Jun 02, 2025
On May 21, The Capitol Forum held a conference call with Cristian Santesteban, Eamon Kelly, and Josh Gray to provide context and discuss takeaways from the federal ad tech case against Google. The full transcript, which has been modified slightly for accuracy, can be found below.
TEDDY DOWNEY: Good morning, everyone. Welcome to our Conference Call on the federal ad tech case against Google. I’m Teddy Downey, Executive Editor here at The Capitol Forum. And joining us today are Josh Gray, Eamon Kelly, and Cristian Santesteban.
Josh Gray is a partner at Sperling Kenny Nachwalter. His insights into antitrust law and federal agency policy stem from two stints at the FTC and two decades in private practice handling government investigations and antitrust litigation.
Eamon Kelly is also a partner at Sperling Kenny. He has helped lead antitrust class actions against Apple and Google on behalf of app developers, earning recognition from the American Antitrust Institute and securing favorable settlements in both cases.
Cristian Santesteban is the CEO and co-founder of ArrowLex AI and CEO and founder of RedPeak Economics Consulting. In his career, he has worked on some of the most important mergers and antitrust cases in the last two decades, including ones involving AT&T, Intel, and Comcast. And thank you for joining us today. This is an exciting call for me.
ALL: Thank you for having us.
TEDDY DOWNEY: Before we get started, if you have questions, please put them in the GoToWebinar control panel. We’ll try to get to them toward the latter half of the call. Or you can email us at editorial@thecapitolforum.com. We’ll try to get to your questions.
Josh, if you could open up the conversation for us talking about the Brinkema decision, the opinion, and that would be a great way to start, I think.
JOSH GRAY: Great. I just want to start with a disclosure that, a viewpoint disclosure, these are our personal views, not the views of any of our firms or our clients, and that all three of us have consulted with clients on issues related to ad tech and other tech platform antitrust issues.
So, the Brinkema decision is an absolutely huge win for the Department of Justice, and they deserve heartfelt congratulations. They did it in a quick timeframe, and it is a really impressive showing by the DOJ. But it’s also a legal and moral victory for publishers, who have been concerned about programmatic advertising for a long time, often not fully understanding the technical context in which they’re operating because a lot of information was withheld from them.
If you look throughout the opinion, the harm to publishers is a recurrent theme. It’s really a unifying theme of Judge Brinkema’s analysis. She makes some interesting points about advertisers as well, but I think that this decision shifts the ad tech discussion squarely to publisher harm in an important way.
Any trial court opinion in a significant antitrust case—and this is very much a significant antitrust case—starts with the trial court’s role as fact finder. This opinion does an admirable job with fact finding. And that’s a very important attribute of any good trial court antitrust decision in a complex case. It’s particularly true here because of the difficulty and complexity of the industry.
Also, the case started out with a complaint that is probably the most complex I have read. If you’re not in the business, if you’re an outsider who doesn’t think about ad tech all the time, it was hard to understand. I don’t know why DOJ took that approach. I suspect that they were concerned about the sort of the black box nature of the programmatic ad chain. The complaint was an opportunity to tell the world what their investigation had found, and was successful in that respect.
But if you’re a lawyer who’s following this from outside, it was a difficult complaint to get your arms around. Brinkema has advanced the fact finding. The story, in her opinion, is clear and concise. And she’s pulled the essence out of that complaint, and presented it in a way that is going to be extraordinarily useful to a court of appeals if the case ends up there, as Google has promised that it will.
You look for a great trial court judge in an important case to do many things. The first thing you want her to do well is the fact finding. This opinion is notably successful in that respect in a very difficult context. So, kudos to her for that.
We’ll start with the product markets. DOJ won on two. She found that publisher ad server for open web display advertising constitute a distinct relevant product market. And she did a number of interesting things to support that finding, including noting that Google was in fact able to degrade the features of DFP and the publishers weren’t happy with it, but they couldn’t switch.
And she also found a relevant market for ad exchanges for open web display advertising. And there she put a lot of emphasis on the 20 percent take rate, which was sustained over a long period of time and is twice as high as the rate that she found is common among third-party exchanges. That gap is really noticeable because it’s so large. It seems to have great significance. And you have to put it in a context where Google’s conduct is also inflicting costs on the third-party exchanges because they are constantly entering auctions that are rigged against them. They’re difficult to win. And you only get paid if you win an auction.
So, that means that you’re having to incur the cost of running a business, but the playing field is tilted against you. And your success is rarer than it should be in a world where there is a fair competition. And as a result, your costs go up, and costs and prices are linked in most markets most of the time. So, it’s possible that the third-party take rate would be substantially lower in a world where Google wasn’t behaving the way it did.
