's a very typical situation. So I just have a few casual questions about the Omnicom IPG acquisition, like how are they going to combine their data capabilities? What happens to IPG's axiom is the ADSTRA lawsuit against Canneso going to be an issue. How's the acquisition going to help you compete with big tech platforms? Are there going to be any client conflicts? How are you going to deal with them? How long is the integration going to take of data assets and people? Are there going to be layoffs? You know, just lay it on me. Just a few questions, right?
- [05:48]
A
Nice and light. Well, Alison, those are all excellent questions, but unfortunately I can't comment on the acquisition while the deal is being finalized. Nor can I speculate in any way about how the acquisition is going to impact Omnicom, the broader marketplace, or anything else.
- [06:04]
B
How many times have you said. Have you had to say that exact sentence?
- [06:08]
A
This is actually my first time. How was it?
- [06:11]
B
It sounded very practiced and polished.
- [06:14]
A
Outstanding.
- [06:15]
B
All right, fair enough. And I know there's really nothing you can say about the acquisition, and I did know that going into this, but zooming out and not at all specific to Omnicom or ipg. Do you think we're in for a wave of agency consolidation over the coming year? Ad Exchanger had three people at ces, not me. I know you were there too, and that was the buzz they were hearing. It came up in multiple of their conversations.
- [06:42]
A
I think it really comes down to whether or not there's an appetite in the marketplace for differentiated agency strategies. Without getting too much into the specifics on this, on this podcast, I think what we've seen over the past couple of years, if we look at which agencies have been having a lot of success in the marketplace in terms of new business, they haven't necessarily had the exact same strategy, which means that different stories are connecting with different clients. And I think as long as that continues to be the case, there's going to continue to be room for multiple competitors. I don't really see like an unlimited accretive logic where we just end up with two mega agencies. As long as it seems like there are multiple different narratives that are good fits for different advertisers. I know that was a little bit vague, but hopefully that's helpful.
- [07:31]
B
I think the indie agencies would be happy to hear that sentiment expressed. Speaking of ces, what was the. Because you were there, right?
- [07:41]
A
Yeah, I was there.
- [07:42]
B
What was the coolest thing you saw or heard and what was the most ridiculous thing you saw or heard?
- [07:49]
A
The coolest thing I saw was actually not advertising related. It was an agricultural robot from a company called Kurobota. If memory serves. It basically looked like a Mars rover that had been adapted to a farm context. And so the they had it in a little fenced off demo area. There was a conical mound maybe 3 or 4ft high. Balanced on top of this Mars Rover type agricultural robot were two big baskets of walnuts that were piled over the brim of the basket. So very easy to spill. Right. And so this rover starts climbing up the side of the mound and it's like a 30 degree incline because that's how steep the mound is. And the bed of the rover is completely level. And the pitch was. In a world of agricultural labor shortages, this robot can be a key part of the solution. So that was one of those CES moments that felt like stepping five or 10 years into the future. I know it's not advertising related, but that that was the most futuristic moment I had at the show. And that's what I enjoy about going.
- [08:55]
B
Was there a screen on it? Could be a media opportunity.
- [08:59]
A
There wasn't a screen on this model. In terms of screens and new media opportunities, I think the obvious thing at CES this year, as it has been for the past couple of years, is the evolution of automotive displays. A lot of companies, including BOE and some others, were demoing very sophisticated in car displays. There's even a fold one that folded in half sideways and then would retract into a glove compartment. It was very James Bond. So I certainly think over the next five to 10 years, as these become more widely available and also propagate down market, like you're not only seeing these in $80,000 cars or whatever, it'll start to become a new endpoint for advertising and that's going to become very interesting.
- [09:43]
B
Have you ever ridden in an av, an autonomous vehicle?
- [09:47]
A
I have not yet. I want to and I have not yet had the opportunity.
- [09:51]
B
I haven't either. I think I read something about Uber possibly including some AVs in their fleet now, or maybe it's coming soon, I really can't recall. So this is a bad story, but I think you can request one in certain markets.
- [10:07]
A
Yeah, I think in markets where they're authorized to operate. I'm New York based, we don't have them here yet, but I am eagerly awaiting their arrival.
- [10:15]
B
I wasn't at ces, but this is also a kind of cool thing that is not advertising related that I read about, that I want, but I definitely don't need. And I can't decide if it's ridiculous or low key brilliant, but it's this little gadget, it looks like a cat and you clip it onto the side of your coffee mug or plate and it Blows on hot food or a drink to cool it down so that you can eat it more quickly or drink it more quickly without burning your tongue. It goes. It's really cute. But like, can't you just blow on your own coffee or whatever? That's fantastic. But it is adorable. I think it is a Japanese company. It's very kawaii, you know.
