How much of a sponsored placement's cost comes from competing bids, and how much from a number the advertiser never sees? Filings unsealed in the United States on November 5, 2026 describe Amazon lifting search advertising prices by as much as 50 percent above auction results on ordinary days and by as much as 250 percent on Prime Day or Black Friday, in a case brought by the Federal Trade Commission and 22 states that, as of November 13, 2026, remains at the pleading stage.
What the unsealed material adds
The dispute is not about whether advertising on a marketplace costs money. It is about who sets the price and how that price is disclosed. According to the filings, the mechanism moved in stages: a first-price share of roughly 4 percent at the end of 2020, 30 to 40 percent during 2021, about 70 percent in 2022 and close to 79 percent by 2024, as described in the case summary. A system that once cleared bids at a second price began clearing them at the bid, and the difference between those two conventions is the difference between a marketplace and a toll booth. For a seller, the change arrives as a higher cost per click with no line item explaining it.
The peak-season detail is the part merchandising teams will recognise. The FTC release reports that a holiday surcharge added about USD 87 million globally in the 2022 season and that an internal email from 2023 described a target of ideally at least USD 40 million. Neither number is large against total ad revenue, and that is the point: the charge was small enough per click to disappear into a cost-per-click curve that sellers already treat as weather. A case update records a seller complaint answered with the explanation that the spike was not a bug or a change in auction style but a direct reflection of it, which is the sentence that turns an operations annoyance into a governance question.
An auction with an invented bidder
The proxy that never existed
An auction is only honest if the participants are real. The filings describe an invented auction participant and a proxy second-price construction, both of which would have the effect of moving the clearing price upward without any advertiser bidding more. Whether a court accepts that characterisation is a separate question from what it does to measurement: a seller comparing its own bids with its own costs cannot see a synthetic participant, so the price increase looks like competition. That is the specific blindness price monitoring is supposed to catch, and it explains why the case matters beyond one platform.
From second price to first price
The shift to first-price clearing changes seller behaviour even when no surcharge is applied. Under a second-price rule, bidding your true value is roughly rational; under a first-price rule, every bid becomes a guess about the next bidder, and the rational response is to spend more time on bid management and less on assortment. Multiply that by a catalogue of thousands of SKUs and the operating cost lands on the same team that is supposed to be watching margins. A surcharge stacked on top of that convention does not create a new behaviour, it accelerates one that was already expensive.
What the peak-season numbers look like
Peak season is where the arithmetic becomes visible, because a percentage surcharge applied to a bidding pool that has already doubled produces a step change rather than a slope. The filings describe the pattern across four windows, and each window maps to a different decision a seller has to make about holding stock or buying visibility for it.
| Window | What the filings describe | Scale |
|---|---|---|
| End of 2020 | First-price clearing is rare | About 4 percent of auctions |
| 2021 holiday season | Average cost per click rises, with some categories far above the average | Up about 20 percent on average |
| 2022 holiday season | A surcharge is applied on top of clearing prices | About USD 87 million added globally |
| 2023 planning | An internal target is described in an email | Ideally at least USD 40 million |
Two of those rows are percentage claims and two are absolute ones, and mixing them is how the story gets overstated in either direction. A 20 percent rise in average cost per click during a peak season is unremarkable on its own; a 250 percent gap between the clearing price and the price charged is not, and the two cannot be averaged into a single number. The useful discipline for a seller is to keep the rows separate: benchmark the click cost against the same weeks last year, then benchmark the charge against the auction record, and only then decide whether the difference is competition or something else.
Why it lands on product prices
Advertising cost does not stay in the advertising line. On a marketplace where most categories run on thin margins, a higher cost per click is recovered through price, through pack size, or through a cheaper specification, and the third option is the one shoppers notice last. That is why an auction dispute is also a price monitoring problem: the same product page can show a stable headline price while the delivered specification quietly changes, and the monitoring that catches it has to read attributes as well as numbers. A seller watching only its own margin sees a cost; a brand watching the category sees a price gap opening between two listings that look identical in the search result.
For the teams building the checks, the practical answer is to treat ad cost as an input to price governance rather than as a separate budget. Connect the click cost to the listing it bought, compare the resulting landed price with the same SKU sold elsewhere, and flag the cases where the gap moved in the same week as a bidding change. None of that requires the litigation to conclude. It requires a record that survives a billing cycle, which is exactly what the filings say was missing.
FAQ
What is actually alleged in the Amazon advertising case?
The complaint alleges that Amazon raised search advertising prices above the auction results, by as much as 50 percent on ordinary days and as much as 250 percent on major shopping days, and that the practice was not disclosed to advertisers. The Federal Trade Commission filed with 22 states; Amazon's defence on the merits is a separate question the public docket does not settle at this stage.
Does the case change how sellers should run price monitoring?
It changes what the record has to contain. A seller that keeps only monthly spend cannot show what a click cost against the auction that produced it, so any price monitoring built on top of that data inherits the gap. Keeping campaign-level cost per click next to listing-level landed price is the minimum that makes a price anomaly arguable rather than anecdotal.
Why does first-price clearing matter to a brand that does not advertise?
Because the cost is passed through. When a category's sellers pay more for visibility, the recovery shows up in price, in pack size or in specification, and a brand comparing its own product page with a marketplace listing can see the effect without buying a single ad. Price monitoring that ignores marketplace ad mechanics tends to attribute those moves to the brand's own channel.
What should a category team ask for first?
The click-level record joined to the listing it promoted, for the peak weeks rather than the quarter. With that file, a team can test whether a price gap opened in the same window as a bidding change, which is the only version of the claim that survives a review with finance or with a retail partner.