She agreed with DOJ and found a worldwide market. And she rejected some Google arguments that would have been very problematic, including the fact that you should just sort of look at the whole ad tech stack as one big market, which I think would have produced an analysis paralysis.
This was a difficult product market definition exercise. If it had appeared on a law exam, the students would have had reason to complain that it was too hard, that the question should have concerned search. Search is also difficult, but comparatively easy. Against that background, she really did a solid analysis and it fulfills the two purposes of market definition.
One is to say, is there potential for monopoly power at stake here? That’s something that we should be concerned about. So, we go into a deeper antitrust analysis. And the other one a table setting exercise to get into the analysis of anti-competitive effects. And so, you frame the analysis of anti-competitive effects with an understanding of what the relevant product market is.
If you judge what she did by those two fundamental purposes of a product market definition in a rule of reason case or a complex Section 2 case, what she did was workmanlike and quite successful.
She found liability for the tie of AdX, the ad server, with DFP. And she found that the tie violated the Sherman Act under both Section 1 and Section 2. And she mixed elements of those two analyses throughout.
If you prefer the Section 1 approach, you’re going to be happy with what she found. If you’re like Herbert Hovenkamp, and you would have liked to see the case decided under Section 2 alone, you look the pieces that are in albeit different parts of the opinion, it’s all there. You can find what might be deemed a Microsoft type analysis within what she did.
Depending on what the judge’s preferences are for how to think about these legal issues, she’s given you the information and the findings that you want. On the Section 1, there’s an agreement requirement and she found that the government presented evidence of agreement. She also found coercion in the policy and technological restrictions that Google placed within AdX.
For Section 2, she did really an admirable job of—well, this is actually really in both, but it’s particularly important for Section 2. She did an admirable job of the ultimate competitive effects analysis. If you’re interested, I would look at page 93 of the slip opinion, where you find the first full citation to Lorraine Journal, talking about Google’s consistent practice of preferencing its own demand.
If she had just found this tie, it would be a huge win for the DOJ. She went further, and she found conduct that she collectively discussed as entrenchment. Following circuit precedent, she considered it sort of altogether as—I don’t know if “course of conduct” is the right phrase—but as interrelated pieces that fit together in a coherent, albeit anti-competitive, way.
She looked at First Look, which was Google’s practice of favoring AdX with a first look at an auction opportunity. And that switched and was replaced after header bidding with Last Look, sell side dynamic revenue sharing, and then later the uniform pricing rule. She found liability for those things because they reinforced the anti-competitive effect of the tie. For the uniform pricing rule—there’s one really remarkable finding When, under antitrust pressure, they finally gave up on the Last Look, but didn’t want to have unconstrained auctions in which third-party ad exchanges might compete on something that looks like a level playing field. So, they prevented publishers from disfavoring AdX, but you were still allowed to disfavor—according to a really remarkable finding—you were still allowed to disfavor third-party exchanges.
That brings us to how do these findings in the government case—what do they tell us about harms and damages to publishers in private cases? Which is something that Eamon’s been thinking about and I pass the microphone to him.
EAMON KELLY: Great. Thanks for that good summary, Josh. So, the biggest takeaway is that her ruling for private plaintiffs gives a giant boost to the publisher cases. It’s interesting this is a case where the federal government’s case came after the state and private plaintiffs have brought their case, but gets to the decision first, and ends up being a big bonus for both a group of class plaintiffs and some individual plaintiffs.
If you’re a publisher, thinking from the perspective of a publisher, the question is what do I do now? There’s a great finding.
The good news is there is a class action that they just filed for class certification earlier this month. They did it with the wind at their backs—or the wind in their sails—from the judge’s ruling.
Another good thing is the class action is helmed by a trio of top-notch class lawyers, Korein Tillery, Berger Montague, and Boies Schiller, and they have done a great job of litigating the case so far.
The interesting thing, though, is there’s a variety of different harms that are here. And the class case is necessarily focused on one of those harms, which is what’s called the inflated take rate.
Google takes generally 20 percent of the ad auction revenue. And the class case is focused predominantly on the excess take rate that they have, the supercompetitive take rate, when you compare it to some of their competitors. And so, if you’re a publisher thinking of what to do, or you’re somebody advising a publisher on what to do, the question you have to ask yourself is that take rate going to be a sufficient recovery for me? And you’re going to want to think about a couple things.
One thing you’re going to want to think about is just your raw size. If you’re a smaller publisher and most of your revenue is programmatic going through ad exchange, the class is probably going to do a pretty good job of protecting your interest.
