For 250 popular items tracked over several weeks, Amazon ranked its own and affiliated products higher about 75% of the time, even when cheaper options existed
Context & Ripple Effects
This 2016 ProPublica analysis is the earliest data point in what became a decade-long evidentiary thread: tracking 250 popular items over several weeks, it found Amazon's own and affiliated products ranked higher about 75% of the time, often above cheaper alternatives. Later reporting kept confirming the pattern from new angles — private labels pitched as shoppers add rival products to baskets in 2019 (basket-add prompts), the top-left search slot used to promote private labels during the pandemic (top-left slot review), and a 2021 investigation finding own-brand placement above better-rated, better-selling competitors despite contrary testimony to Congress (Markup investigation).
What makes the original finding matter is that it predates all of that: it established the mechanism — ranking, not price or quality — as the lever, which is exactly what antitrust experts flagged in 2019 when they warned that burying items priced lower elsewhere would draw scrutiny (antitrust scrutiny warning).
First-order effects
- Shoppers searching for the 250 tracked items were steered toward Amazon's own and affiliated products in roughly three-quarters of cases, paying more than the cheapest listed option in instances where one existed.
- Third-party sellers of those popular items lost search visibility to house brands regardless of their ratings or prices, with placement — not merit — deciding who sold.
Second-order effects
- Sellers respond by distorting pricing across platforms, since the related reporting shows items become harder to find on Amazon when priced lower elsewhere — effectively taxing sellers into price parity.
- The findings give antitrust experts a concrete self-preferencing record to argue from, converting an abstract 'platform conflict' complaint into measurable ranking data regulators can act on.
Third-order effects
- If the pattern holds — and the 2016-to-2025 arc of investigations suggests it does — marketplace search stops functioning as a neutral ranking system and becomes a retail channel for the operator's own inventory, forcing regulators to treat ranking itself as a competition issue.
- Sellers rationally treat Amazon less as a storefront they compete in and more as a gatekeeper whose private-label ambitions cap their upside, reshaping where merchants list and how much margin they cede.
The trend: Platform self-preferencing at Amazon has persisted across a decade of investigations, with each new dataset strengthening the case that search ranking is the company's most powerful retail weapon.