- [10:59]
A
Yes, it is.
- [11:01]
B
So, yeah, zero segue from that to industry calls for transparency, but that is what I want to spend some time talking about because I know you've done a lot of work on this, but as more of a scene setter first, I mean, advertisers, they obviously want more clarity on the quality of their impressions. They have questions about verification, increasingly loud questions, media costs like how programmatic spend is being allocated, their supply chain complexity, questions about data accuracy, demands for access to log file data if you want to identify inefficiencies and even malfeasance in some cases. But even though all of this is obviously very important and we've spilled I don't even know how many gallons of digital ink on the topic of transparency. I kind of feel like the word transparency is almost starting to sound like background noise or a buzzword, like, oh, marketers are lamenting their lack of transparency again, you know, like this. Here's another story about how we need transparency and I don't think that's a good thing. So obviously it's important to talk about a topic, but you don't want to become like a droning sound in people's ears or to, I don't know, fall into this place where the more you talk about something and the more change hasn't happened, the more cynical people get. So I'm just curious what you think about that and also what you think when you hear the word transparency, like, what it really means to you.
- [12:33]
A
Transparency is in the eye of the beholder. It means different things to different people. From my point of view, transparency expressed certain ways, is critical to a healthy relationship between an advertiser and an agency, and certainly between advertisers and the people that they're ultimately buying media from. Back in 2020, when I was in my marketplace intelligence role at Omnicom, we launched a program called casa. It's. It's short for the Council on Accountability and Standards in Advertising. And it was based around laying down some pillars of advertiser rights, things we believed advertisers were entitled to. And I'm not going to get into all of it, but a key thing that we believed advertisers were entitled to from sellers was knowing where exactly their media delivered. And so in the context of a social platform, that means what adjacency was there with respect to the advertising. Meaning if an advertiser bought something in Facebook newsfeed, what was the content above and below it? Or if they bought a Snap ad, what preceded and followed the ad? So that type of transparency I think is critical. There's certainly other forms of transparency that we could definitely talk about.
- [13:50]
B
Well, there's CASA and there's also this MRC Auction Standards Working Group that you're the steering chairman of. Right. So tell me about that effort.
- [14:01]
A
Yeah, happy to. And I think, and we can explore this more. But there's an interesting relationship between the concept of standards and the concept of transparency. Because the more you have standards that people trust and the more you have companies conforming to the standards, the less transparency you actually need. Because then the standard has become a stand in for being able to review everything in detail yourself. So just something that I wanted to.
- [14:25]
B
Put out there, the standard has become a stand in is.
- [14:28]
A
Exactly.
- [14:29]
B
Let me take out my pen and make a note to myself that will go in the write up about this podcast. That's very quotable. Go on.
- [14:34]
A
But you know what I mean. It's like when you buy a toaster off of Amazon or wherever and you plug it in, you're not worried that it's going to burst into flames and destroy your home, because you know that wasn't before. Right, right. But I don't think you need to, because there are all these standards regulating the design and sale of electronics. So you're aware in some abstract sense that these standards exist and therefore you're not like, well, I need transparency into the exact specifications of this toaster and I need to see the electrical engineering diagram. You see what I mean?
- [15:05]
B
Yeah.