If you’re a publisher that has a lot of premium content, has a lot of direct sales, has experienced a substantial decline over the last period of time, since the mid-2010 to 2020 period, then you’re going to want to think do I have unique harms? Because it’s obvious from the ruling the harm is not just limited to the extra take rate, but there are things that have been impacting directly the price, the auction price.
If you’re a publisher that has seen a decrease in your revenue, you’ve got a lot of auction revenue, you have premium content, you’re going to want to think clearly is the class case going to do enough to protect my interest? Not because they’re not doing a good job. They are. But they’re necessarily going to be focused on the vast majority of publishers who are harmed primarily by the take rate.
If you find yourself in a bucket where you’ve got a different set of harms, that’s when you’re going to want to think about whether or not you should ultimately opt out and pursue your own claim. You’re not on the clock yet, in the sense from the class certification. The motion for class certification just got filed. Although, if you want to join in the current case in the Southern District of New York, that’s a decision you’ll have to be taking soon.
Those are the things, Josh, to answer your question, that people, that publishers and people advising publishers, are going to need to think about. I will say, obviously, when one of the things that happens in private actions is pursuit of remedies. But we’re going to get more information about remedies soon from the government case before anything happens in the private cases.
So, Cristian is going to take a moment now. Cristian, if you could just talk through what’s at stake in the remedies phase of the DOJ’s case.
CRISTIAN SANTESTEBAN: Well, the remedies implications are pretty huge and there’s two wildly different proposals on the table from DOJ and from Google.
I’ll start with the government’s proposals first. The government’s proposing a mix of structural remedies that will involve the divestitures and quite extensive conduct remedies. I think those have been downplayed a bit, but they are quite wide ranging.
In terms of the structural remedies, number one, the first thing you need to say off the bat is the government’s proposing to divest AdX, the ad exchange that Google owns. And that’s in the center of the vertical stack, crucial to direct buyer demand and to publishers. So, it’s kind of the matchmaker. And Google’s control of that is fundamental to the judge’s finding of monopolization. So, that is the number one proposal from the government, just divest AdX.
Next, the government doesn’t think that’s enough. So, there would continue to be incentive to self-preference, even in the case that AdX changes hands from a legal perspective. So, the government is proposing a divestiture of the publisher ad server, what has been traditionally called DFP, from the legacy of DoubleClick.
The DFP divestiture they are proposing, at first read, is a rather complicated three phase divestiture. And so, that’s just to allow for the complexity of the technological aspect of this case. And so, the first phase has a goal of making it cheap to multi-home for publishers by allowing them to migrate their data out of the DFP.
They’re requiring, in this first phase, the creation of an API to allow publishers to transfer the data out of DFP to some other ad server. And that is in addition to providing another API so that you can open up the DFP to the Pre-Bid header bidding system.
Right now, AdX is advantaged in the DFP. So, in order to level the playing field somewhat, immediately in this first phase, they would force Google to allow the Pre-Bid open source header bidding system that has a real‑time auction among exchanges to provide their bid on an equal footing to the AdX bid through an API access.
In summary, the first phase, two APIs, one to get the data out of DFP and one to allow header bidding to provide real-time biddings in parallel with AdX.
Then the second phase is, for me, a little bit more opaque because you need to get into the details of code and things that are very technical. And so, this involves licensing the part of the DFP software code that does — what the DOJ is calling — the final auction logic. So, basically once you have an impression, how do you select what ad actually gets put on the website of the publisher? And that auction logic code is where Google has a lot of opportunity to self-preference.
The auction logic code in phase two would be licensed through an open source license and it would be run by a third-party organization that would host what they call the final auction to basically place the ad on the website. And that would be out of the control of Google by that point in phase two. So, the final call to decide which ad gets put on the website, that’s taken out of Google’s hands.
From an economics perspective, that basically says you turn this very strategic code into a public good to eliminate information asymmetries and prevent Google from recreating any of the clandestine discriminatory rules. So, it’s a very important part of the DOJ’s logic behind the remedies. This opening up of the auction logic code.
And finally, once you have done that, you pretty much have de-toothed DFP’s ability to discriminate and self‑preference. But at that point, the third phase involves a full-on divestiture of DFP, even not just to a single buyer, but to potentially multiple buyers. So, you’d be replicating, doing clones, of this more neutral DFP.
The idea is you split the natural monopoly that currently exists due to the network effects before it tips again into a potential monopoly, before it solidifies again. So, it creates multiple buyers to create competition. But even a single buyer of DFP would mean structural separation from the major exchanges and eliminate the conflict of interest that currently exists there.
