Search ranking manipulation boosts sales of pricier own-brand products, study finds

Consumers shopping online are heavily influenced by platform rankings, often purchasing products listed at the top, regardless of price or quality, according to a new study. The Korea Fair Trade Commission (KFTC) released a report on "Consumer Behavior Experiments on Platform Algorithmic Self-Preferential Treatment," revealing that platforms can significantly boost sales of their own products simply by manipulating algorithms.

“

Consumers are strongly dependent on the rankings presented by platforms.

The study involved 3,072 participants using a virtual online shopping mall designed to mimic real platforms. In the first round, without algorithmic manipulation, sales were concentrated among the top five products, with 94.6% of purchases made on the first page. Only 25.2% of users changed the default sorting from "SC Ranking" to "low price," and a staggering 83.8% did not use filter functions to search for specific product features or price ranges.

When platforms manipulated algorithms to favor their own products, either by elevating their ranking or adding labels like 'SCpay +1% bonus', the purchase rate of these favored items dramatically increased. A product initially ranked in the middle or lower search results, with identical core attributes but a 10% price increase, saw its purchase rate jump from 1% to 35%. Conversely, competing products that were previously ranked higher experienced a 32 percentage point drop in sales, falling from 52% to 20%.

“

Consumers are easily misled into mistaking the rankings presented by platforms as reflecting certain quality signals, such as product quality or suitability.

Even when platforms disclosed that rankings might reflect business interests, the effect was minimal or counterproductive. The 'SCpay +1% bonus' label actually reduced active consumer search behavior, such as using filters, and further increased the purchase rate of self-preferential products by 4.5 percentage points. Only 10.7% of consumers checked the disclosure about ranking criteria. The researchers concluded that consumers easily mistake platform rankings for quality signals, leading to significant distortions in purchasing decisions. Worryingly, consumers who bought the 10% more expensive copied products perceived the platform rankings as fairer and more helpful than those who did not purchase, indicating a structural issue where consumers are unaware of the harm caused by algorithmic bias.

“

The consumer harm caused by self-preferential treatment is not just a distortion of choice, but a structural issue where consumers do not recognize the harm itself.