- [15:05]
A
Because you have that comfort that the existence of standards and your knowledge that they've been adopted provides. And so in the same vein, so you brought up the standards effort that I've been leading. So the background on this is that ad auctions are very, very important because they are the beating heart of modern digital advertising. And the scale of ad auctions is tremendous. This isn't widely known and we couldn't find any good publicly available data on this. So we, we built out our own bottoms up internal analysis of this and the analysis indicates that presently global ad auction spend is about 650 billion. And it's going to scale, based on past growth Trends, to about 1 trillion per year by the end of the decade worldwide. And contrary to popular belief, only about a quarter of that spend is in programmatic auctions. The other three quarters is in so called walled gardens or closed ecosystems, whatever term you want to use, like Google search or Meta, Snap, TikTok, whatever. Now, an ad auction has two jobs. First, set the price of an ad impression and second, determine what advertiser gets the impression. And we call those two jobs pricing and allocation. And different auction systems have different rules for pricing and allocation. There's a lot of variation. And so this is where the need for a standard comes in. How the prices are set is so important because it dictates how buyers should behave. So for example, and I promise I won't go all the way down the auction theory rabbit hole with you, in a first price auction where you pay what you bid if you win, game theory indicates that you should bid less than what it's worth to you. But in a second price auction where you pay less than what you bid if you win, game theory indicates a different strategy. But what if an advertiser doesn't know the rules of the auction game? How are they supposed to figure out how they should bid? Now it might sound like a crazy hypothetical, but that actually describes the reality of most of the advertising world today. Most auctions are not fully disclosed in terms of how prices are set. And that also means that advertisers can't be sure that a seller isn't manipulating that ad auction to boost their revenues at an advertiser's expense. And so historically there's been no standard governing how ad auctions get disclosed either. What are the rules up front before advertisers play the auction game? Or standards about what gets disclosed after the fact? Like okay, well how did this bid that I input get translated into a final price that I paid? So because of that, about one year ago, I reached out to the MRC and some other key industry stakeholders like the Forays, the ANA and the WFA to try and solve this problem once and for all for the industry. And so as you said, that has turned into an auction standards working group under the aegis of the MRC's normal standard setting process, Chairman of the Steering committee. But I'm very proud to say today that we have more than 70 companies from around industry from all over the world participating. So it's been a great process so far and we're going to have a lot more to share publicly later this year, including a draft and a final standard.
- [18:12]
B
I think it is really noteworthy as you just pointed out that there are so many companies involved because there are often opposing points of view that players on different sides of the supply chain will have. What's good for buyers isn't always good for sellers, and vice versa. And you also have other holding companies involved here. But standards will help everybody. It's not easy getting everybody to work together, though. It is great that everyone's signed up, but I'd imagine that there are competing interests that come up when you guys meet. And getting to something that even vaguely resembles consensus is really hard. I mean, I, I have compassion for Tony Katzer at the Ivy Tech Lab.
- [18:52]
A
Yeah, he's actually part of the steering committee and he's, he's been a terrific contributor to the process. Now that you mention it, as you.
- [18:59]
B
Were sharing your analogy about the exploding toaster, I was thinking, yeah, but at least in the case of an exploding toaster, there's immediate evidence that something is seriously wrong. And you write an angry letter, get a new toaster. But it's probably fairly easy to keep spending in an opaque ecosystem and not really looking under the rock, not really understanding what's happening in the sort of pea soup of the supply chain unless you make a real effort to dig in. You could just keep spending and keep going as you normally do, and it's not as efficient as it could be. There's just not like as much evidence as something exploding in your face.
- [19:38]
A
It's true. What I would say is that it has caused friction in agency advertiser relationships through no fault of either party. Because what will happen sometimes is an advertiser will say to us, hey, what happened to our prices in Google Search? What happened to our price in Meta? Like, whatever. The specific platform doesn't really matter. And then we'll say, well, we don't know. We didn't change anything in our strategy. Everything seemed normal. And then there was just this discontinuity of cost. So then of course, we'll go talk to the media seller and ask them, like, hey, we're seeing this, we can't really account for it. It doesn't seem like there's anything that should have led to this. What's going on? And because the auction is so opaque, even to the people we talk to on the sell side, they'll say, oh, it's the auction. And at certain times I've gotten so frustrated, I'll say, oh, you mean the auction that you designed, Maintain and control that auction. That's the one that you can't give any insight into. So even Though it can be hard to spot these problems, it does crop up. And the lack of an ability to get answers to these questions can cause conflict. And so that's some of what we're trying to address with this. Because the reason that we've been able to build such a large coalition around this is we're not trying to say, you know, you should only have these three different auction designs and that's it. We recognize that the world needs a multiplicity of auction types, if only because there's so many different app formats out there. Right. Like to give a really obvious example, search and CTV are going to have very different inputs because clicks are a vital input to how a search auction is run, and clicks are not a relevant concept in CTV advertising. Right. So instead of trying to specify this is how you have to run it, we're just focused on this is how you need to disclose it. Like, you design whatever auction you like. You just need to tell people what the rules are. And you need to make it possible for advertisers to understand what happened after the fact, to try to neutralize these very unpleasant conversations where maybe an advertiser thinks that the agency forgot how to run a search campaign or something. But the reality is something has changed in the auction process that no one is aware of.
- [21:48]
B
It's too funny saying, oh, it's the auction. As if man did not create this thing. We're not talking about a natural force. It's right.
- [21:55]
A
Exactly.
- [21:56]
B
What has man wrought? It has wrought the auction. You made it. You made it happen.
