Amazon’s response to the FTC’s lawsuit regarding Sponsored Ads

Key takeaways The FTC wants the public to believe this case is about higher prices for consumers. It is not. Amazon's approach to pricing contradicts any suggestion of consumer harm.

We provide customers the lowest prices every day across the widest selection of products, and work to ensure our retail and grocery prices meet or beat those offered by other retailers.

On top of everyday low prices at Amazon, customers saved over $230 a year on average last year on essentials and other products through deals, coupons, and our Subscribe & Save reorder program. The FTC’s own complaint cites no evidence of consumer price increases, and consumers are only mentioned a handful of times in over 150 pages.

There is no advertiser harm either. From 2019 through 2024, the average cost-per-click for Amazon’s Sponsored Products search ads remained flat adjusted for inflation, while conversion rates grew 24% from 2021 to 2025. Advertisers paid the same and got more as we meaningfully improved ad relevancy and therefore performance.

The FTC’s claim fundamentally misunderstands how advertisers operate. Advertisers adjust bids based on real-world performance, not descriptions of auction mechanics.

Even accepting the FTC’s flawed premise that advertisers do not adjust bids, we estimate they saved over $8 billion from 2021 to 2025 as a result of Amazon prioritizing ad relevancy over selecting ads on bid price alone.

Average winning bids fell 50% from 2019 to 2025 on Sponsored Products search ads, and roughly 92% of placed ads are not given to the highest bid.

In 2026, we estimate that advertisers will deliver at least 58% higher sales with this model—rather than being ranked by highest bid amount—and at least 46% better return on ad spend because of our focus on relevancy.

Additionally, shoppers are 58% more likely to see ads for products they would consider clicking on or purchasing because we prioritize ad relevancy rather than simply the highest bid. The case centers on generalized second price auction dynamics, which the complaint itself concedes have been “the industry standard for decades.”

In no scenario does an advertiser pay more than their bid. After reviewing approximately 1. 5 million pages spanning six years, the FTC leans on a handful of simplified communications to allege a companywide effort to deceive. That is patently false. In the early years, Amazon. com only provided basic search results.

We, of course, wanted to show customers the best products, and our shopping results favored items with attributes such as many customer reviews, positive reviews, high click rates, and high purchase rates.

Over time, we found that existing items in our product catalog had many of these favorable attributes and became so favored in our systems that they inhibited new, sometimes more helpful products from surfacing in our search results. We also heard from brands that they wanted tools to enhance the discoverability of their products, particularly new ones.

One of the ways we worked to address this challenge was to introduce advertising so that shoppers could more easily discover new products and brands that better served their needs or more affordable alternatives they may not have known about, giving brands a new way to reach customers.

For example, customers searching for USB power cables for their phones saw best-selling, positively reviewed power cables that had been around for years. Advertising surfaced a new innovation—portable charging banks.

These were not yet showing near the top of search results but solved a significant customer need, demonstrated by their rise in popularity once they were advertised. Our early advertising was helpful for customers because it surfaced new products, but was powered by fairly rudimentary systems.

We used simple rules-based software systems to help ensure ads were relevant to shoppers, but these were imprecise, restrictive, and limiting for advertisers. Given the early stage of the advertising program at the time (and online advertising in general), it worked well enough across a limited range of categories and products.

However, over the following years, as the selection in our Store grew to hundreds of millions of products across 35+ categories (from books to electronics to toys to fashion to everyday essentials), the simple rules-based advertising software system did not effectively scale to enable advertisers to place relevant ads across our wide selection, and we started to invent with machine learning and AI to benefit both shoppers and advertisers.

We knew there were two things we had to prioritize: ad relevancy for shoppers and high-performing advertising for brands. Our philosophy was simple: customers discover new brands and engage with ads they are interested in, and advertisers get value and results when our customers are engaging with these ads.

In 2014, we started testing advanced machine learning-based relevance models in an attempt to predict how likely a sho