The core dispute is not whether reserve prices exist in digital ad auctions — they do, and they can raise publisher revenue — but whether Amazon’s ad marketplace matched what it told advertisers it was running. The FTC and 22 states say it did not, alleging years of undisclosed price floors and auction tweaks that inflated what 1.2 million advertisers paid; Amazon says the agency misreads how sophisticated buyers behave and how its auctions actually work.
At a Glance
- The FTC and 22 states allege Amazon secretly imposed “soft reserve” floors and related mechanisms that inflated ad prices paid by about 1.2 million advertisers since 2019.
- According to the complaint, Amazon represented a second-price auction while using undisclosed pricing controls that pushed charges above true competitive outcomes.
- Amazon disputes deception, pointing to documentation that reserves and pricing adjustments exist in its system and arguing that advertisers buy on performance, not auction labels.
- Economics supports that reserve prices lift platform revenue; the legal question is disclosure, consistency with stated rules, and whether practices were unfair or deceptive.
What the case is really about: disclosure, mechanism, and trust
Digital ad auctions are designed markets — code sets the rules, and those rules determine who pays what. The FTC’s lawsuit contends Amazon said one thing and did another: it told advertisers they were buying through a second-price mechanism — where the winner pays just above the runner-up bid — while internally applying “soft reserve” floors and other adjustments that raised the amounts charged beyond what open bidding would have produced. The agency pegs the affected population at roughly 1.2 million advertisers and characterizes the practice as systematic since 2019. Amazon rejects the charge, saying the portrayal is inaccurate and advertisers never pay more than their bid; it also emphasizes that pricing controls, including various reserves, are embedded in its documented systems.
In a second-price auction, truthful bidding is theoretically optimal because you pay near the next-highest valuation. Reserve prices — minimum acceptable charges — change the calculus. Set a floor above the true second price and the winner can be charged that higher amount. This is not exotic; it’s auction design 101. The crux, therefore, is not “did reserves exist,” but whether Amazon disclosed their presence and operated them in line with what advertisers reasonably believed they were buying. If the platform said “second price,” but silently layered algorithmic floors that persistently lifted charges, regulators have a straightforward deception theory; if Amazon adequately warned that dynamic reserves and pricing adjustments apply, then price elevation looks like standard yield management rather than a hidden surcharge.
How Amazon’s auctions are supposed to work — and where the friction lies
Amazon’s public materials describe auction-based buying across Sponsored Products and Sponsored Brands, with bidders setting maximums and the platform determining the final cost-per-click (CPC) or cost-per-thousand (CPM) through market competition and system controls. Those materials explicitly reference “reserves” that influence final CPC or vCPM based on predicted conversion, return on ad spend, placement, and other context signals. They also note that the charged price may exceed the runner-up bid but never the advertiser’s authorized maximum. In other words, Amazon’s own documentation acknowledges price shaping beyond a textbook second-price outcome.
The FTC’s case doesn’t hinge on the existence of controls; it asserts undisclosed mechanics — including “soft reserve” price floors — that, in effect, replaced the second-price logic advertisers were led to expect. News reports summarizing the complaint echo two load-bearing allegations: that Amazon continued to represent second-price auctions and that its internal pricing floors and adjustments systematically lifted advertiser charges above competitive levels. If proven, those facts would support a claim that the platform’s stated rules and its actual rules diverged in a way that mattered to buyers.
What economics says about reserve pricing — and why that does not settle the case
Large-scale field experiments and theoretical work in sponsored search and display show that reserve prices can materially boost publisher revenue; when set intelligently, they filter low-value impressions and extract more surplus from high-value bidders. That is a feature of optimal auction design, not a bug. Platforms regularly tune floors, test variants, and personalize thresholds by inventory segment or user context; such tuning is endemic in ad-tech and often invisible to individual bidders even when disclosed in principle.
But the legality turns on what was promised and what was actually delivered. If advertisers were told they were in a second-price world — a description that conveys a specific pricing intuition — and the platform simultaneously used undisclosed reserve logic that persistently elevated charges above the second-highest bid, regulators can argue the representation was misleading regardless of whether the reserves improved outcomes for some buyers. The literature justifies the tool; consumer-protection law scrutinizes the truthfulness and materiality of the platform’s statements about that tool.
Amazon’s counter-arguments and how to evaluate them
Amazon’s position, distilled, has three planks. First, advertisers never pay more than their bid, so there is no overcharge relative to the buyer’s explicit cap. Second, its documentation already signals that reserve-like adjustments exist and that prices can exceed the runner-up bid, undermining the claim of concealment. Third, sophisticated advertisers optimize to performance outcomes — conversions, incremental sales — not to auction labels; if return on ad spend holds or improves, then “harm” is illusory.
Each point has traction but limits. Bid caps constrain total exposure, yet they do not cure misrepresentation if the charging rule was described inaccurately. Documentation that gestures at reserves helps, but the FTC will test whether those disclosures were sufficiently specific and consistent with actual practice; vague nods to “other reserves” may not rescue explicit “second-price” messaging if buyers reasonably understood that to mean charges track the next-highest bid absent transparent modifiers. As for performance, courts often treat deception claims as turning on disclosure and materiality; strong outcomes do not immunize undisclosed pricing methods if buyers would have behaved differently knowing the facts.
Consequences for advertisers, platforms, and the market
If the FTC prevails, expect prescriptive remedies: clear disclosures of pricing mechanics, auditability of auction logic changes, and potentially restrictions or reporting around reserve-price governance. Monetary relief could be significant given the scale of Amazon’s ad business and the complaint’s scope, though any award would depend on proving both conduct and quantifiable harm. Even short of a courtroom win, the regulatory signal is unambiguous: in retail media and search-style auctions, it is no longer enough to say “market-driven”; platforms must document how system-level price controls interact with stated auction formats, in plain language intelligible to buyers.
For advertisers, the practical response is to negotiate for data and clarity: ask whether a second-price label is literal or shorthand; request documentation of reserve policies, update cadences, and the conditions that lift charged prices above the runner-up bid; and monitor diagnostics that can surface floor effects, such as discontinuities in win rates near specific bid levels. The playbook used for supply-path optimization in programmatic display — mapping paths, identifying floors, and testing — belongs in retail media as these networks scale.
Amazon is accused of rigging ad prices worth billions. In response, Amazon says the US FTC “misunderstands” their ad market. The lawsuit highlights concerns about competition, pricing tactics, and transparency in digital advertising. https://t.co/G3ZJbKAzMN pic.twitter.com/W7LKG1CgRZ
— Drew Grimaldi (@Grimillionaire) September 1, 2026
The broader pattern: designed markets demand designed transparency
The suit against Amazon joins a lineage of disputes where platform-controlled auctions evolved from simple stories to complex, yield-optimized systems — and buyers balked when the story and the system diverged. Reserve pricing, floor tuning, and predictive multipliers are legitimate tools of auction design supported by decades of research; deployed under a banner of “second price” without precise caveats, they become legal risks. The lesson for platforms is straightforward: make the invisible visible, not in a white paper nobody reads, but in the commercial terms and interfaces where buyers place their bets. The lesson for advertisers is equally clear: treat auction rules as product features to be diligenced like any other input to return, not as black-box lore blessed by a three-word label.
Sources:
ftc.gov, finance.yahoo.com, bclplaw.com, reuters.com, cryptobriefing.com, techpolicy.press, milkeninstitute.org