- [21:58]
A
Yes. This is not some mystery of quantum mechanics. It is a computer system built by people.
- [22:04]
B
So we're going to take a quick break and when we're back, we'll talk a little bit more about standards, retail media standards specifically, and also what's become a, I don't know, third rail term, principle based media buying. We'll talk a little bit about that, so stick with us. All right, we're back. And principle based media buying, the practice of buying media directly from media owners and then reselling it at a markup. There has been a lot of backlash against it because of potential conflicts of interest, and there's concern about how transparent the process is. And Omnicom Media Group does principle based media buying, but you've been very public about it, so it's not like some kind of secret, but to your mind, what, what is the case for and what is the case against principle based media?
- [23:06]
A
Well, let me start with a case against, because I am never one to shy away from an argument. And also just because I feel like it gets an unfair share of ink, it feels very cyclical where like clockwork, every quarter, sometimes every month, we'll see a story published in a trade publication decrying the practice. So I just want to take the most serious objections head on. So the most serious one is that the practice of principle based media buying can create a financial conflict of interest between the agency and the client. So, for example, if we would make more money buying from media seller A than media seller B, maybe we would recommend media seller A even if it's not a good fit for our client. And then the secondary argument against principle based media buying, which is less concrete, but I still think it's a reasonable argument, is that the lack of transparency about the agency's underlying cost basis for the media can undermine trust between advertiser and client. All right, so the case for principle based media buying is that it can deliver material financial benefit to clients. And there's really no magic to how it works. It's quite straightforward. We've been public about it. As you said, agencies have the scale to make hard dollar commitments to media sellers, and sometimes that will even extend as far as up front payments. Please note that this is distinct from a JBP or a spend endeavor, because the commitments that we're talking about for principal based buying contractually obligate us to pay media sellers a certain amount over a given time period. So those hard commitments, which by definition entail financial risk for the agency, are what unlock material discounts on the media versus the prevailing market price. And so principle based buying is what allows us to pass some of that savings onto our clients. All right, so those are the cases for and against. And I do want to directly address the against objections, especially the one about conflict of interest, because it is a real risk when you're designing a principle based media offering. I would freely agree that it is possible to design a principle based media offering that is just rife with conflict of interest issues. I'm happy to say that Omnicom's principal offering does not have those problems because it was designed at the outset to prevent that conflict from occurring. So, first of all, we have a firewall wall between our principal arm and our brand agencies. So that means that our principal deals, like who we have them with and what the terms are, are not known to the brand agency teams. So concretely, someone working on an account here has no idea what media sellers we have principal deals with, so they don't have the information they would need in the first place, to steer budgets in a way that would make the agency more money. And second, as an additional safeguard, we have a double opt in process. So that is even after a client has opted in to our principal based offering. In addition, they have to sign off on each individual buy that leverages principal media. And then on the transparency point, whether or not having full transparency into the cost basis diminishes trust or not, that's something that we leave up to advertisers to decide. Now, my personal view is that it's completely possible to trust a business partner without having 100% transparency into their cost basis. Because that actually describes most commercial relationships. Right, like, keep going back to the toaster. Do you need to know what the toaster manufacturer paid for the aluminum and printed circuit boards that went into it to have confidence in the toaster? Probably not, because there are other things that make you confident in the toaster, but different advertisers have different points of view on this and not the toasters, but on transparency and media cost basis. And we respect that. And that's why our principal offering is opt In. So if an advertiser wants 100% cost basis transparency at the expense of foregoing the savings that principle based buying can offer, that path is open to them. We have clients that do it. It's completely fine. But I would note that the vast majority of our clients do choose to take advantage of our principle based offering.
- [27:01]
B
I do think the opt in, the double opt in neutralizes most of the arguments against, because we are all adults here. So read the information and make the choice yes or no. And trust informed, concise, informed consent. Exactly.
- [27:18]
A
Yes.
- [27:18]
B
Well, sticking with our transparency theme, like I alluded to before the break, OMD Worldwide is also pushing for retail media standards specifically. Is that through casa.
- [27:29]
A
Yes, we have a CASA retail arm. CASA is really about driving capabilities development. And then when it makes sense, we'll work with standards bodies to try to push that out into industry. To rewind a little bit. We've done the same thing with ad auction standards. So ad auction transparency was actually part of the CASA program for a couple years before we reached out to the MRC and these other entities. Because what we realized was that we were never going to have any success on this without partnering with industry. And so we're doing something similar in the retail space now, where we launched CASA retail, I want to say, two years ago. And that includes a number of capabilities pillars and where it makes sense, we're partnering with industry bodies to try to get Those to be adopted at a broader scale.
