Trial Update, September 19: In relentless pursuit of power and profit, Google leadership directives compromised even well-meaning efforts to serve interests of publisher customers

US v. Google day 9 kicked off with some lawyer-to-lawyer drama. Google shared their witness list with DOJ, where 6 out of 19 relate to government advertising. Some are government advertisers while others are Google account managers on the government accounts. Reminder that Google paid $2.3M to settle government damages claim, which is why this is no longer a jury trial.  

DOJ’s Ms. Wood flagged this to the court, along with Google’s prior argument that government advertisers are not representative of all advertisers, along with the fact that none of the advertisers, during their depositions, have actually used Google’s tools. They use agencies, and we’ve heard cross-examination by Google of those agency witnesses (Joshua Lowcock re: USPS, and Luke Lambert re: US Army). Judge Brinkema says 1, maybe 2 can testify.

Ms. Dunn pushes for 3. Unclear if that’s 3 advertisers or 3 of the total 6.  If she's holding out for the latter, she’ll be sorely disappointed as Judge Brinkema makes  it clear again that she’s done with cumulative (redundant) testimony, and will be more “draconian” moving forward.

Dr. Timothy Simcoe, Econometrics and Industrial Organization Expert (Part II)

We then get into the cross-exam of Dr. Timothy Simcoe, DOJ expert witness. Karen Dunn spends two hours attempting to poke holes in his testimony and methodology, often using weird math to draw faulty comparison, getting her wrist slapped a few times for out-of-line comments where she tried to signal to the Court the “importance” of a particular issue or question.

Jonathan Bellack former Director of Product Management (Live)

We finally get to Jonathan Bellack, long-time Google PM that came over with the acquisition of DoubleClick. No fireworks, but a more complicated story emerges: one where the product team wanted to do right by publishers, but leadership only cared about the bottom line. It’s reminiscent of what we heard from Eisar Lipkovitz. At one point, there’s an email about Jedi++ that mentions that the goal with Exchange Bidding was to roll out a product “slightly better” than Header Bidding. The email also says that the industry doesn’t seem to care about latency. By this point, HB innovation had continued, so EB would have been “inferior.”

Bellack explains that his goal was to make a product that’s as much “better” as possible than Header Bidding, but leadership’s directive was to make it only “slightly” better. DOJ asks if that’s because AdX is so profitable, and Bellack says that DOJ would need to ask leadership about their rationale.  Bellack also states that he considered HB a threat in the sense that he had heard from publishers including Amazon that they didn’t trust Google and found its structure “unacceptable.” We see an email where he had proposed the creation of a “neutral” ad exchange organization.

He is the first Xoogler that did not seem overly coached, and this sense seems validated by Dunn’s comments that she wants to consider during lunch recess whether Google will ask him any questions on cross-exam. I wrote earlier that if I were a betting gal, I’d say they won’t. I would have won that bet.

Next, we hear a series of read-ins, primarily to get minor points on record and evidence entered.

Google/Xoogle:

1. George Levitte, Director of Product Management: The most substantive one. There was a document we didn’t see and is not entered into evidence but that he spoke to, which he says was a pitch for a job promotion. In it, he talked about how he led the strategic effort to launch a HB alternative that is favorable to Google to counter the HB long-term threat. He explains that EB is favorable to Google in terms of fees and data. He also said it turned competitors into customers. Another doc we didn’t see mentions “secret sauce” - data that is reserved for Google’s buy-side, like the cross-device graph
2. Woojin Kim, VP of Product Management: Reference to a Media Review document that shows that “non-agency, sub-scale” advertisers are only the target of Google Ads, not DV360. The significance is to support the unique demand point, and the reason that DV360 is not a substitute for Google Ads. Google’s side marked for inclusion the part where she speaks to why Google Ads bid exclusively on AdX before AWbid, where the “benefits” were stated to be avoiding self-competition and spammy inventory and traffic
3. Sam Cox, former Product Manager for Authorized Buyers, EB, OB, AdX:  Explaining the difference between instream and outstream video
4. Bonita Stewart, former VP of Global Partnerships: They literally just read-in her stating her title, and that she never turned off default chat deletion

Market Participants:  Deposition Read-ins

1. Bo Bradbury, SVP at GDS&M (US Air Force client lead):  More market definition stuff on channel selection and how they are not interchangeable, as well as a bit of discussion of buying tool use (e.g. DV360 vs. Google Ads) by the Air Force/advertisers
2. Michael Shaughnessy, COO at Kargo, formerly VP, Revenue at Bauer Excel Media: He explains the bigger picture around Google’s market power, referencing Search and Google Analytics, and why this matters. He says adtech is important because it facilitates the value exchange between consumers and content, and is the backbone of fact-based information online, which is key to democracy. Amen.

Robin S. Lee, Professor of Economics at Harvard

DOJ’s final expert, Dr. Lee, takes the stand.  He was asked to analyze (1) whether the markets defined by the plaintiffs are appropriate, relevant markets, (2) the extent of Google’s market power, (3) whether Google’s conduct created/sustained  market power, harmed competition, and harmed customers.  He goes through each market explaining his findings and analysis. Of course, he did find that the ad server, ad exchange, and network are each appropriate, relevant markets where Google has substantial and sustained market power, protected by significant barriers to entry.

The one I was most interested in is the ad network monopolization claim that we’ve heard the least about explicitly. He began by explaining how Google’s market power in the ad network market - i.e. the millions of primarily smaller advertisers forming the basis of its unique demand - originated with Search. He then explained the direct evidence he found, where he used Google’s own experiment data to show that they could profitably increase Google Ads fees by +7%. He also explains that Google was able to reduce “quality” by restricting access to inventory, where quality for advertisers is defined as ROI.  The rest of the analysis focused on publisher.  For example, how Google Ads won 28% of auctions, and removing Google Ads from the auction would reduce bids by 14%.  Analyzing Google’s advertiser ad network market power from the perspective of publisher impact is an interesting approach. Here’s his expert report where most of this comes from for those that are interested.

After an undoubtedly lengthy and circular cross-examination ends tomorrow, the DOJ will rest its case-in-chief, and Google will call their first witness.

‍