Amazon's "Soft Reserve" Explained: What the FTC Lawsuit Means for How You Bid on Sponsored Products

TL;DR

On August 31, 2026, the FTC and 22 state attorneys general sued Amazon in federal court in Washington, alleging it added a hidden surcharge to its ad auctions starting in 2019 and concealed it from advertisers. Amazon denies wrongdoing and published a detailed rebuttal the same day. Buried in that rebuttal is the most useful thing either side has published for sellers: a clear explanation of how the auction actually prices your clicks. If you were taught to bid your maximum because "second price protects you," that advice needs updating. Here's why, and how to check your own account in about twenty minutes, ideally before Black Friday.

What the FTC Is Alleging

Amazon has told advertisers for years that Sponsored Ads run on a generalized second-price auction, the long-standing standard in digital advertising. In that model, winning doesn't mean paying your full bid. Amazon's own materials described the winner paying one cent more than the next highest bidder.

The FTC's complaint says that changed in 2019, when Amazon added an undisclosed pricing floor it called a "soft reserve price" internally, without telling advertisers. The agency argues this quietly converted a second-price auction into something functioning much like a first-price auction.

The number the complaint leans on: for Sponsored Products, the share of the time advertisers paid their own bid amount rose from between 30% and 40% in 2021, to 70% in 2022, to roughly 80% by 2024. The FTC says the surcharges were ramped up further ahead of high-volume days like Prime Day and Black Friday, and that the practice has run for over seven years across more than a million brands and sellers, including over 500,000 small and medium-sized businesses. The complaint puts the total extracted from advertisers at more than $20 billion.

The complaint quotes internal Amazon documents describing an "invented auction participant" used to raise prices, and alleges Amazon gave misleading answers to advertisers who asked directly whether the auction format had changed. The FTC is seeking a permanent injunction, and the states are seeking remedies including restitution and civil penalties.

Amazon calls the suit misguided and says it looks forward to making its case in court. Nothing here is proven.

What a Soft Reserve Actually Does

This is the part worth understanding regardless of how the case ends, and Amazon explained it plainly in its response.

There are two floors in the auction. A hard reserve is the minimum bid required to enter at all, which Amazon says covers its costs. A soft reserve is a real-time estimate of what that specific placement is actually worth.

Then the pricing works like this:

• If your winning bid clears both the hard and the soft reserve, you pay the soft reserve.
• If your winning bid clears the hard reserve but falls below the soft reserve, you still win the placement and you pay your full bid.
• Either way, Amazon says, "In no scenario does an advertiser pay more than their bid."

Read that second case again, because it's the whole ballgame. When the soft reserve sits above the second-highest bid, classic second-price logic stops applying. Your price is no longer set by your closest competitor. It's set by Amazon's real-time valuation of the placement, or by your own bid, whichever binds first.

Amazon says reserve prices are common across the ad industry and that its use of them is consistent with other platforms. That's fair. The dispute is not really about whether reserves exist. It's about timing and disclosure. A conventional reserve is a floor published before the auction runs. According to the complaint, Amazon's soft reserve is calculated after the auction has already picked a winner and a runner-up, and it is capped only by the winning bid. Whether advertisers were told that, and whether the gap between the marketing materials and the mechanics was material, is what the court will decide.

Why Both Sides' Numbers Can Be True at the Same Time

Most coverage of this story picks a side's statistics and runs with them. You don't have to.

The FTC says advertisers paid their full bid about 80% of the time in 2024. Amazon says the average winning bid for Sponsored Products search ads fell 50% from 2019 to 2024, that average CPC stayed flat adjusted for inflation over the same period, that conversion rates rose more than 24% from 2021 to 2025, and that roughly 92% of placed ads in 2024 did not come from the highest bidder.

Those datasets don't contradict each other. They describe two different things happening at once.

Amazon's account is that it shifted auction ranking to weight relevance much more heavily than bid amount. That would genuinely lower the bid required to win, which is what its 50% figure captures. But lower winning bids also meant placements were clearing below what Amazon thought they were worth, which is precisely the problem the soft reserve was built to solve. So the reserve catches the difference.

Net result: you can be winning at a lower bid than you would have in 2019 and paying your full bid far more often than you used to. The first is a real benefit. The second is a real change in how you should bid. Whether Amazon adequately disclosed the second is what a court will decide.

