Abstract: Technological advancements enable e-commerce platforms to reach users 24/7, fostering the formation of Digital Habitus—consistent patterns of behavior and cognitive adaptations—through user interaction with digital environments. Research on digital habitus remains nascent. Although a few studies suggest that adhering to shopping-time habits can enhance consumer satisfaction and revisit intention, little is known about how digital habitus can be leveraged for user relationship management. Against this background, this study focuses on post-purchase review and repurchase decision-making, adopting Schema Theory to clarify the cognitive logic of user behavioral responses. Based on one-year real e-commerce data covering 14,858 users, over two million product interactions and 73,405 order comments, this paper constructs empirical models to explore the joint relationship between review-time habits, comment valence and repurchase performance. Combining the Logit model, deep learning instrumental variable approach (DNN-IV) and Weibull accelerated failure time (Weibull-AFT) model, this study systematically identifies the core effect, moderating mechanism and heterogeneous characteristics of review-time habit adherence. Key findings reveal: (1) Compliance with review-time habits significantly inhibits negative comment tendency, and this finding remains robust after addressing endogeneity. A U-shaped nonlinear correlation exists between habit intensity and negative evaluation probability. (2) Product involvement positively moderates this linear inhibitory effect. (3) Heterogeneity tests further verify that such suppression effect is more salient for high-value orders, non-promotion periods and basic member groups. (4) Review-time habit adherence generates positive spillover effects on repurchase behavior by shortening repurchase intervals, while negative comments exert limited influence within this pathway. (5) Users without stable review-time habits exhibit a higher negative evaluation tendency and longer repurchase cycles. This study clarifies the internal mechanism and boundary conditions of digital habits in driving consumer review and repurchase behaviors. With objective platform transaction data and multi-method empirical design, it supplements theoretical evidence in the digital habit domain, deepens the understanding of differentiated user responses under time-based behavioral patterns, and provides theoretical reference and practical implications for e-commerce platforms to optimize information system design and refined user operation strategies.

Key words: digital habitus, negative word-of-mouth management, repurchase, product involvement, schema theory, DNN-IV, Weibull-AFT