Almost every account of the Federal Trade Commission’s new lawsuit against Amazon leads with the same number: $20 billion, allegedly extracted from advertisers since 2019. It is a good number. It is also the least interesting one in the complaint, and reporting it without the mechanism underneath leaves readers with a vague impression of corporate greed rather than the specific, documented, and unusually provable thing the FTC says happened.
The number that matters is 79.1%. That is the share of Sponsored Products clicks in 2024 for which Amazon charged advertisers their own full bid, rather than the runner-up price its own marketing had promised since 2014. And the reason it matters is the trajectory: roughly 4% in late 2020, 30 to 40% through 2021, about 70% by 2022, and 79.1% by 2024. That is not a bug that crept in. That is a dial being turned.
What Amazon Said It Was Selling
For more than a decade Amazon told advertisers it ran a second-price auction, the standard mechanism across digital advertising. You bid, and if you win you pay a penny more than the next highest bidder. The appeal is that honest bidding is the optimal strategy, so advertisers can bid their true value without fear of overpaying.
The FTC alleges that in December 2018, starting with Sponsored Brands over the Christmas period, Amazon inserted an undisclosed floor it called a “soft reserve price.” It reached Sponsored Products in mid-2019 and Display Ads in 2023. The complaint describes a two-stage process: Amazon runs the generalized second-price auction it advertised, then applies the undisclosed reserve on top, and charges the result.
Amazon’s own people described it plainly in internal documents. A senior scientist called it an invented auction participant representing how much Amazon thinks a particular ad slot is worth. A senior vice president called it a proxy second price that Amazon calculates.
A phantom bidder that always bids roughly what you bid is not a bidder. It is a price.
Why This Is Harder to Defend Than the Monopoly Case
This is Amazon’s third major FTC suit, and it is structurally different from the others in a way that should worry the company more.
Monopolization cases turn on contested economics: market definition, whether conduct is exclusionary or merely competitive, whether consumers were harmed on net. Expert witnesses disagree for years. Amazon has fought that ground before and has real arguments.
This case turns on a comparison between a representation and a log file. Amazon said winners pay slightly more than the second-place bidder. Amazon’s systems recorded what winners actually paid. If the FTC’s 79.1% figure survives discovery, the gap between the promise and the practice is arithmetic, not opinion. The internal quotes make the intent question difficult too, because they show engineers and executives describing the mechanism accurately in private while the public documentation said something else.
The scale is what turns a pricing dispute into a political problem. More than 500,000 of the 1.2 million affected advertisers are small and medium-sized businesses, the constituency both parties compete to defend, and Amazon is now the third-largest advertising platform in the United States.
Amazon’s Defense Answers a Different Question
Amazon calls the suit misguided, says it fundamentally misunderstands how advertisers operate, and notes the complaint offers no evidence that consumer prices rose. It also says its auction systems saved advertisers roughly $8 billion between 2021 and 2025.
The consumer-price point is fair, and the FTC’s pass-through theory is the weakest part of the case. The $8 billion claim is not fair, and it is worth being precise about why. Comparing outcomes against a hypothetical first-price auction says nothing about whether Amazon disclosed the mechanism it was actually running. A locksmith who charges you more than he quoted has not answered the complaint by proving he charged less than a different locksmith would have. The question is what Amazon told 1.2 million customers it was doing.
Here is where BusinessTech.News lands. The pass-through claim should probably fail, and the FTC weakened its own filing by reaching for it. The disclosure claim should succeed, because the ratchet from 4% to 79% over four years is very hard to characterize as anything other than a deliberate, incremental migration away from the advertised mechanism, executed quietly enough that the advertisers paying for it did not notice. Amazon did not have to run a second-price auction. It had to stop saying it did.
The remedy that would actually matter is not a fine. It is compelled disclosure of auction mechanics, audited, on an ongoing basis, for any platform of this scale. The FTC has recently shown appetite for structural remedies over cash, having made Redfin re-enter a business Zillow paid it $100 million to leave. A cheque from Amazon is a rounding error. A permanent obligation to show advertisers the real price-setting logic would reset an industry where roughly nobody can currently verify what they are being charged.
Watch whether Amazon moves to dismiss on the pass-through theory alone. If it does, that is a signal it likes its odds on consumer harm and does not want the 79.1% number argued in open court.