What This Changes About Your Bidding Strategy

Here is the practical consequence, and it holds whether or not the FTC wins.

The standard advice for second-price auctions is to bid your true maximum willingness to pay, because the mechanism protects you from actually being charged it. That's the entire logic behind "bid aggressively, the auction sorts it out." Economists call the opposite behavior bid shading: in a first-price auction, you deliberately bid below your true value, because your bid is your price.

If the FTC's 80% figure is anywhere close to right, a large share of your Sponsored Products clicks are priced first-price style. Which means bidding your theoretical maximum is not a protected strategy. It's just a high price.

The fix isn't to slash bids across the board. It's to change what number you're deriving your bid from. Instead of "the most this click could possibly be worth to me," anchor on target CPC computed from your actual margin:

Breakeven CPC = (profit per unit after all fees) × (conversion rate)

Then set your target below that by whatever margin you need. A product converting at 10% with $6 of profit per unit has a breakeven CPC of $0.60. If you're bidding $1.20 because you assumed second-price mechanics would settle you somewhere in the middle, and the soft reserve is instead landing you near your full bid, you're buying unprofitable clicks and the reporting will look like a conversion problem rather than a pricing problem.

Amazon's counterargument here deserves a hearing: it says sophisticated advertisers already adjust bids from real-world outcomes rather than auction descriptions, and that 80% of bid changes on clicked Sponsored Products ads happen within a day of a prior change. That's true of well-managed accounts. It's much less true of the seller running campaigns from a template with bids set once and revisited quarterly. Those are the accounts where this matters most.

Where the Reserve Is Most Likely to Bind

You can't predict a soft reserve in advance. Amazon says nobody can, itself included, because reserves are set in real time. But Amazon's own Help content now tells you what goes into them. Its Ads auction page says reserves can factor in the likelihood that an ad leads to a sale, your predicted return on ad spend, competing bids, placement, and whether it's a peak shopping period.

Read that list as a seller and three pressure points stand out:

• Your best-converting keywords. If a reserve rises with predicted sales and ROAS, the keywords where you convert best are the ones where Amazon values the placement most, and where your bid is most likely to become your price.
• Premium placements, especially top of search. Placement is an explicit input, and top-of-search is the placement Amazon prices highest.
• Event days. Amazon's own wording says the same placement can be worth more in a peak period. The FTC goes further and alleges surcharges were deliberately raised for Prime Day and Black Friday.

This is our reading of Amazon's description, not something Amazon has confirmed keyword by keyword. Which is exactly why the check below is worth running on your own data.

The Twenty-Minute Check: How Often Are You Paying Your Full Bid?

You can approximate the FTC's headline metric in your own account. It's one of the fastest diagnostics we run, and most sellers have never run it.

1. Pull your Search Term or Targeting report for the last 30 to 60 days, at keyword or target level, with cost and clicks.
2. Calculate average CPC per keyword (cost divided by clicks).
3. Work out your effective maximum bid for each keyword. This is the step people get wrong. Your effective ceiling is not your base bid. It's your base bid adjusted for placement modifiers and your bidding strategy. Amazon describes dynamic bidding as adjusting bids up to 100% for certain placements, and some automated strategies can go much higher. If you run "dynamic bids, up and down" with a top-of-search modifier, your true ceiling can be substantially above the number in the bid column. Compute it properly or the comparison is meaningless. (Worth knowing: according to the complaint, Amazon's own team noted internally that many advertisers chose up-and-down bidding precisely because they assumed they'd never be charged their full bid.)
4. Compare. Divide average CPC by effective max bid. The closer that ratio sits to 1.0, the more you are effectively in a first-price auction on that keyword.
5. Segment the results. Sort by ratio and check whether the highest ratios cluster on your best converters and your top-of-search placements, where the pressure points above say they should. Those are also where the most money sits.
6. Compare an event window to a normal week. Prime Big Deal Days ran October 6 and 7. Once that data settles, run the same ratio for the event days against a normal week in September. If the ratio jumps during the event, that's the pattern to plan around before Black Friday and Cyber Monday (November 27 to 30), the biggest ad week of the year.

If a large share of your spend is running at a ratio near 1.0, your bids are functioning as prices, and every bid should be set from margin math rather than from an aspirational ceiling. For the full Q4 calendar those bids need to survive, see our Q4 2026 Amazon deadlines breakdown.

