Matt Wasserlauf, now the CEO of Blockboard, remembers it well. 

At the time he was an ad tech operator fielding a request from one of the world’s largest advertisers. Procter & Gamble wanted him to verify whether DoubleClick — the gold standard for display ad serving — was really delivering on its promises. The marketers suspected they were being overcharged for banner ads and DoubleClick wasn’t offering any real answers. With Google circling the company, P&G didn’t expect things to get any more transparent. 

So they asked Wasserlauf to put some checks and balances in place. The result was Vindico, an alternative ad server created to shine a light on what many suspected was a black box.

Even back then, ad execs had a gut feeling about Google running DoubleClick — and it wasn’t a good one.

Now, that gut feeling has become common knowledge, and as messy as it is, here’s the gist: Google’s purchase of DoubleClick gave it the tech that powered how ads were served on the web. That meant it could control the tools used by publishers then later build tools for advertisers and eventually run the ad exchange (the auction itself). By owning all three layers — buyer, seller and auctioneer — Google could steer the market to its advantage often behind the scenes. 

The implications weren’t immediately visible but the industry felt them. Three years later, when Google acquired Admeld, a supply-side platform used by publishers, it looked less like a business deal and more like a power move. 

Publicly, Google painted the $400 million acquisition as a way to help publishers make more money from their inventory. Privately, it had more Machiavellian uses. The Department of Justice would later unearth discussions from Google execs pondering whether they should buy Amdeld simply to “park it somewhere”. 

With Admeld in Google’s hands, publishers feared the worst. They saw the deal as Google doubling down on its DoubleClick playbook: buying inventory from publishers at low prices, selling it to advertisers at higher ones, and pocketing the difference — all while keeping publishers in the dark.

Naturally, publishers have more concerns — some even ditched Admeld altogether.

“After the [Admeld] acquisition, we received a lot of publisher clients that fled Admeld,” said a former AppNexus exec who exchanged anonymity for candor on their time at this particular ad tech business.

Fast forward to late 2013, and Admeld had found its parking spot at Google — only to be shut down soon after — just three years after it was bought.

Google’s next steps were less subtle. It began manipulating auction dynamics through mechanisms that bore the names of internal projects like Bernanke and Global Bernanke. These were designed to inflate bids when Google needed to win and redistribute ad revenue between publishers without transparency. Another effort project Bell punished publishers that dared to work with rival ad tech providers by discounting their inventory by up to 20%

With that kind of control, Google could tilt the scales, growing its ads business at the expense of competitors, partners, and even its own advertisers. Yet, few called it out. Sometimes it was ignorance, sometimes willful ignorance, and sometimes? Just plain greed. Whatever the reason, a large swath of the ad industry stood by — complicit, whether they knew it or not.

If that sounds overblown, this ad exec’s unfiltered perspective offers a stark reality check:

“There’s a frustration that the Google tech we’re using to buy ad impressions works best when buying them from its own marketplace of publishers using its ad server,” said the exec, who requested anonymity to avoid jeopardizing commercial relationships. In 2015, when these concerns peaked, the exec and their colleagues were pushing to consolidate their ad spend into fewer programmatic marketplaces. The trouble was, they couldn’t tell if Google’s marketplace outperformed because it was superior — or because Google was making it look that way.

Still, they spent the money. Because in digital media, if something appears to work, no one asks too many questions. As the exec explained: “When something works in media buying, it gains traction — no questions asked.”

Publishers, however, had fewer illusions. And unlike advertisers, they began fighting back. 

In 2014, a workaround emerged: header bidding. The technique allowed publishers to run simultaneous auctions in multiple ad exchanges before passing inventory to Google, cutting into AdX’s advantage and giving rival marketplaces a fair shot.

It worked — briefly. Publishers began earning more. Competition increased. For the first time in years, Google was no longer guaranteed a front-row seat. 

Google didn’t take it lightly.

The company’s retaliation unfolded in phases. First came DFP First Look in 2015, a feature that allowed Adx to outbid exchanges at the last minute, often by just a penny. When publishers resisted, Google pivoted. In 2016, it introduced Open Bidding, an alternative to header bidding that let other ad tech vendors into its own auctions — with limitations, of course. It still controlled the auction, kept data advantages, charged fees and gave Adx subtle edges that preserved its dominance. 

Then came the scorched-earth campaigns.

Project Poirot identified and targeted header bidding partners Google deemed risky — ad tech vendors were riddled with price manipulation, costing its own platform money every time ads were bought through them. They had to go. Eventually, they did. Google’s bids into rival exchanges dropped anywhere from 10% to 40%, and eventually as much as 90%, leaving its own marketplace in pole position to scoop up that ad spend. When publishers tried to push back and implement custom rules to regain some control, Google unveiled Unified Pricing Rules, which banned floor-price tweaks that allowed non-Google bidders to compete fairly. The official line was simplicity. The internal motive, as the DoJ later confirmed, was to kill header bidding. 

For Stefan Havik, head of digital at European media company DPG Media, that was the breaking point. The company began a strategic overhaul to minimize its dependence on Google. To continue playing by Google’s rules, he feared, would mean surrendering autonomy entirely. 

“Those rules — that was the moment for us when we decided that the dynamic was going the wrong way and so we developed a plan to be less dependent on Google,” said Havik at Digiday’s European Publisher Summit last fall. 

It’s a sentiment that’s grown louder in recent years, now echoed not just in hushed meetings but on legal dockets and court transcripts. The receipts are on record.