Following a four-and-a-half-hour cross-examination of economist Robin S. Lee, the DOJ rested its case-in-chief Friday September 20. Friday, we also got to see what Google has in store for its defense, which continues in earnest this week.

If Google’s first witness serves as any indication, it’ll be a long week of the company’s textbook weaponized complexity, jazz-hands, and word games. So, some of us in the courtroom made our own word games to help pass the time: I’ve created more Buzzword Bingo cards for the week for those that are in EDVA. But why should we get all the fun?

Click here to download our US v. Google II crossword puzzle.

Now, as week three of this monumental trail begins, here’s a recap from inside the courtroom of week two along with predictions on what we can expect this week.

---

Revelations and receipts from week 2

More detail can be found in my daily updates, but here are some key moments from last week, with receipts (i.e. Trial Exhibits).

1. Google thought Yield Manager tech was “irrelevant.” Yet they paid $100M over valuation to “park it somewhere.”

While some have pointed out that Google integrated the Yield Management capabilities, the discussion at trial was focused on the RTB capabilities Google chose to deprecate to perpetuate AdX’s advantage via Dynamic Allocation. (See: PTX0085, PTX0112, PTX0060, PTX0058).

2. The IAB Tech Lab declined to take in Prebid.js because Google, its largest funder at the time, “vehemently objected.”

For more on this, have a listen to the Marketecture podcast featuring Brian O’Kelley, and check out AdExchanger’s coverage. It would have been nice to be able to point to a response here from IAB Tech Lab, but it doesn’t seem like that is going to happen.

3. The sell-side product team wanted to make Exchange Bidding much better than Header Bidding to legitimately serve the interests of publisher customers. Leadership told them to aim lower.

Jonathan Bellack’s testimony shed more light on internal dynamics, reminiscent of Eisar Lipkovitz’ comments that “the machine won.” Bellack wanted EB to be a superior product that wins over publishers, but leadership said “slightly better” is good enough.

What is the incentive to invest in innovation that serves the interests of customers when they have nowhere else to go? The market power is further illustrated by a 2014 experime which revealed that Google could profitably raise GDN prices by +7% for buyers to hit a higher margin target without a care in the world.

What to expect during week 3

Google’s first witness was sell-side sales VP, Scott Sheffer, who has been with the company for 18 years and counting. They began with a charade that came to be known as the adtech Spaghetti Football, where Sheffer walked through a clearly-rehearsed exercise of mapping out Google’s view of the adtech ecosystem. Naturally, Google is everywhere.

Those in the courtroom seriously questioned the strategy behind the use of this visual because it so clearly illustrates that Google’s control over the adtech ecosystem extends well beyond even the scope of this trial. Google has been known to weaponize complexity. The less people feel they understand, the more susceptible they become to Google’s narrative filling in the blanks. However, if their big plan is to confuse Judge Brinkema into siding with them, they may want to consider that strategy.

This approach further backfired when the DOJ untangled the graphic in order to leave only the products on-screen that allow publishers to indirectly sell display ad inventory on their websites, and enable advertisers to buy it.

Indeed, Google’s approach at this point appeared to be off to a very rocky start. As for where Google may be headed, the company faces some significant roadblocks:

1. Before they started their case-in-chief, Judge Brinkema already made it clear that there’s no way Google wpi;d be allowed to call six witnesses related to government advertising. “One, maybe two,” she said. Three of the planned six were advertisers, and three were Google account managers on government business.
2. If Google had been planning to call a lot of their long-time employees, they’re likely reconsidering that after Judge Brinkema suggested much of Sheffer’s testimony was “tainted” and “highly questionable” given Google had already been on notice about litigation during the events he attempted to recount. The DOJ has done a great job of ensuring the spoliation issues remain a constant rain cloud over the heads of their opponents and their witnesses. Nevertheless, I’m sure we’ll hear from at least one engineer about how daunting it would have been to enable unrestricted real-time bidding on third party exchanges, or to allow publishers to get AdX real-time pricing in ad servers other than DFP.
3. Google’s full witness list also included a lot of “Grow with Google” businesses and small publishers, and Sheffer’s testimony leaned heavily into Google’s purported goal to foster the creator economy. If they’re planning on calling more than one or two of each, they are likely in for another rude awakening.
4. We can expect that Google will over-index on expert witnesses. The core of their defense is to play games with market definition and to reframe tying and exclusionary conduct as legitimate refusals to deal. We already have a preview of we’ll hear (at length) from Mark Israel.
5. We can expect that Google will rely significantly on deposition read-ins, where they don’t have to worry about unexpected cross-examination, and can cherry-pick what they feel they need to be part of the official record, with a keen eye on the inevitable appeal.

It’s slim pickings on what Judge Brinkema will allow, much less find compelling, so it makes sense that they expect to wrap up their case-in-chief by Thursday. I wouldn’t be surprised if it’s sooner, now that they’ve been told that testimony by conflicted loyalists will not be persuasive.

Google’s games and tricks

One of my observations throughout Google’s defense is their affinity for jumping from side to side as it suits their argument. They bob-and-weave around the DOJ’s allegations by speaking to alleged advertiser benefits of conduct that harmed publishers, and publisher benefits of conduct that harmed rivals. The most frustrating manifestation of this is via their argument that the DOJ’s markets are overly-narrow and gerrymandered. Google loves to try to land the point that certain channels can be interchangeable to advertisers, and that they do face competition in the single two-sided market they insist adtech is. This completely evades the DOJ’s core point.

The DOJ has focused heavily on publishers as the key victims of Google’s conduct, with a specific focus on the majority of publishers that have both direct sold and indirectly sold display ad inventory on their websites, and thus need an ad server. No amount of alleged advertiser alternatives addresses the stronghold Google has publishers in. Only when pressed does Google point to a swath of (impractical and incomplete) alternatives available to publishers (e.g. running prebid.js without an ad server, or using two ad servers, where one of them is still GAM).

It often makes this trial feel like watching chickens and ducks try to have a conversation, and Google revels in this game of rhetorical fallacies. Ultimately, however, dancing around the core issues and harm to publishers only makes Google’s alleged stated ambition to fund the world’s information appear that much more disingenuous.

---

earlier coverage:

A recap of week one of US v. Google II