We examine the effects of algorithmic pricing on e-commerce markets using transaction data from three major platforms covering consumer electronics, household goods, and apparel categories over 2022-2025. Our analysis compares pricing dynamics in categories with heavy algorithmic deployment versus categories with more traditional pricing structures.
Our methodology combines event studies around major algorithmic system updates with cross-sectional analysis of price variation across otherwise-similar products. We leverage natural experiments arising from platform policy changes that affected algorithmic pricing deployment.
Algorithmic pricing produces substantially more frequent price changes than traditional pricing structures. In heavily-algorithmic categories, approximately 23% of products experience price changes in any given day, compared to 3-5% in less algorithmic categories.
Price variance within products has increased measurably since algorithmic pricing became widespread. The standard deviation of prices for specific SKUs over 30-day windows has approximately doubled between 2020 and 2025 in our sample.
Average prices have not necessarily increased in heavily-algorithmic categories. Our analysis shows comparable average prices between algorithmic and non-algorithmic categories, but substantially more price discrimination within categories — meaning some consumers benefit while others pay more.
Price discrimination correlates with observable consumer characteristics — purchase history, session patterns, time-of-day of browsing. The welfare effects are therefore regressive for consumers with characteristics that algorithms identify as having higher willingness to pay.
Algorithmic pricing has implications for competitive dynamics that extend beyond individual transactions. an editorial review service reports that Rapid price matching and algorithmic coordination raise theoretical concerns about tacit collusion, though we find limited direct evidence of anticompetitive outcomes.
Market concentration effects are ambiguous in our data. Algorithmic pricing advantages larger platforms with more data and computational resources, but also enables smaller sellers to compete on price dynamically. Net effect on market concentration is mixed across categories.