"Bid your true value. The auction will protect you." Amazon told its advertisers this for years. The FTC alleges that promise was false. A study published twenty-two years earlier names the mechanism behind it.

Here is how a second-price auction works. The ad industry calls the format GSP. You bid your max. The winner pays one cent more than the second-highest bid, not their own number. Bidding high is safe. The format keeps you from paying it.

Now picture an operator selling kitchen gear through Amazon Sponsored Products in 2021. Her cost-per-click starts climbing, week after week. She calls her Amazon rep and asks if the auction changed. The rep says no. More competition. She raises her bid, trusting the format to hold.

That scene played out across thousands of accounts. The trust made sense. The format did not.

The research on this is clear. In 2004, Chakraborty and Kosmopoulou published a paper in Economic Theory that asked a sharp question: what happens when the seller in the auction also bids in it? The practice is called shill bidding. Their finding was direct. In any auction with shill bidding, only the auctioneer gains. Bidders lose surplus. The house wins.

The paper came out twenty-two years before the FTC filed its complaint.

The problem is not that advertisers trusted a second-price auction. The problem is that the label replaced the need to check who else was bidding.

The FTC's case, filed late in the summer of 2026, alleges Amazon did what the study described. Internal documents quoted in the complaint show Amazon staff calling the pricing tool an "invented auction participant." The head of Amazon Ads allegedly said the second price "isn't set by an actual bidder" but by a "proxy 2nd price that we calculate."

The first-price rate is the share of auctions where the winner paid their full bid. The FTC alleges that rate went from 4% in late 2020 to 30 or 40% in 2021. By 2022, it hit 70%. By 2024, roughly 80%. That shift hit roughly 1.2 million U.S. advertisers. The FTC estimates the total extraction at over $20 billion.

A 2023 internal document notes that many advertisers set bids as high as "$100 or $1,000 because they trust Amazon is not running a first-price auction." They were bidding as if the format still protected them. The FTC alleges the format had changed. The instruction had not.

When advertisers asked directly if something had changed, an internal message stated: "We don't tell them about the surcharge, we let them assume GSP-based." One employee noted that advertisers would blame higher costs on competition, adding: "This also helps with external communication."

Most operators treated rising ad costs as a competition problem. It is a system problem.

The question worth asking is not "how do I bid smarter in this format?" It is "can I verify the format I am bidding in?"

Once that is clear, three moves follow from it.

Before you raise a bid on any platform auction, check one thing: does the platform that collects the revenue also set the reserve price? If yes, you are not in a clean second-price auction. You are in a format where the house controls the floor and takes the spread. That takes five minutes to look up. But it changes every number that comes after it.

Pull your actual cost-per-click data by week. Compare it to your stated bid. If you are paying your full bid more than half the time, the auction is not working the way a second-price format should. The FTC alleges Amazon's first-price rate climbed from 4% to 80% in four years. Your own data will tell you whether your platform follows the same drift.

Set a hard ceiling on what any single platform auction can cost you per quarter. The FTC complaint describes surcharges reaching 250% above auction results on peak days like Prime Day, with $87 million in extra revenue from one holiday window alone. A cap does not fix the format. It limits the damage when the format shifts without notice.

Running this for 90 days does something the advice never did:

It shows you which platform auctions charge you your own bid most of the time. It shows you which channels produce real margin and which ones extract it. It shows you where rising spend tracks actual new competitors and where it tracks nothing visible at all. And it puts a number on the gap between what you were told and what you paid.

At the end of the quarter, ask three things:

→ On which channels did your actual cost stay below your bid?

→ Where did rising costs track new competitors entering, and where did they track nothing you could see?

→ Which cap saved you money you would have spent on reflex?

That is the difference between advice that sounds right and a system that proves itself.

Built for the most ambitious companies in GTM, Attio is an end to end revenue platform that allows teams to run multiple plays across different revenue motions and lifecycles.

Teams at the frontier are already building on it: Granola, turbopuffer, Modal, and many many more.

Start building on Attio.

The operator who called her rep got the answer Amazon wanted her to hear. The mechanism the FTC describes was named and studied two decades before any complaint was filed. Your instinct was sound. The system was the variable.