Two honest caveats. First, this is a diagnostic, not evidence. A high ratio is consistent with the FTC's theory but also with a keyword where you're simply outbidding everyone by a wide margin. Second, if you want cleaner data, Amazon Marketing Stream delivers hourly performance at the keyword and placement level, which makes this analysis considerably more precise than a pulled report.

Amazon's Defense, Fairly Stated

Amazon's rebuttal makes several arguments that will matter in court.

It says the Ad Console campaign builder, where advertisers actually set bids, has stated since 2018 that a bid represents the maximum an advertiser could be charged. It characterizes the training videos and courses the FTC cites as low-reach materials that were missed in an audit and later corrected, noting that three of the cited courses drew fewer than 1,900 enrollments combined over their lifetimes.

It argues there is no consumer harm, pointing out that the FTC's own complaint barely addresses consumers across more than 150 pages and that its damages model assumes no pass-through.

And it argues the FTC built a company-wide deception narrative out of a handful of simplified communications found in roughly 1.5 million pages of documents, which Amazon calls patently false.

Amazon also says it has since updated its Help content to explicitly describe reserve prices, and now runs regular audits and sales trainings on auction communication. According to the complaint, the first mention of reserve pricing appeared on a single Support Center page on October 30, 2025, about a year after Amazon learned of the investigation, and was revised in March 2026 to say your price may exceed the runner-up bid but never your maximum bid. Worth noting: that update is itself the clearest confirmation that the earlier guidance wasn't explicit about this.

What Happens Next

As of mid-October 2026, the case is at its earliest stage. Expect Amazon to move to dismiss. If the case survives, discovery is where the internal documents behind the complaint get tested in full. This is the FTC's third major action against Amazon, and cases at this scale run for years.

Three things to watch: whether the case clears dismissal, whether Amazon makes further voluntary changes to auction disclosures, and whether private class actions get filed and consolidated on behalf of advertisers.

FAQ

What is a soft reserve price on Amazon Ads?

A real-time minimum price Amazon sets for a specific ad placement, reflecting its estimate of that placement's value. If your winning bid clears it, you pay the soft reserve. If your bid falls below it but clears the hard reserve, you win and pay your full bid.

Does Amazon charge more than my bid?

No. Amazon states an advertiser never pays more than their bid, and the FTC does not claim otherwise. The dispute is about how often advertisers pay their full bid rather than a second-price amount, and whether Amazon disclosed that.

Does the soft reserve go up on Prime Day and Black Friday?

The FTC alleges Amazon raised surcharges further on high-volume days like Prime Day and Black Friday. Amazon's own help content says the same placement can have a different value during a peak shopping period than a non-peak one. Either way, event days are where your bids are most likely to function as prices.

Should I lower my Amazon Ads bids because of this lawsuit?

Not reflexively. Recalculate bids from your breakeven CPC rather than from a maximum willingness to pay, then compare that to what you're currently bidding. For some keywords that means lower bids, for others it means higher ones.

Am I part of a class action?

No. This is a government enforcement action, not a class action. The FTC is seeking a permanent injunction, and the states are seeking remedies including restitution and civil penalties. If private class actions follow, membership and notice would work separately.

Was every advertiser affected equally?

Unlikely. The alleged effect concentrates in competitive placements and in peak periods like Prime Day and Black Friday, so exposure varies by category, keyword competitiveness, and how much spend runs during events.

Bottom Line

The FTC says Amazon hid a surcharge in its ad auction for seven years and made tens of billions doing it. Amazon says the auction works as designed, advertisers never pay more than they bid, and the complaint misreads how ad buyers behave.

That argument will take years. The takeaway for sellers doesn't have to wait. Both sides now agree on the mechanic: a real-time price floor sits in the auction, and when it binds, you pay your bid. So stop treating your bid as a theoretical ceiling you'll never actually be charged. Treat it as the price. Then run the numbers before Black Friday and find out how much of your spend has been priced that way all along.

Sources

• FTC, States Sue Amazon Over Secret Ad Surcharge Scheme (FTC press release, August 31, 2026)
• FTC et al. v. Amazon.com, Inc., Complaint, Case 2:26-cv-03097 (W.D. Wash., filed August 31, 2026)
• Amazon's response to the FTC's lawsuit regarding Sponsored Ads (About Amazon, August 31, 2026)