Day 1: A Tale of Two Remedies

This is the first day of coverage of the remedies trial in US v. Google. I’ll be in Virginia writing every day on the courtroom activities.

Day 1: A Tale of Two Remedies

If the liability trial in US et al v. Google LLC (aka “The ad tech trial”) was a dramatic telling of big tech abuse and the poor, suffering journalists whose very livelihood was being snatched from their inky maws, then the remedies phase, which kicked off this morning in balmy Arlington, VA, is more like a very long and boring scrum planning session where the product manager is drunk and the engineering manager is trying to hide their crushing gambling debts.

Or, put another way, our adversarial legal system is not exactly well suited to figuring out a solution to this seriously complex technical and business problem.

The day started with half hour presentations from the plaintiffs (the DOJ plus states) and Google, each summarizing their proposed remedies and the flaws in those proposed by the other parties. I won’t spend too much time on the actual proposed remedies, since I summarized that in yesterday’s Marketecture Newsletter (“What to Watch”).

Whack-a-Mole

Julia Tarver Wood started things out with a Churchill quote (“those who forget history…blah blah”), referring to Google as a “recidivist monopolist” and outlining how Google’s proposed behavioral remedies were necessary, but not sufficient in this case. Behavioral remedies would be “a never-ending game of whack-a-mole”, she claimed, with an amazing illustration of cartoon moles, each with a “G” logo on their heads, and with names like “Last Look” and “UPR.” If this image makes it into the record, I promise to post it ASAP.

In addition to the headline-grabbing “break them up” mantra of the DOJ’s proposed remedies, she highlighted additional behavioral constraints:

• No discriminatory bidding without customer consent (e.g. Google cannot favor one exchange over another)

• No direct bidding (e.g. once AdX is spun, they can’t just let DV360 bid into GAM and skip the mediating step)

• No use of first-party data to tie the buy- and sell-sides.

We’ll come back to these, because I think they are actually more important than the spin-out.

“Google has listened”

Karen Dunn, who is important enough that I’m linking to her Wikipedia instead of LinkedIn (that’s a flex), is the lead for Google out of her brand new firm, Dunn Isaacson Rhee (there’s a story, you can Google it). In her opening she characterized the DOJ remedies as “Radical and Reckless,” and while I always enjoy hating on Google in these instances, I’ve got to admit she was very persuasive.

Her argument was very tight:

• “Google has listened”

• Google is accepting and offering remedies that solve the monopoly allegations

• They can all be implemented within a year, as opposed to DOJ’s which could take up to 14 years

• The spin-out is not supported by law since a) the combination of DoubleClick into Google was not illegal; b) plus a bunch of other reasons.

• The spin-out assets are not just for display, but also video and other stuff which was not part of the market definition (I told you this was going to be a big deal).

She ended her statement with a quote, that I’m really not sure I transcribed correctly because if I did it really isn’t great:

Witness #1: Grant Whitmore, Advance Local

Any hopes that this trial would move swiftly was dashed by the endless interview with Mr. Whitmore. Whitmore is an experienced publisher ad ops guy so the DOJ felt like asking him his opinion about every single aspect of every remedy, to which he gave informed, but largely data-free, opinions.

Q: Would you like AdX spun out?

A: Definitely.

Me: OK, cool, we’ll get right on that.

But where this conversation got interesting is when the DOJ used Whitmore as a foil for bringing into view the obvious flaws in Google’s proposed remedies. While Mrs. Dunn made their proposal seem like it tied up the whole monopoly in a bow, there were enough holes in the bow to drive a truck through (mixed metaphor alert!).

The headline remedy that Google proposed, is for AdX to bid into Prebid. Nice! That will let publishers switch ad servers since they won’t be tied to AdX! Nice! Not so fast. Google only promised to bid “Qualified Open Web Display Inventory” into Prebid. What is that, exactly? It’s open auction only, and display only. So if you still want PMP demand, video demand, programmatic guarantees, etc. that demand is only available through the direct tie between the two monopoly products. Not so great any more.

Other truck-sized holes that came up include first party data usage, bid throttling, latency differences, and a whole category of stuff I’ll call “shenanigans no one has thought of yet, but which we are sure someone smart enough to work at Google could probably come up with if given a proper OKR.”

Google’s cross-examination was mostly around the question of whether alternative auction types, like “private deals” were best described as “direct” or “indirect.” This semantic argument was rooted in the idea that Google’s monopoly was only described as part of the “indirect” world of programmatic, so the remedy should be limited as such.

Google’s counsel Bill Isaacson (as an aside, all of the lawyers in this case need to update their headshots) also pushed the point that AdX demand into Prebid would eliminate the possibility of auction manipulation since Prebid is trustworthy and open source. He made this point with a crude diagram showing AdX with an arrow going into Prebid, which then went into DFP. “How can you manipulate the auction when its going through Prebid” he more or less argued.

On redirect, however, Mrs. Tarver Wood took that same diagram plus a ballpoint pen and just started drawing boxes and arrows all over it, like Perry Mason with a whiteboard.

“AdWords bids into DFP, doesn’t it?” <Draws Box>

“DV360 bids into DFP, doesn’t it?” <Draws Box>

“PMPs bid into DFP, don’t they?” <Draws Box>

“Let’s not forget Amazon TAM” <Draws Box>

The dramatic display dispelled any idea that AdX bidding into Prebid was the solution to all the problems in DFP transparency.

Witness #2: Andrew Casale

Casale’s testimony was notable for his assertion that there could only be a fair playing field if AdX and DFP were owned by different companies. This framing is different from the DOJs, where they want AdX spun first, and also different from Google’s, where they don’t want a spin at all. Casale reiterated a lot of points we have already heard, about the advantages that accrue to owning both the exchange and the ad server. He also had a great quote about the limitations of behavioral remedies:

The tea, though, was the discussion around making a bid for AdX. Casale was asked if he was interested in making a bid for AdX and innocuously said he would. Fine, nothing much there. But on cross, Mrs. Dunn brought out some deets that had me wondering if there was more to the story. Dunn got Casale to admit he had discussed making a bid with the DOJ, and recently. She also made it seem like Casale has changed his opinion about the merits of a spin-out in the same timespan that these discussions may had happened. She then pulled out quotes from his deposition and other public statements about how great it would be if the exchange market consolidated into “three to five players.” She painted the picture of a power-mad Canadian rubbing his maple-sticky hands together in anticipation of total advertising dominance. Plus also, AI and curation, but that’s dull in comparison.

Honorable mentions

• Beet.tv got a shout out

• OpenAds.ai got a shout out

• “Would Applovin be interested in buying Adx?”

• Judge Brinkema: “Explain it to me like I’m your favorite 15-year old relative, a smart nephew.”