ANALYSING THE IMPACT OF MULTI-CRITERIA RECOMMENDER SYSTEMS ON USER ENGAGEMENT IN E-COMMERCE PLATFORMS

Abstract

Recommender systems have become integral to modern e-commerce platforms but often yield inconsistent user engagement and satisfaction. This study investigates how multi-criteria recommendations, factoring in user preferences, browsing history, and demographics, influence both perceived usefulness and continued platform use. Employing a mixed-methods approach—combining reliability checks (Cronbach’s Alpha, KMO) and structural modeling—we analyzed user feedback from two major Indonesian marketplaces (Tokopedia and Shopee). Results show that with well-targeted recommendations and higher engagement, it significantly enhances optimal user experience, yet poorly tuned suggestions can diminish trust. These findings highlight the importance of personalization quality, affirming that recommender systems must be accurate, context-aware, and adaptable to varying user preferences for optimal effectiveness.