Already that’s pretty extensive. But in addition to that, they’ve gone onto impose conduct remedies while these divestitures are going on. And I won’t mention all of them, but some of the highlights are that the remedies are going beyond the narrow call of the judge’s anti-competitive conduct in that they’re imposing some remedies on the advertiser side.
One of the conduct remedies is for the buying tools to work well with all the exchanges, not just AdX. So, there should be no preferential routing of the buy-side advertiser demand to say AdX. And the conduct remedies are asking to provide a neutral API to access AdX. So, Google AdWords would no longer have preferential access to AdX.
Some of the other conduct remedies, no tying of products, of course, no tying. And enjoining any conduct that has been deemed anti-competitive by the court. And last but not least, they are imposing a conduct remedy of just data sharing, like significant data sharing.
In addition to the code that we just talked about, I just talked about, DFP has all these relevant signals for the ad buying tools, especially one of the sort of not well-known ad buying tools advantages — or maybe well-known by industry participants — is that Google AdWords and all the Google buying tools, they get this instant feedback on conversion from DFP, from the publisher ad server.
These data signals on what clicks happened and what ads converted well, all that gets instantly shared with Google’s ad tools, ad buying tools. And so, those data signals, the DOJ conduct remedies say should be shared equally to all ad buying tools, not just being preferentially given to Google’s tools.
Those are pretty wide ranging structural and conduct remedies that the DOJ is proposing. In contrast, Google is proposing quite limited remedies and, in fact, no structural remedies. And they’re proposing two conduct remedies.
One is that the real-time bids that AdX provides to DFP right now — so their ad exchange is giving real-time signals, real-time bids, to their own ad server — that should be made available to all publisher ad servers. So, any competing ad servers on the publisher side should be getting AdX’s real-time bids. That’s what Google’s first remedy is.
Of course, Google has 90 percent share of the ad server. So, there are very few ad servers that would be getting these real-time bids. And the market has already tipped. So, it’s not obvious that the status quo would change with that.
And the second thing is the deprecating unified pricing rules. So, deprecating unified pricing rules means that you should be able to impose whatever floor you want to whichever ad exchange if you are a publisher. And that’s something that’s currently prohibited under UPR.
So, that would go away. It would let publishers set different price floors. But, in my opinion, that would still leave Google in control of auction sequencing, latency requirements, data privileges. So, there would still be information asymmetry.
And I think also proposing a three-year compliance trustee to make sure that Google complies with the two conduct remedies that I just outlined, three years seems awfully short, given the timing of market equilibrium settling here and migration to ad servers if there is to be migration. But anyway, that’s in large part the description of the remedies proposed both by the DOJ and Google.
TEDDY DOWNEY: This is a really great dive into the decision and the proposed remedies. We’ve got some questions here. And I think one of the questions from the audience kind of gets at the heart of this incentive for publishers to sue Google directly, which Judge Brinkema focused on the near constant 20 percent take rate by Google’s AdX as evidence of monopoly power. How do you think about damages in regards to that percentage in a healthy market? And isn’t the auction rigging likely much more consequential? You brought this up a little bit, talking about publishers, premium publishers, but would love to sort of get everyone’s thoughts on that.
EAMON KELLY: Yeah, this is Eamon, happy to jump in. We don’t see exactly what the classes damage model is. Our guess is that it’s going to be roughly half of it. So, about 10 percent of the ad revenue is likely what the class is going to seek. There’s some basis upon what you could seek more. We’ll be interested — right now, the class certification papers are redacted in some important aspects. We’ll have to get more information about what exactly their damage model is. Obviously, that’s a substantial amount. It’s 10 percent of the AdX’s revenue, if we’re right about that thesis about what the class is going to do.
And there’s great evidence to support that. And that’s a really common sense model to take if you’re just trying to get every — like a class lawyer, which I am often, you’re seeking damages for everybody. But we do think that some of these manipulations plainly had the effect of decreasing the auction clearing price. And we know, kind of over the last eight years or more, we’ve seen a very substantial decrease in digital advertising revenue and a shift of it more and more towards Google’s system.
We think it’s likely that there are bigger damages than just the take rate for premium publishers and for publishers that have enough revenue. If you’re a smaller publisher, you rely mostly on programmatic. The tradeoff for you in pursuing those bigger damages is going to be something you’re going to have to think about. But the bigger you are as a publisher, the more premium your content is the more compelling it is going to be for you to think about whether or not you should pursue your own claim.
JOSH GRAY: Cristian, do you want to add anything to that?
EAMON KELLY: Teddy, I don’t know if that got it. Josh or Cristian, I don’t know if there’s any follow-up there.
