What you need to know:

• A newly unredacted lawsuit filed in New York by 16 US states, plus Puerto Rico, paints a detailed picture of Google’s alleged price fixing, backroom deals and misleading statements over the past decade.
• It alleges collusion with Facebook, market manipulation and anticompetitive behaviour at a grand scale that have handed Google hugely inflated cuts of advertisers spend, up to 42 cents on the dollar, while crushing competition.
• One Google employee admitted “[t]he analogy would be if Goldman or Citibank owned the (New York Stock Exchange)”.
• The lawsuit also alleges that deprecation of cookies are a “ruse” to create a closed system out of the open web, an end game it states Google has long intended.
• Google says the lawsuit is “riddled with inaccuracies”.

States pull no punches

Documents unsealed by New York judge on Friday expose explosive evidence and allegations 16 US states are using in a bid to take down Google on antitrust charges.

The documents claim Google manipulated advertising exchange results through bid rigging, cut backroom deals with Facebook to kill header bidding, and designed at least five programs that deliberately undercut competition and harmed publishers’ revenue, while taking up to 42 per cent of money paid for online ads. It’s a clinical exposé which – if proven – shows how effectively advertisers have been fleeced and how publishers were sold lies over and again.

“Having reached its monopoly position, Google now uses its immense market power to extract a very high tax of 22 to 42 per cent of the ad dollars otherwise flowing to the countless online publishers and content producers,” the lawsuit states. “The open internet is now threatened by a single company.”

The 173-page complaint filed by 15 US states, plus Commonwealths of Kentucky and Puerto Rico, in August in a New York District Court was unsealed after judge P. Kevin Castel ruled last week the material should be made public, counter to Google’s wishes.

The lawsuit is a laundry list of alleged actions taken by Google that, it claims, disadvantaged publishers for almost a decade, driving billions in revenue and entrenching Google’s market dominance. It alleges in detail, using subpoenaed internal communications from both Google and Facebook the manoeuvres Google took to “kill” the free market publishers and the broader ad market had tried to build through header bidding – which for a short time effectively freed them from being bound to Google.

It asserts that Google created an alternative to header bidding and then punished publishers that attempted to resist its efforts by throttling their loading speeds and manipulating their search rankings to cripple web traffic.

Google told Bloomberg the lawsuit was “riddled with inaccuracies”.

“In reality, our advertising technologies help websites and apps fund their content, and enable small businesses to reach customers around the world,” per spokesperson.

“There is vigorous competition in online advertising, which has reduced ad tech fees, and expanded options for publishers and advertisers.”

The US states' Attorneys General disagree and the filing pulls no punches in alleging insider trading and market manipulation at a grand scale.

‘Kill header bidding’, Facebook deal

The case details internal Google plans known as “Jedi” to “kill HB [header bidding]”, which allowed publishers to circumvent Google’s exchange and create a more competitive market, generating more revenue by using a piece of JavaScript code that effectively opened up their inventory to the broader market.

While publicly stating Google didn’t “see header bidding as a threat to our business”, privately, employees described it as an “existential threat”.  

In response, Google allegedly signed a secret deal with Facebook in 2018 known as “Jedi Blue” that gave Facebook an advantage in Google’s online auctions. Per the document: “Google and Facebook agreed to manipulate publisher auctions in Facebook’s favor through secret bid, spend, and win commitments.”

In return, Facebook stepped away from header bidding – and committed to spend at least $500m per annum in auctions from year four of the agreement, according to the filings.

The documents state that Facebook actually saw Google coming, recognised the leverage to be gained from being seen to support header bidding, and laid a ruse for Google in order to secure multiyear preferential terms and insider information.

Facebook would otherwise have had to make significant investment to compete. The court documents detail the alleged agreements between Facebook and Google to jointly fight any subsequent antitrust and data privacy actions they may face as a result of collusion. Within that document, the word ‘antitrust’ is mentioned 20 times.

“Any collaboration between two competitors of such magnitude should have set off the loudest alarm bells in terms of antitrust compliance,” the lawsuit states. “Apparently, it did not.”