- [28:14]
B
I think the partnership aspect is so important because like you do have a lot of different players working on frameworks for retail media standards. You have Albertsons, they've been very vocal. The IAB has been working on stuff and so has the mrc. And then everyone's claiming their efforts aren't competitive, they're fully complementary. It would be the ultimate irony if all of these efforts to standardize retail media ended up creating standards fragmentation.
- [28:44]
A
Yeah, exactly. It was a formative experience for me very early in my omnicom career. There was a big push around viewability at the time. I'm sure you remember this. This was like in the 2014, 2015.
- [28:56]
B
Period, right when I started at Ad Exchanger. Some of my first stories were about the IAB and viewability.
- [29:02]
A
Yeah. And so there was a two second viewability standard that was adopted for video and one second for static. And there was one major agency that came out and said, that's not good enough. We're pushing out our own standard. This is nonsense, we're going our own way. And it was really damaging to the industry. I mean, sure, it got this agency some press for a time. So I mean, maybe that was good for them in the short term.
- [29:25]
B
No doubt wrote about it.
- [29:27]
A
Yeah. So no doubt it was good for them in the short term, but it caused harm to the industry and really delayed the adoption of the MRC viewability standard by a number of years. And that always stuck with me and it really informed our approach to taking on ad auctions and other issues. Because the way to actually create a standard is to do the painstaking long term work of putting together a broad buy side coalition that can develop a standard that will actually achieve broad adoption. So I'm actually proud to say that we have participation from every holding company in the Auction standards working group. We, we even have someone from another agency on the steering committee.
- [30:08]
B
What about today though, the experience of buying retail media? I've asked this question of so many people, like, have we reached peak retail media? And it always feels like, yeah, we probably have, but no, the answer is no. I mean, just recently Mohegan the Casino launched a casino media network. I mean, there seems to be a new one at a very regular cadence. They're more than 600 or so, so it's obvious why the industry needs retail media standards. But they're still in the works. So I mean, without them are buyers just consolidating their spending on a handful to preserve their sanity? Or how does that even work.
- [30:48]
A
It really depends on the advertiser and it is an important part of our service offering to help them navigate the increasingly fragmented retail media landscape. But we've got partnerships with every significant retail media network out there. We're continuing to ink new agreements and develop new capabilities that are exclusive between Omnicom and them.
- [31:11]
B
Well, I want to talk about AI now. AI powered ad products like Performance Max, which is Google's thing, and Advantage plus, which is Meta's thing. And I've made this joke before, but doesn't it sound like a health plan? Doesn't it?
- [31:27]
A
Anyway, I guess Advantage plus does sound like a health plan now that you mentioned.
- [31:31]
B
Yeah, I can't argue with you, but yeah, lots of other big platforms have their own versions of PMAX and Advantage plus and they're on the ascendancy. They're also black boxes. And then meanwhile, I'm stealing this insight from an Arete research note that I read recently. But they're predicting that we're going to see AI used more and more to cut down on ad tech arbitrage and non working media. That'll be a big trend in 2025. And I just find those two trends to be very interesting because you have AI as a catalyst both for more transparency and then also as the reason why there's less transparency now into how some media dollars are being spent. Really the majority of biddable, as you, as you mentioned, is going to the walled gardens.