CRISTIAN SANTESTEBAN: Well, just on the specifics of the judge’s ruling with regard to like UPR, Uniform Pricing Rule, there’s potential for some damages there on behalf of some premium publishers who historically set higher floors for their most valuable impressions. That ability to set floors is very important to publishers that have different kinds of types of impressions, some of which are very valuable given their user base.
Their ability to control those price floors, we think that there is a potential floor lift loss that is unique to publishers who rely on these sophisticated price discrimination strategies. And premium publishers are such publishers. So, that would be one direct link to the judge’s decision involving damages.
JOSH GRAY: Another that struck us is the DOJ, at the last status conference, spent a lot of time on the second step of the divestiture that Cristian explained, taking away Google’s control of that decision logic of ultimately which ad to place. And DOJ seemed to reflect a concern that the decision logic was biased and favored, not only programmatic, but Google’s demand within programmatic. And it follows from that, that Google may have been giving premium inventory to itself, and discriminating even around the margin, against direct sales that would be at much higher prices.
And that could have been a very serious market distortion that comes from the monopoly and Google’s pattern of self-preferencing. We don’t have full access to what DOJ knows. From what Julia Tarver said at that hearing, it certainly sounded like DOJ knows there’s a problem there.
And there’s some other things that we’ve been looking at, and it’s hard to exactly handicap them, but they all seem relevant and compelling, at least in concept.
TEDDY DOWNEY: Can I ask a similar question? We’re talking billions of dollars here, right? I mean, what are the dollar amounts that we’re really talking about when it comes to both on the risk to Google from follow-on litigation, direct lawsuits, but also just like kind of opportunity costs for big publicly traded premium publishers? I mean, are we talking about big enough numbers here that investors are going to be going to the companies and say, hey? Why aren’t you bringing a direct action here? This is hundreds of millions, if not billions, of dollars on the table. Are we talking about big enough numbers there? Or am I getting ahead of myself in terms of the size of the liability?
EAMON KELLY: I think my gut is two things on this. One, it’s not a decision about whether or not you pursue your damages or not. Because there’s a class that’s pursuing damages, right? So, I think businesses have to make their own call about whether or not it’s compelling. And there’s no one right answer to whether or not you have to do it.
The damages can be huge. Although, sometimes people that you would think have big damages, really because of the impact of the decreasing price of digital advertising over time, don’t have as big damages as you would think. But I think it’s always a judgment call, particularly when you’ve got good class lawyers.
When you’re starting to see tens of millions to hundreds of millions, the high figures to hundreds of millions in affected commerce, you should really seriously consider whether or not you have a direct claim. But with good class lawyers, it’s always a judgment call. And I don’t think you have to – it’s a business judgment call that you could make a lot of different ways, would be my take on it.
TEDDY DOWNEY: Are the class lawyers going to be able to get into the auction rigging details the way that a direct could, though? That’s the issue, right? Is that they’re just going to leave that out because that’s specific to each firm.
EAMON KELLY: Yeah. So, as best I can tell from reading the class certification papers, they are absolutely not going after the impact on the auction clearing price. They’re just going after the take rate. So, there is absolutely, particularly for larger publishers, a lot of their claim that isn’t being pursued, which is totally reasonable by the class lawyers. It presents an opportunity that I think premium publishers or publishers with big spends really ought to be thinking about. I guess my point is, Teddy, the decision they ultimately make is kind of a business judgment in my mind.
TEDDY DOWNEY: Do they also get any remedy, like another bite at a remedy, if they go direct? Like we’re going to get a remedy in the DOJ’s case besides the damages? Could they ask for anything else specific that would be useful?
JOSH GRAY: So, let me start on that quickly, Teddy, and then pass it. I’ve got something framing to say that I know Cristian and Eamon are both also going to want to weigh in on this.
We’re in an extraordinary situation because we have the MDL private cases. There’s the Texas action, which is seeking injunctive relief. And now we have this fairly remarkable remedy phase that Cristian described in really interesting detail.
Three things are going on, and it’s important not to overlook the Texas case. The other thing that’s going on in the background is Europe has promised a decision on this issue. And they’ve talked about structural remedies, which is very unusual for Europe.
In the past, European remedies have been limited to the community and not global. But we now have a finding from Judge Brinkema that it’s a global market. So, there’s at least a possibility that Europe is going to be thinking about remedies on a global basis.
There’s a lot going on with remedies. There’s always opportunities for – and Eamon should speak to this after. But there’s always opportunities for private plaintiffs to either participate in the remedies phase of a government case or to ask for some incremental relief that will make it more effective to that particular private entity’s business. You know, sort of reinforcing relief. But we’re in a fairly extraordinary situation with all the pending actions right now.
