From purchase to return: How personalized E-commerce recommendations shape consumer behavior

As fashion e-commerce grows, personalized recommendation systems (PRS) are increasingly influential in shaping consumer decisions. Based on the DeLone and McLean IS Success Model, this study investigates how information, system, and service quality-along with immersion-affect perceived value, purchase intention, post-purchase satisfaction, and return intention. Using data from 299 online fashion shoppers and PLS-SEM analysis, the findings highlight perceived value as a key dual mediator: it partially mediates the effects of PRS quality on purchase intention and fully mediates the effect of immersion on purchase intention. While PRS quality and immersion enhance perceived value, only system and service quality directly influence purchase intention. However, neither perceived value nor purchase intention significantly affects post-purchase satisfaction, revealing a gap between expectation and experience. Post-purchase satisfaction is the sole significant predictor of return intention, with dissatisfaction driving returns. These results underscore the need for e-commerce platforms to align recommendation strategies with accurate product representation to reduce returns and build lasting trust.