- [32:19]
A
So the buying agents are fascinating. And I spent a, I spent a lot of time over the last year working on that particular area. The first thing was just coming up with a proper umbrella term for what these things are. Like, what would you call PMAX, Advantage plus and Smart plus from TikTok and all the other similar offerings. And so where we netted out was AI buying agents is a good umbrella term for these things. And, and then the next step from there was, okay, well how would we evaluate these things if we wanted to take an objective look at them and see what are the pros and cons? And so we came up with five evaluation pillars and then used that as the basis to launch a new branch of casa. We rolled that out in the marketplace in October of last year during Ad Week New York. But here's what it boils down to. Number one, is the agent constrainable? Like can the advertiser constrain the AI buying agent from buying an audience that the advertiser knows just won't work for reasons that the agent may not understand? Can the advertiser constrain what creative the buying agent is using or what combinations of creative elements. Can the advertiser constrain the agent's approach to bid strategy? What about placement and so on and so forth. So constrainable. Very important. Second is directable. That just means what objectives can the agent pursue effectively? And that's more just a gauge of capability because over time we expect each of these agents to support more and more advertising objectives. Another one is transparent. To your point about it being a black box, Alison, to what extent can the advertiser see delivery and performance after the fact? Can the advertiser see the details of the audience profile or delivery and performance by site and by placement? Because historically that has not been available and in some cases it creates a contractual block to us using one of these tools. If we have a contract where we're obligated to disclose site level reporting to a client on demand, which is not uncommon, then we can't use some of these tools without a waiver or without changing the contract. Fourth pillar is verifiable. Like is there a third party that can come in and confirm that the agent is working as declared? And then fifth one, the most important one in my opinion is declared alignment. And what we mean by that is that we expect the operators of these AI buying agents to disclose whether or not these systems are designed to consider a seller's own commercial interests as a decision factor. And it's so important, kind of analogous to the point about auctions and game theory earlier, for advertisers to know if these AI agents are working faithfully on their behalf or not. Because if you have a faithful agent, you should give it different information than you should give an unfaithful agent. So game theory of this is if you know your agent is working for you faithfully to do the best it possibly can for you without considering the agent's own interests, then an advertiser should be truthful about their willingness to pay for a business outcome, whether it's a conversion or a site visit or whatever. But if you think your agent is not faithful, then actually the optimal strategy for an advertiser is to be deceptive. It's to be misleading about willingness to pay on the downside in order to secure the maximum benefit. So we're pushing partners hard to come out with declarations of alignment. It's been a really great set of conversations because it's an exploration of a problem that agencies are very used to, which is the principal agent problem. Like how do we make sure that the agency's interests are very closely aligned with our clients interests. Right. Because we're different businesses. So we have to do some work through various things like incentives and, and structuring the relationship to make sure that our incentives are as aligned as possible. That's a new conversation for some of these platforms to be having. But philosophically it's very much the same.
- [36:02]
B
And the most well known of these agents, pmax and Meta's health plan, do they tick all the boxes or. That's a journey.
- [36:11]
A
It's a journey. But the good news is that the direction of travel is positive, the same as it has been with casa. You know, we launched CASA with CASA social back in 2020. It took about two years for all four pillars to see wide adoption, which, depending on how you look at it, is a long time or a very short time to get these very large organizations to implement substantial product changes. So I think this is going to play out similarly. But the important thing is just to have the right framework and good relationships. And just to start.
- [36:43]
B
Do you use any AI tools yourself in your day to day? Something that's made your life better?
- [36:51]
A
Sometimes I'll use GPT as a sounding board or I'll use it to condense some prose. Those are probably my two primary use cases today.
- [37:01]
B
I've extolled the virtues of otter AI on this podcast before. It saves me. I don't even know how many hours of work every week because I don't have to manually transcribe interviews anymore, which is a horror beyond horrors to have to re listen to clips of audio over and over again until you get them down. It's just your own voice. Everyone's verbal tics. It's a nightmare and it's just so time consuming.
- [37:27]
A
Yeah, that's an amazing use case. Transcription has gotten so good.
- [37:31]
B
So just changing tacks a bit. You were on a panel at our Pro Guyo event in New York last fall. It was with me and Kaval Khan from Group Black and the inimitable Ari or Ari Paparo. It depends on. I think he answers to both. So we, we were talking about the future of the open Internet as one does at Pro Gu. And I want to re ask you one question that we talked about a bit on stage, which is it's meant to be a little facetious and a little provocative, but I think it's a legitimate question. Do advertisers even really like the open Internet, like in their hearts? Because they don't necessarily love it as much as they could with their wallets and actions speak louder than words because you can just look at all the money that's being shoveled at Meta and Google, just the billions and billions and billions.
- [38:26]
A
Well, I think one can never really know what's in someone else's heart. But if I had to guess, I would guess that advertisers do like the open web. I'd also say though, that advertisers like print media and broadcast TV too. So you probably see where I'm going with this. Ultimately, advertisers, and that includes businesses and NGOs or governments, they're all accountable for getting results out of their media investments. And in the long run, that is always going to outweigh their personal feelings, whatever might be in their hearts about a media channel. I think I said something similar on stage, but my view is that the only high ground for media is performance. And I don't mean, just to be clear, that every part of a media buy should be direct response or bottom funnel or whatever. But what I do mean is that every part of a media buy is increasingly being evaluated for its real contribution. Is it delivering a measurable improvement in awareness, favorability, likelihood to recommend purchase, repurchase, loyalty, whatever, it doesn't matter. The point is that money will always flow to the media properties that deliver on advertiser objectives the best. And that's why part of our philosophy at OMD is that all media is performance media. That's how we have to look at it.