EAMON KELLY: Yeah. So, my thing is, if you want to use a private case, individual action, to achieve a conduct remedy, it probably has to be something that’s pretty specific to you. Because we’re going to get pretty good litigations and rulings that the judges are going to start to look to what happened in the DOJ case? What happened in the Texas case?
So, if you’re asking for a remedy that’s a big industry wide remedy, that affects a lot of people, and it wasn’t given in the DOJ and it wasn’t given in the Texas case, then you’re going to have an uphill battle getting that remedy. The difference is if there’s a way in which Google’s conduct is particularly harming you, or a unique group of users, that’s a good reason to pursue a private path. If there’s something that’s special about you, that’s unique to you, that isn’t getting lumped in.
So, the structural remedies Cristian talked about, I think private cases are tough ways to get those kind of structural remedies. Now, we saw one. I mean, it’s in the news. Epic got a great remedy in the Apple litigation that had industry-wide impact and right today is having industry-wide impact for app developers. But that’s more unusual than it is typical. So, it can happen, but that’s unusual.
JOSH GRAY: Cristian, anything more on that? It’s sort of a lawyer topic.
CRISTIAN SANTESTEBAN: Well, the Epic case was a private case and there was no equivalent government case. Here we have a government case that’s asking for some pretty significant remedies that could impact the industry in a big way. There’s still ways in which individual claimants, large publishers, premium publishers, could ask for things on top of that.
One of the distinct features of premium publishers is that they have a user base that’s very valuable. And one of the allegations that I have heard is that Google is able to monetize that user base in its own properties because it has information from the publishers themselves. And that erodes the value of those users for the publishers and their ability to monetize.
There’s potential for some conduct remedies there in the sense of prohibiting Google from taking potentially private data on the user base of publishers for itself and for its own monetary gain. So, those are damages as well as potential conduct remedies.
JOSH GRAY: So, There’s a really interesting point behind your interesting point, which is the DOJ litigated the search case and the ad tech case on separate paths. And the analysis is within the complaints, and within each of those businesses, in a discrete way. But there’s clearly a relationship between the two in the background. And if it had been brought as one big antitrust case, we might have learned things and there might be findings that are not there because of the separate paths.
One of the mysteries of this is, I think, 75 percent of Google’s revenue was search advertising. Still, I think it’s like 13 percent on ad tech, but those might be old numbers. The order of magnitude for ad tech is a much less important part of Google’s business in search advertising. Yet, Google has done everything that Judge Brinkema described. And clearly, having a presence in programmatic open web advertising was important to Google.
We really, from these opinions, don’t know why it was so important to Google. I can guess that the DFP provided Google with a lot of data. I could guess that if somebody else had that data, maybe it was going to erode some of Google’s market power in search advertising or audience profiling or whatever. But that record hasn’t been developed because of the way this unfolded. So, there’s some missing pieces that are out there.
TEDDY DOWNEY: One other sort of related question here. If you take a step back — and you guys mentioned there’s European cases, there’s state cases, there’s class action, there’s the search case. If you take a step back, is there going to be, at the end of all of this, a competitive market where publishers and competitors will be able to participate in these markets?
And how do you think about how big that opportunity is? Or are you still in a wait and see mode? I’m curious if you’re confident in that this says we’re ultimately going to get enough of a weakening of Google’s power in these markets that you will have actual fair competition. And how do you think about the size of that opportunity?
JOSH GRAY: Cristian, that’s got your name all over it.
CRISTIAN SANTESTEBAN: It all depends on how these remedies that are being proposed play out. I suspect that if just a few of the conduct remedies that are being proposed out there, especially from Google, are implemented and that’s all, I don’t quite see, as I sit here, how that’s going to restore competition given that market has already tipped. People are going to continue to use DFP even if Google decides to give access to its real-time bids to other publishers.
AdX has still the demand flowing from Google AdWords. So, that’s a very valuable source of demand that is tied with AdX. I don’t think the Google remedies, by themselves, are going to restore competition. If instead some version of the government remedies are implemented, especially AdX divestiture, and some form of the DFP divestiture, that has some opportunity to transform the industry and move the industry from its current market structure of a dominant firm that has significant market power to extract 30 to 40 percent of value to a much more competitive regime in which publishers would get much more of what the advertisers are willing to pay. And there’s real hope for that to happen, but only if some significant changes happen from both structural ownership and on the conduct side.