The documents allege that: “The parties agreed up front on when and how often Facebook would bid in auctions, and when and how often Facebook would ultimately win … Google also provided Facebook with a speed advantage in auctions. Google subjects other marketplaces competing for publishers’ inventory in Open Bidding to 160 millisecond timeouts. Competitors have actively complained that 160ms is not enough time to recognize users in auctions and return bids before they are excluded. By comparison, Google nearly doubled timeouts, extending them to 300 milliseconds, for Facebook.”

Apple turnover

The lawsuit also alleges that Google gave Facebook a sweetener by enabling it to have direct billing and contractual relationships with publishers, an advantage it does not afford to other networks and exchanges. Then Google gave Facebook another “leg up”, the filings claim:

“On top of special pricing, longer timeouts, and a direct billing relationship exception, Google further induced Facebook to help it shut down competition from header bidding by informing Facebook which impressions are likely targeted to spam (e.g., impressions targeted to bots, rather than humans). Facebook does not have to pay for those impressions. Other networks have asked Google for the same information, but Google has refused. So now Facebook has a further leg up over the competition in Google auctions: Facebook knows which impressions sold through Google are fake and worthless.”

Meanwhile, the lawsuit claims Facebook and Google have continued to collaborate to identify users in auctions – and tag team against Apple, which is attempting to compete by claiming to offer users greater privacy.

One penny more

The filings also allege that Google traded ahead of other exchanges, excluding competition from header bidding:

“A publisher like USA Today would route their inventory to multiple exchanges through header bidding, then route the winning exchange bid into their Google ad server. Google programmed its ad server to let its exchange displace the winning header bidding exchange bid by paying one penny more,” per the document.

“Put another way, Google’s ad server let Google’s exchange peak at the winning header bidding exchange’s bid, then displace the trade. Industry participants called this Google’s “Last Look”. Other industries call analogous conduct by intermediaries “insider trading” and “front running.”

“With Last Look, and Google’s absolute monopoly in the ad server market, Google successfully foreclosed competition in the exchange market and ensured a system where Google always prevailed. Google’s exchange cherry picked the best impressions, leaving rival exchanges the low value impressions left behind by Google’s exchange.”

The Jedi mind trick: publishers duped?

According to the filings claims, while striking deals with Facebook, Google set about driving publishers away from header bidding and convincing them to re-route ads through its pipes. The lawsuit alleges that in 2018 Google started redacting data from the auction results it shared with publishers. That made it “nearly impossible for publishers to compare the relative performance of exchanges in header bidding with the performance of exchanges going through Google’s ad server. Consequently, Google renders the entire reason publishers use header bidding … unobservable and unmeasurable”.

Meanwhile the lawsuit claims Google then used search market dominance to “strong-arm publishers and advertisers to stop using header bidding and re-route trading through Google’s ad server”.

Header bidding relies on publishers inserting JavaScript code into webpages. According to the filing:

“Google created Accelerated Mobile Pages (AMP), a framework for developing mobile web pages, and made AMP essentially incompatible with JavaScript and header bidding. Google then used its power in the search market to effectively force publishers into using AMP.”

The documents allege it told publishers that using AMP would make their load times faster, but that was not always true, and that “Google throttles the load time of non-AMP ads by giving them artificial one-second delays in order to give Google AMP a “nice comparative boost.” Throttling non-AMP ads slows down header bidding, which Google then uses to denigrate header bidding for being too slow.”

While publishers are not forced to use AMP, the lawsuit claims: “Google uses its scale in search to punish publishers that do not choose AMP. Specifically, Google Search ranks non-AMP pages lower in search results and reserves the top placements in the “Search AMP Carousel”—the top search results placements with pictures—to publishers using AMP.”

The filing continues: “Google gave publishers a Faustian bargain: (1) publishers who used header bidding would see the traffic to their site drop precipitously from Google suppressing their ranking in search and re-directing traffic to AMP-compatible publishers; or (2) publishers could adopt AMP pages to maintain traffic flow but forgo exchange competition in header bidding, which would make them more money on an impression-by-impression basis. Either option was far inferior to the options available to publishers before Google introduced AMP. Just how inferior? According to Google’s internal documents, 40 percent less revenue on AMP pages.”

gTrade, price rigging, other programs

The case alleges Google created a series of programs designed to manipulate prices for its benefit. It states a team in Google’s New York office called gTrade designed a feature, called Reserve Price Optimization (RPO), that automatically increased the floor price of an ad spot to what a buyer was predicted to be willing to pay. If a publisher had set a floor price to a $10 CPM, RPO could increase the floor price to just below a small business’s predicted willingness to pay, which could be a $14.50 CPM.

