The history of advertising is a history of unmeasured waste. My whole career rests on one uncomfortable question: of every dollar a merchant hands to a platform, how much bought something, and how much evaporated? The Federal Trade Commission and a coalition of state attorneys general have now converted that question into a legal complaint about Amazon's advertising auction. The charge is not that the ads failed to sell. The charge is that the pricing rule itself was misrepresented — that sellers bidding for visibility were told one thing about how they would be charged and experienced another.
The framing is the story. Under the Section 5 framework of the FTC Act, the alleged victims are sellers rather than consumers 7, and the challenge aims at advertising-auction mechanics in which sellers are the paying parties 7. This is not a case about shelf prices. It is a case about the take-rate economics of the marketplace itself — the rent for the shelf, not the price of the goods on it.
The FTC alleges that advertising on the platform operates as a fixed, non-discretionary cost of visibility for the majority of private-label and brand-registered sellers 11. In my trade we had a word for a charge a merchant cannot decline. We called it rent.
What the Case Alleges
The mechanics
The core allegation is narrow and technical, which is what makes it expensive. Winners in the auction, the FTC contends, were made to pay amounts closer to their maximum bids than a true second-price auction would require 19. The dispute sits inside a long-running argument in auction design: the debate over second-price versus first-price formats 3. One strand of the record puts it without decoration — misrepresenting a second-price auction as a first-price auction is deceptive 3.
Note what the FTC's own analysis assumes, as characterized in the record: that bids are fixed and buyers do not respond to the floor when comparing a move from second-price to first-price formats 5. Hold that assumption. The auction literature on the same record does not treat bidders as furniture.
"Auction" as a marketing word
Sarah Caputo, founder of the consultancy Fraction Method 22, observes that "auction" has become "a marketing word instead of a defined term" in the industry 22. A defined term is the first requirement of an honest invoice. If the word no longer describes the mechanism, the seller cannot audit the bill — and a bill that cannot be audited is an open invitation to the waste fraction this complaint describes.
The Auction Theory Cuts Both Ways
The auction economics on the record deliver no verdict. They supply exhibits for both sides, and a careful reader should inspect both piles.
For the defense stands Vickrey's (1961) Revenue Equivalence Theorem, which holds that the two formats yield the same expected revenue, with truthful bidding under second-price rules and shading under first-price rules 5. If that result governs, the format label alone does not change what the auction collects. That reading favors the platform, and I flag it as inference rather than finding.
For the plaintiffs stand two related results. The "linkage principle" holds that a rival's bid conveys information about the winner's value 5. A soft floor, on this account, removes that information and lets bidders gain once they adjust their behavior 5 — the combination on which Brian C. Albrecht concludes that soft floors help bidders 5. Set that against the complaint's static assumption, which fixes bids and freezes buyer responses to the floor 5, and the argument becomes behavioral, not merely semantic. That behavioral dispute is not academic decoration. Technical and economic disagreements over the auction mechanics are themselves identified as a litigation risk 11.
The Second Front: Consumers Who Paid Twice
The seller is not the only alleged victim, and this is where the case reaches past the marketplace. The coalition alleges hidden surcharges raised advertisers' costs, many of which were passed on to consumers 15. Stated plainly, the allegation is that consumers were harmed through costs passed to them 17. On that theory, potential consumer-harm claims exist 24, and consumer class actions could follow if the FTC prevails 24. The FTC has also identified food and grocery items as essential goods affected by the alleged conduct 12 — a detail that widens the aperture from marketing budgets to household bills.
Any retailer knows this arithmetic without a courtroom. When the cost of the shelf rises, the ticket price follows. The seller signs the invoice; the customer pays the markup. A cost passed through is not a cost absorbed. It is a cost displaced — onto the party with the least visibility into why.
Remedies on the Table
What the plaintiffs ask for is broad. Remedies sought include an injunction and monetary relief 13,23, specified as a court order barring continuation of the alleged practices together with financial penalties, restitution and other damages 16 — in short, a halt to the alleged illegal practices 20. The coalition spans red and blue states 9, and the case could produce significant damages or remedies 18.