JOSH GRAY: I have a little historical perspective on this. The CMA studied this industry first, I believe. And then the ACCC in Australia issued a fairly extensive report. And both those agencies, and actually Dina Srinivasan in the U.S., who’s sort of a one-woman antitrust agency on this issue, all studied this. And Dina’s article is brilliantly footnoted and sourced. And they all concluded that antitrust had a role to play, but they also thought that some kind of regulatory apparatus, sort of an SEC for open web programmatic advertising, was going to be appropriate too. And I have a bias against regulation, if it can be avoided, because I’m an antitrust lawyer.
One of the things that I have in mind, looking at what’s happened in Judge Brigham’s courtroom in Virginia and what DG Comp may do in Europe, is, if we solve the monopoly power problem, and we separate the buy-side from the sell-side, and we open pathways for more competition, are the things that the CMA and the ACCC, the problems that they concluded required an SEC-type regulator, going to go away? It would be fantastic if, 10 years from now, when people were having this discussion, they say we solved the monopoly power problem, then suddenly these markets started working. If you look at those early reports, there are trading practices that look an awful lot like problematic trading practices in financial markets, and that may require some rules and ongoing supervision. We don’t know yet whether solving the antitrust problems is going to completely resolve the problems in these markets. It’s hard to believe it won’t be a huge step forward no matter what.
EAMON KELLY: It’s likely this is going to result in a more competitive market. But then there’s the question of what happens over time. And I think for us to really get competitive tech markets, we’re going to have to see sustained antitrust enforcement and the remedies are going to have to be significant enough. So, we’re all watching the remedies. But if they’re significant enough, then they may condition conduct going forward. So, I think you need both. They’ll make a difference.
It is important not just to correct the market, but also to send a signal that in the future, if there remains a commitment to antitrust enforcement in the tech sector — and we have had more antitrust enforcement in the tech sector in the last 10 years, in the last five years, than in any period since Reagan — then I think we’ve got a hope of keeping these markets more competitive.
TEDDY DOWNEY: And we’ve got a question here. Where is liability decision most vulnerable and what remedy is most likely? So, we looked at obviously, it’s worse for Google if they go with DOJ’s remedy.
EAMON KELLY: As an enforcer, there’s nothing vulnerable at all.
TEDDY DOWNEY: Getting your thoughts on those questions, I think maybe — and then maybe we’ll get hopefully get to one more after that.
EAMON KELLY: Yeah. Josh, do you want to jump in?
JOSH GRAY: I really don’t like to predict what the courts of appeals are going to do. Judge Brinkema is a very experienced trial court judge, and she clearly knows her circuit well. And I think there’s always a conversation between an experienced and excellent trial court judge like she is and her circuit, which is the Fourth. And my suspicion is that she knows the way that the judges who will be reviewing her decision think about the law, and she knows what their policy preferences are, so she’s crafted that decision to fit within what they will be comfortable with.
That may be one of the reasons, for instance, that she didn’t find liability prospectively for the acquisitions, the Double Click acquisition. And instead, she focused on post-acquisition conduct that was the basis of her liability findings. And it’s quite obvious, and notable, that all that post acquisition conduct would have been impossible if the acquisition hadn’t taken place.
So, it’s almost immaterial whether you find the acquisition illegal under a prospective, hypothetical Section 7 type analysis. All the bad things that the FTC didn’t think were going to happen immediately happened. And with the exception of – I’m forgetting her name, the Commissioner from New York [ed. Pamela Jones Harbour] who wrote that wonderful dissent, who turned out to be completely right.
I have a high-level of confidence that Brinkema has framed that decision in a way that’s going to work for the reviewing judges.
TEDDY DOWNEY: And then do we have any thoughts on, again, I mean, I sort of asked it at a high-level. Do you think a remedy ultimately is going to work? But any thoughts on at least which, you know, obviously, they’ve sided with DOJ so far, the judge. Is it more likely at all? Any guesses here on which side Brinkema will land on for a remedy? I mean, you make a pretty compelling point that if they go with Google’s remedy, there is not really a solution to the problem.
So, ostensibly, she didn’t come up with a liability decision with no interest in a functional remedy here. Any aspects of the DOJ remedy that the judge might like? Right now, there seems to be a heavy interest in structural over behavioral stuff. But they’re sort of thinking that structural is a little bit harder to get appellate courts to go along with. What are you what are you thinking in terms of how to think about what remedy’s most likely?