- [39:38]
B
Right? Or all media can be performance media.
- [39:41]
A
Right.
- [39:42]
B
It just depends on your strategy.
- [39:43]
A
Exactly. But if you're just evaluating media on the basis of like, how many impressions did we buy? You're missing something.
- [39:50]
B
I evaluate media based on nostalgia for a simpler time. That's my metric.
- [39:57]
A
I do have nostalgia for the glory days of print. Back in the 90s.
- [40:01]
B
I did actually write for a couple of print publications back. Back in the day. I mean, I don't know, like 17 years ago now. I wrote about certified public accountancy for a CPA society in New York for their newsletter. They called it a newspaper, but it wasn't really a newspaper. And there was something really satisfying, I'll say, about having a very specific word count and having to hit that word count. Like word counts really matter because you can't say too much and you can't say too little. It was an interesting exercise and I think it trained me for having a good sense of how a piece should go and what needs to be included and killing my darlings and all of that.
- [40:45]
A
Constraints breed creativity.
- [40:47]
B
Yeah, absolutely. I have a friend who writes in form. He writes poetry, and he doesn't ever write freeform. He writes like sonnets and villanelles because the constraints are satisfying when you can do it. Yeah, I don't, I didn't think I was going to talk about Villanelles on ad exchanger talks, but let's do a, let's do a lightning round. Like a for real lightning round. 5 second answers or less, if you're game to do it.
- [41:14]
A
Absolutely. Let's go.
- [41:15]
B
All right. Do you still care about the Chrome privacy sandbox?
- [41:20]
A
Yes. Mostly out of nostalgia.
- [41:23]
B
Okay. Do clients still care or ask you about the Chrome privacy sandbox?
- [41:29]
A
They used to, but not anymore.
- [41:31]
B
What was the first thought that went through your head when you heard in July that Google was no longer going to deprecate third party cookies in Chrome?
- [41:39]
A
Oh, come on.
- [41:42]
B
Yeah, same. Did you predict it, by the way? Did you? No, No, I didn't.
- [41:48]
A
I, I, I thought it was going to happen. I really did.
- [41:51]
B
I really did, too. There are so many episodes of talks where I say they're definitely going to do it though, right? They're doing it. They're totally doing it.
- [41:58]
A
Well, the problem is if you're at the agency, you don't, you can't be the person who says it's not going to happen. And then it happens.
- [42:05]
B
Yeah. Oh, my God.
- [42:07]
A
So we were just in this state of constant elevated vigilance for years, and then after the events of last year, we sort of stood down because that was it for us.
- [42:20]
B
I mean, you do have to focus on other things, for God's sakes. I mean, the amount of mindshare.
- [42:24]
A
Exactly. It's not that signal loss in the browser doesn't matter anymore. It's just we have to backburner it if there's zero certainty and total ambiguity.
- [42:32]
B
All right, back to the lightning round. What is the number one thing that most of your clients care about right now?
- [42:39]
A
Connecting with consumers to help their businesses thrive in a very dramatic era of change.
- [42:45]
B
Okay, what are they most worried about?
- [42:48]
A
They're worried about their ability to manage the complexity of the landscape and also how to keep their own talent up to speed.
- [42:55]
B
What are they most excited about and feeling most passionate about?
- [42:58]
A
Technology advances that lead to greater ad effectiveness and definitely greater efficiency in operations. And there's a lot of passion around AI right now.
- [43:07]
B
What are they bored with? Like, what's something that makes them want to roll their eyes?
- [43:12]
A
Stereotypical fluffy agency presentations. Just like a lot of buzzwords and hand waving. Can't stand It.
- [43:18]
B
Well, speaking of buzzwords, if you could kill one ad industry buzzword with fire and never have to hear it again, what would it be?
- [43:25]
A
Web3.
- [43:27]
B
I mean, I think you're already living that dream. I haven't heard anyone mention Web3 in a while.
- [43:31]
A
It still comes up periodically.
- [43:33]
B
Really?
- [43:34]
A
Yeah.
- [43:35]
B
I usually use my own work inbox as a gauge and a sort of litmus test for what's top of mind for people. Because if I'm getting pitched on it, it's either relevant or someone wants it to be relevant and it's not really relevant. But if I see a lot of pitches about it, it's a sign. And I haven't gotten a pitch about Web3 in quite some time.
