Google rigged ad auctions and inflated costs to boost revenue, hurting advertisers, the Justice Department argued last week in the US antitrust lawsuit against Google.
What follows is a summary and some slides from the DOJ’s closing platform, specific to search advertising, that support the DOJ’s argument.
Google’s monopoly power
This was defined by the DOJ as “the power to control prices or exclude competition.” Also, monopolists don’t have to take into account rivals’ ad prices, which testimony and internal documents showed Google doesn’t.
To make its case, the DOJ showed quotes from several Googlers discussing raising ad prices to boost the company’s revenue.
Dr Adam Juda said Google tried to find “better prices or fairer prices, where these new prices are higher than the old ones”.
Dr. Hal Varian indicated that Google had many levers it could use to change the design of the ad auction to achieve the desired result.
Juda and Jerry Dischler confirmed it. Dischler was quoted discussing the impact of price increases of 5% to 15% in these two slides:
Other deck slides the DOJ used to make its case:
CPCs for Google Search ads more than doubled between 2013 and 2020.
Advertiser Damages
Google has the power to raise prices whenever it wants, according to the DOJ. Google called this “tuning” in internal documents. The DOJ called it “manipulation.”
According to the DOJ, format pricing, crushing, and RGSP are three things that hurt advertisers:
Price format
“Advertisers never pay more than their maximum bid,” according to Google.
Yes, but: What Google didn’t mention is “Project Momiji,” which launched very quietly in 2017.
What is Momiji: He artificially inflated the offer made by the runner-up.
The result: A 15% increase for the “winning” advertiser. More advertising revenue for Google.
relevant slides: From the DOJ platform:
Crush
How it worked: Google increased an advertiser’s lifetime value based on how far their predicted click-through rate (pCTR) was from the highest pCTR. According to a 2017 paper touting a new product called “Kumamon,” Google had been doing this using “a simple algorithm consisting of bid, three quality signals, and some (mostly) hand-tuned parameters.” (A screenshot of that document appeared to indicate that Kumamon would be adding more machine learning signals to the auction.)
In other words: Google raised “the price to the highest bidder.”
Google’s goal: To create a “broader price increase”.
The result: The winner of the Google ad auction paid more than he would have if the crush wasn’t part of the ad auction.
And: The DOJ stated that all of this led to a “negative user experience” as Google ranked ads “suboptimally in exchange for more revenue.”
Relevant slides: From the DOJ platform:
RGSP
What is: The second generalized randomized price was introduced in 2019. (You deeper What is RGSP? Google’s Random Second Price Ad Auctions Explained)
How it worked: Google referred to it as the ability to “increase prices (shift the curve up or make it steeper at the top end) in small increments over time (aka ‘inflation’). No led to better quality, according to 2019 Google emails.
How Google talked about it: “A better pricing button than format pricing.”
The result: It incentivized advertisers to bid higher. Google increased revenue by 10%.
Relevant slides: From the DOJ platform:
Search query reports
The lack of visibility of queries also hurts advertisers, according to the DOJ. Google makes it almost impossible for search marketers to “identify non-matching queries” with negative keywords.
Presentation of the DOJ. You can view all 143 DOJ slides: Conclusion: Search Advertising: United States and Plaintiff States v. Google LLC (PDF)
Add Search Engine Land to your Google News feed.
New in Search Engine Land
About the author
Danny Goodwin has been the editor-in-chief of Search Engine Land & Search Marketing Expo – SMX since 2022. He joined Search Engine Land in 2022 as a senior editor. In addition to reporting on the latest search marketing news, he manages Search Engine Land’s Subject Matter Expert (SME) program. Also helps schedule US SMX events.
Goodwin has been editing and writing about the latest developments and trends in search and digital marketing since 2007. He was previously executive editor of Search Engine Journal (2017-2022), editor-in-chief of Momentology (2014-2016), and editor from Search. Engine Watch (from 2007 to 2014). He has spoken at numerous major search conferences and virtual events, and has brought his expertise to bear in a wide range of publications and podcasts.