Leveraging Users’ Digital Habits: Real‑platform Data Reveals New Insights for Refined E‑commerce Operations

Fierce competition within the e‑commerce industry has left platforms grappling with persistent challenges of negative word‑of‑mouth and rising customer‑acquisition costs for repeat buyers. How to reduce negative reviews and boost repurchase performance at low cost has become a key operational pain‑point. An empirical study built on one‑year real transaction data from a leading domestic e‑commerce platform identifies the Digital Habitus Effect: users’ inherent review‑time routines exert significant influence over post‑purchase review sentiment and repurchase behaviour. These findings deliver actionable digital‑operation solutions for refined user management and negative‑comment governance.

This collaborative research involves Dr Chenwei Li from the International Business School Suzhou (IBSS), Xi’an Jiaotong‑Liverpool University. The paper, titled The Effect of Digital Habitus: How Adherence to Review‑Time Habits Influences Review Valence and Repurchase Among E‑commerce Users, has been published in Chinese Journal of Management Science, a top‑tier Chinese CSSCI and Peking‑University‑core management journal.

Existing e‑commerce literature mostly focuses on external drivers such as products and promotional subsidies, while paying limited attention to users’ intrinsic digital behavioural habits. Grounded in Schema Theory, this study analyses real‑world platform datasets covering 14,858 users and 73,405 orders. Multiple econometric approaches including the Logit model, deep‑learning instrumental variable (DNN‑IV) and Weibull‑AFT survival model are adopted for rigorous empirical testing. Results show that users are substantially less likely to post negative reviews when writing feedback within their habitual time windows. Habit‑adherence intensity and negative‑review probability follow a U‑shaped curve, with users at moderate habit intensity demonstrating the lowest tendency to leave negative feedback. Product involvement positively strengthens such inhibitory effect, whereas order value shows no significant moderating impact. Furthermore, adhering to personal review‑time routines generates positive spillover benefits: it shortens repurchase intervals and raises repeat‑buy frequency. By contrast, users without stable review‑time habits tend to produce more negative comments and exhibit longer repurchase cycles, representing relatively lower operational value. Heterogeneity tests indicate the effect is more pronounced for high‑value orders, non‑promotional periods and basic‑member user groups.

Drawing on empirical outputs, the study puts forward practical recommendations for e‑commerce platforms. Operators may build identification systems to map users’ preferred review hours and dispatch review reminders within individual habitual time‑windows to mitigate negative feedback under low‑cognitive‑load conditions. For high‑involvement categories such as home appliances and digital electronics, timed outreach aligned with user routines should be prioritised. For users lacking stable reviewing routines, timed notifications, point rewards and badge incentives can cultivate consistent digital‑review habits, improving long‑term review sentiment and repurchase outcomes. Platforms may also adopt the DNN‑IV modelling framework to build user‑behaviour prediction systems for early detection of high‑risk users prone to negative comments and low repurchase rates, enabling proactive intervention.

Against the backdrop of soaring customer‑acquisition costs, identifying, utilising and nurturing users’ review‑time digital habitus is no longer an optional auxiliary tactic. Instead, it constitutes a core strategy for platforms to implement refined operations, cut reputation‑governance expenses and enhance customer‑lifetime value. This research provides solid empirical evidence helping domestic e‑commerce platforms move beyond crude promotion‑driven growth toward low‑cost and sustainable user‑relationship management.

Dr Chenwei Li is now an associate professor in the International Business School Suzhou at Xi'an Jiaotong-Liverpool University, the deputy director of RCI-Emerging Technologies and Innovation and the deputy program director of MSc Project Management. Her research interests focus on social media and consumer behavior, human-computer interaction, and information security and privacy. Her work has been published in journals like Decision Support Systems, International Journal of Electronic Commerce, International Journal of Information Management, Information Processing and Management, etc., and conference proceedings such as ICIS, PACIS, WeB, CSWIM, and CNAIS, etc. She is serving as the editorial board member for Pacific Asia Journal of the Association for Information Systems (ABS-2, ABDC-B), guest editor for Internet Research (ABS-3, ABDC-A), and reviewer for many top academic journals. In addition, she has extensive experience in presiding over government research projects and industrial cooperation initiatives, with a number of external projects successfully approved.

Journal Introduction

Chinese Journal of Management Science, supervised and sponsored by the Chinese Academy of Sciences, is a leading domestic Chinese‑language journal for management science. It is indexed in CSSCI, Peking University Core and CSCD Core databases. It publishes high‑quality theoretical and empirical papers across management science, systems engineering, digital economy and business administration, and enjoys outstanding academic authority among China’s economics and management research communities.