- [43:54]
A
It's not front and center anymore. People still mention it in passing, though. They'd be like, oh, we're launching this new thing and it's got Web3 integrations. I'm just like, you had me until.
- [44:05]
B
Okay, how many more analytics reports will have to come out before there is real, sustainable and sustained change in the ad industry?
- [44:14]
A
I don't think analytics reports are capable of creating change in the industry because of their nature.
- [44:21]
B
Lightning round over. Expand on that.
- [44:24]
A
The way that analytics conducts their process and the nature of the reports makes it very, very difficult to ascertain what's true, what's not, and if any action should be taken at all.
- [44:39]
B
Do you feel like it almost stymies people because it confirms their existing fears and assumptions about the lack of transparency, and then they're not sure how to move forward, so they just don't do anything?
- [44:54]
A
I think that's true for some stakeholders, and then for others, the frustration is just the lack of ability to have a dialogue. It doesn't really seem like there's a good dialogue between analytics and the platforms that they're writing about. And so again, it makes it very, very difficult to determine. Like, are they publishing something because they don't understand what's actually going on, or have they found a real problem? And then that coupled with the extreme verbosity of their publications, it makes it very difficult for them to yield anything but flash in the pan of press.
- [45:30]
B
Though I will say we cover them all. But I do hear what you're saying. That's an interesting perspective.
- [45:35]
A
Well, but you know what I mean. And I don't mean to be critical of your coverage, but all of the coverage hasn't led to change. And I wouldn't say it's because people aren't open to change. I would say it's because of the nature of the work that's being done.
- [45:49]
B
One last question of my lightning round. Unless you feel like expanding, would a breakup of Google, quote unquote, fix the open Internet?
- [45:59]
A
No. To expand on that, I would argue that four things are what ail the open Internet the most signal loss, the publisher ad load death spiral where publishers put more ads on their site to get more revenue, but then in aggregate that turns to lower prices and lower revenue which then leads to more ads. Third, an unsophisticated ads marketplace. Like there's no persistence of creative right. There's no creative relevance score or anything quality score that gets baked into how ads are bought and sold and not very effective ad formats in a lot of cases. I don't think breaking up Google solves for any of those four things. Now you could argue, as I think Ari did on stage last year, that Google's very high market share of sell side technology has led to higher ad tech prices for publishers than they might otherwise have paid. And I agree that that's a very reasonable argument. But let's assume that ad tech fees paid by publishers drop by half in the wake of a Google Ad tech breakup. Let's say it drops from the 40% average take to 20%. I don't think that would be enough to fix the open web. And so that is why my answer.
- [47:04]
B
Is now is there anything that's enough to fix the open web? Whatever fixing really means. I mean, I'm putting it in air quotes.
- [47:13]
A
I don't think you'd have to fix all four of those things, but I think you'd need to make headway on at least a couple. And I think that is possible through collective effort creating more standards. I just think that some of these problems, especially related to ad formats and some gaps in the design of the marketplace, need more attention.
- [47:33]
B
Well, let's play alternate history. This is my last question because we're nearly out of time. Doc Brown pulls up in the DeLorean, offers you the chance to travel back to a specific moment in time. You can make one strategic change that would alter the online advertising industry for the better. Where are you going to travel to and what would you change?
- [47:56]
A
All right, I'm assuming I'm omnipotent in this scenario, so I could change anything at this point in time. I think I traveled back in time to the Google IPO in 2004 and I would somehow convince Google's leadership that they should definitely buy DoubleClick in a couple of years, but when they do, they shouldn't isolate the search ads business from the rest of the Google Ads business, and it should all be folded into one giant open federated ecosystem. The reason I would make that change is I think if you did that, then you could probably have prevented the fragmentation that we're all dealing with today. There are a lot of advantages to walled gardens. I've talked a lot about them. But the great disadvantage that we're still living with is that advertiser buys remain fragmented. There's no reach and frequency control. If you're buying on a Google and a meta and a TikTok and a Snap, Right. You just sort of have to buy individually across each of those and hope for the best in terms of overlap and redundancy and so on. So I think that with this one historical intervention, maybe I could prevent that from happening, which ultimately I think would be better for advertisers, because that's an important capability that is still missing in the ads marketplace today.
- [49:10]
B
You could also buy some shares, be rich.
- [49:14]
A
I thought it was about altering the industry for the better, not.
- [49:18]
B
Not my own, not my own fortunes while you're there.
- [49:21]
A
Right, exactly. By the way, could you give me a tranche of shares in exchange for this?