“By adjusting floors in this manner, Google ensures that its own exchange transacts publishers’ most valuable impressions, even though an advertiser in a non-Google exchange would have otherwise won,” the lawsuit alleges.

It also created programs that allowed its own exchanges to win ad auctions – even if they were offering lower bids, in a situation that staff admitted “generates suboptimal yields for publishers and serious [sic] risks of negative media coverage if exposed externally”.

Per the lawsuit: “Google’s gTrade team launched another program called Dynamic Revenue Share (DRS) that leverages exclusive access to publishers’ ad server user IDs to exclude exchange competition in a second way. Google automatically opted publishers into the DRS program under the misrepresentation that it would make publishers more money.

“DRS dynamically adjusts the take rate that Google’s exchange charges in order to win more impressions, most particularly the high-value impressions.

“For example, if a publisher offers an impression for sale in Google’s exchange, but the highest bid cannot clear the publisher’s price floor due to Google’s take rate, DRS can dynamically lower Google’s take rate to ensure that the impression will still transact in Google’s exchange,” the filing continues.

“In order to know when and by how much Google should vary its take rate with DRS, Google must be able to accurately determine the value of impressions, which depends upon its access to publishers’ ad server user IDs. Google forecloses competition in the exchange market by blocking publishers from sharing their ad server user IDs with non-Google exchanges.”

There was also Project Bernanke, another called Bell and another called Elmo, which “also use inside information to privilege Google’s exchange over rival exchanges”.  

“In addition to representing both the buyers and the sellers of online display advertising, Google also operates the largest exchange, AdX. In this electronically traded market, Google is pitcher, batter, and umpire, all at the same time,” the complaint says.

“Google operates the largest electronic trading market in existence… [it] processes about 11 billion online ad spaces each day.”

NERA and “anticompetitive” Privacy Sandbox 

Another initiative highlighted by the case is Project NERA, which allegedly aimed to “create a closed ecosystem out of the open internet”. Using Chrome, rather than third-party cookies cookies, Google allegedly tracked publishers’ users and offered publishers access to their deeper trove of data – if they gave Google exclusive control over ad space. “For Google, Project NERA represented a win-win,” the case alleges.

Per the filing: “To get publishers to give Google exclusive access over their ad inventory, Google set publishers up for a lose/lose scenario. First, Google started to leverage its ownership of the largest web browser, Chrome, to track and target publishers’ audiences in order to sell Google’s advertising inventory.

“To make this happen, Google first introduced the ability for users to log into the Chrome browser. Then, Google began to steer users into doing this by using deceptive and coercive tactics. For example, Google started to automatically log users into Chrome if they logged into any Google service (e.g., Gmail or YouTube).

"In this way, Google took the users that choose not to log into Chrome and logged them in anyways. If a user tried to log out of Chrome in response, Google punished them by kicking them out of a Google product they were in the process of using (e.g., Gmail or YouTube). On top this, through another deceptive pattern, Google got these users to give the Chrome browser permission to track them across the open web and on independent publisher sites like The Dallas Morning News. These users also had to give Google permission to use this new Chrome tracking data to sell Google’s own ad space, permitting Google to use Chrome to circumvent reliance on cookie-tracking technology."

The extension of Project NERA was the Privacy Sandbox and the deprecation of third-party cookies. It means the company could “wall off the entire portion of the internet that consumers access through Google’s Chrome browser”.  

To summarise, the case states, “Google’s upcoming cookie changes in the name of privacy are a ruse to further Google’s longstanding plan to advantage itself by creating a closed ecosystem out of the open web.”

Google made $147bn in ad revenue in 2020 and will vigorously contest the allegations. See the full document here.