How Do Personalized Recommendation Systems Influence Consumer Purchase Intention? The Mediating Role of Perceived Value

Abstract

With the continuous growth of the number of products and user behavior data on e-commerce platforms, personalized recommendation systems have become an important tool connecting information filtering and consumption decisions. This study revolves around the following five parts: 1) Research Background: Existing research focuses primarily on algorithm accuracy; how consumers form value judgments requires further explanation. 2) Research Objectives: To examine how the personalization, accuracy, and timeliness of a recommendation system influence purchase intention through perceived value. 3) Research Methods: Based on a stimulus-organism-response model, 340 questionnaires were used to conduct reliability, validity, correlation, regression, and Bootstrap mediation tests. 4) Research Results: All three system characteristics significantly enhanced perceived value, with personalization having the largest standardization effect (Beta=0.412). Perceived value significantly promoted purchase intention (Beta=0.905). The three indirect effects were 0.442, 0.476, and 0.480, respectively, with 95% confidence intervals not including 0. 5) Research Conclusions: Perceived value played a partial mediating role in all three relationships, indicating that for a recommendation system to achieve purchase conversion, it cannot only pursue technical matching but also needs to ensure that consumers genuinely feel the recommendations are useful, timely, and meet their needs.