CRISTIAN SANTESTEBAN: That’s quite a difficult question for me to answer. I don’t know how the ruling is going to come down. In terms of effectiveness of restoring, at least having the potential to restore competition, a remedy, as Google is framing it, that it should be narrowly targeted to the exact conduct mentioned in the liability finding that was anti-competitive and imposing conduct remedies on that, which would translate into the two conduct remedies that Google is proposing, I don’t think that that would restore competition. And that is kind of a backward-looking remedy.
What you want is a forward-looking remedy that looks forward towards restoring competition in the way the markets are actually functioning. And in this case, the judge would be most interested in the effectiveness of the remedy, not symmetry with the liability definition.
Some form of structural remedy, even if it involves something as drastic as eliminating Google’s ownership over two of the three market segments that they own and operate, is probably the only way to restore competition. You can’t just get rid of AdX because then Google would still have the buy side and sell side incentives due to the strong indirect network effects taking place in this market. So, some form of AdX plus DFP opening up are necessary in my mind. Whether that’s going to happen —
EAMON KELLY: I’m going to call my shot. I don’t know which, if it’ll be here or in search. But my gut is sooner or later Google is going to be forced to make some divestitures. I think five years from now — I’ll say I predicted it here — we’re going to see some form of divestiture for Google. And look, my best guess is Google’s probably preparing for that reality. In a weird way, the cases aren’t totally unrelated. I predict five years from now, we’ll have some divestitures. Cristian’s better able to call the shots on which ones than I am.
The other one I would say is, look, I think Brinkema opinion was great. I suspect if I were to put — I think what Google is going to fight the hardest about the liability, to not totally dodge your question earlier, Teddy, is I think they are going to put a lot of energy into pushing back on her conclusion to apply a per se standard to some of the content. So, I think that’s where they’re going to focus a lot of their energy in terms of the liability finding.
JOSH GRAY: Where they’re [Google] going to have trouble is that the DOJ staff that litigated that case did just a brilliant job in undermining Google’s justifications, business justifications, for the tie and for the conduct that collectively she calls entrenchment. Once you have fact findings that your justifications are balderdash, it’s very hard for a defendant to dig itself out.
On the remedies, I want to put a legal gloss on there, which is I think remedies are most difficult in situations where the monopolistic position grew organically. We saw that in the Microsoft case. Microsoft had built that OS monopoly itself. It had not done it through acquisitions. And there’s a lot of reasons to presume strong efficiencies when a firm develops a monopoly organically.
Those assumptions do not apply when a monopolistic position has been put together through a series of acquisitions, whether or not the acquisitions ex ante at the time are illegal. And that’s exactly what you see in Google’s position across the ad tech stack, it is created by acquisitions. That should make it easier to break up. You’re just less concerned that you’re destroying efficiencies when that’s the historical context.
TEDDY DOWNEY: I’ve got one last question here that sort of plays on the same point, which is Google, and I think the other Big Tech companies as well that are being sued for antitrust, are arguing that they are too intertwined and that these divestitures are too difficult to pull off.
You know I’ve been on watching Wall Street for a long time. It seems like divestitures, spinoffs, happen all the time, either as part of mergers or just the regular course of corporate decision making. But I’m curious to get your views on how easy and typical or difficult and atypical it is from a legal and logistical perspective for courts to order big divestitures. And, Josh, you touched on this, but I just want to sort of put it out there. Like, is this typical and easy? Or is it, like the Big Tech companies say, super hard because they’re so intertwined?
JOSH GRAY: I’m a lawyer. I find it very hard to have an opinion on a sophisticated business question like that. My life experience is limited to practicing law.
EAMON KELLY: I mean, look, divestitures are hard. But just think how many willing buyers are there for these business units that would happily pay money to own these units and operate them? That’ll tell you whether or not you can get over those challenges. I’m sure there are people who would pay big money to get these units.
CRISTIAN SANTESTEBAN: Exactly. As Eamon was saying, there’s a market test for this. If these assets are valuable, and the technological cost of transferring the code out of the Google infrastructure into a third-party’s infrastructure is not prohibitive, then there’s going to be bids that are significant for these assets. And if that’s the case, then that’s market proof that whatever hurdles exist are not so high to overcome that a buyer would be willing to pay for it.
TEDDY DOWNEY: All right. Well, this is a very fascinating conversation. I learned a ton. I’m super interested to see how this plays out. And thank you so much for doing this. And thanks to everyone for joining the call today.
CRISTIAN SANTESTEBAN: Thanks for having us.
JOSH GRAY: Thanks for inviting us.
EAMON KELLY: Thanks for having us, Teddy. Good fun.
JOSH GRAY: It’s been fun.
TEDDY DOWNEY: All right. Bye. Thanks again. Bye-bye. This concludes the call.
(END OF TRANSCRIPT)