The downside scenarios run longer. Worst-case outcomes include ongoing operational restrictions on advertising practices 14 and cascading regulatory action in other jurisdictions 14. The broadest consequence would be standard-setting: the outcome could establish pricing standards for the digital advertising industry as a whole 8 — in plain terms, the rulebook for retail-media network economics.
The precedent weather
Enforcement in digital advertising has a record of modest winnings. The government has beaten Google twice, yet both victories are described as producing barely more than consolation prizes 21, with Judge Leonie Brinkema approving limits on how publishers use Google's ad tools — remedies short of divestiture 6. Meta, separately, faces a lawsuit brought by 30 U.S. states alleging it endangered children's safety through an addictive product 1, with an existential legal penalty identified as a distinct risk 1. Amazon, for its part, is already named among the public retailers facing regulatory and legal liability risk, alongside Kroger, Walmart, Albertsons, Ahold Delhaize and Costco 4.
A ruling on auction mechanics would cut deeper than any of those. Tell the industry that "auction" must mean what it says, and every platform's ad rate card becomes open to audit.
Why Sellers Keep Bidding: The Friction That Feeds the Auction
A defender of the status quo asks the obvious question: if advertising is supposedly a fixed cost, why do sellers not simply stop buying it? The record supplies the answer. Knockoffs and illegal resellers on Amazon are a problem even for larger brands selling their own proprietary products 3, and their presence forces merchants to maintain advertising spend for visibility and defense 3. That mechanism is the backbone of the argument that passing inflated advertising costs to consumers creates a "waterfall of harm" 3, and of the dismissal of counterarguments as technically true but practically false 3.
I recognize this pattern from my own sales floor. The brand pays for the window display to outsell the counterfeiter at the door. Defensive advertising is not demand creation. It is protection money booked as marketing — the purest form of attribution collapse, spend that defends position and gets recorded as growth, with no incrementality test anywhere in the ledger.
The harshest frame on the record comes from Cory Doctorow, whose analysis of Amazon supports the "Enshittocene" thesis that the worst ideas of the worst people now generate the most money 10, resting on a definition of enshittification as any taking from a customer that a competitor could win that business back by reversing 10. One need not adopt the label to see the structure it describes: a marketplace that monetizes the defensive spending its own frictions create has built a machine that bills for its own leaks.
What a Fair Reader Must Hold in Reserve
Three qualifications belong in any honest account, and I will not bury them.
First, the underlying material flags its own tilt: the seller-side content is described as one-sided and sympathetic to the seller 2. These are allegations. The record reads this case from the merchant's side of the counter, and a court may read it differently.
Second, the read-across to Amazon's own advertising docket is interpretive. The claims describe the action's mechanics without naming the defendant. In an Amazon-centered claim set the mapping is hard to avoid, but it is a mapping, not a fact, and I present it as such.
Third, the economics are unsettled. The soft-floor analysis 5 and Albrecht's conclusion from it 5 point one direction; the revenue-equivalence tradition 5 points the other; and the complaint's static-bidding assumption 5 is contestable on its face. Both parties will find their expert witnesses in this literature. Cost-per-acquisition integrity in this channel will be litigated, not decreed.
The Takeaway
Strip the legal vocabulary and the complaint describes a measurement failure with intent attached. Sellers were allegedly charged by rules other than the ones advertised; the resulting costs were allegedly rolled into shelf prices; and shoppers paid a second time without knowing they had bought anything. If the allegations hold, the waste in this channel was not the half nobody could measure. It was the half somebody designed.
The question is not whether Amazon's ads work. The question is how a seller is supposed to know what they cost, and whether the pricing rule is what the platform says it is. If a court rules that this auction was not an auction, every seller on every retail-media network will want yesterday's invoices re-audited. Which half of your advertising spend would you like to check first?