Abstract
This study empirically investigates the impact of platform encroachment and algorithmic recommendation adjustments on third-party (3PS) seller welfare within multi-sided e-commerce ecosystems. Utilizing a high-frequency panel dataset spanning 42,000 product listings on Amazon's marketplace from 2021 to 2023, we exploit a quasi-experimental policy shift in the proprietary 'Buy Box' algorithm that altered weighting parameters for platform-fulfilled and private-label products. Employing a staggered difference-in-differences estimation framework, our results reveal that algorithmic changes favoring platform-owned products reduce 3PS Buy Box win rates by 14.2% and decrease quarterly merchant revenues by an average of 11.8%. We find that independent sellers attempt to compensate for diminished algorithmic visibility by engaging in aggressive price discounting, which leads to significant margin compression and an elevated probability of seller exit among small-to-medium enterprises. However, participation in Fulfillment by Amazon (FBA) partially mitigates these adverse outcomes, highlighting a strategic dependency wherein sellers surrender logistics margins to preserve organic buy-box positioning. These findings contribute to the literature on platform governance, competitive marketing strategy, and antitrust policy, demonstrating how algorithmic mechanisms can institutionalize platform encroachment and reshape market dynamics to the detriment of competitive seller ecosystems.
