Live streaming shopping (LSS) has rapidly grown in China, becoming a significant e-commerce method. However, high return rates and low repurchase rates pose challenges to its sustainability. This study addresses these challenges by exploring the relationship between self-perceived customer experience and repurchase intention among LSS users. The dramatic rise in LSS users in China, reaching 526 million in June 2023, is accompanied by concerningly high return and cancellation rates (47.4%). This discrepancy highlights a critical need to understand and improve the customer experience. LSS differs from traditional e-commerce in several key aspects: the shift of user attention from products to streamers, the high level of real-time interaction, and the emphasis on discounted pricing and after-sales service. While prior research has focused on purchase intention, this study examines repurchase intention and the role of self-perceived experience by investigating whether users' perceptions align with their initial expectations. The study uses expectation-confirmation theory (ECT) to develop a framework for understanding user behavior in the post-purchase stage and aims to provide valuable insights for improving user retention and the long-term growth of the LSS model.

The study draws on two primary theoretical frameworks: the Theory of Reasoned Action (TRA) and the Expectation-Confirmation Theory (ECT). TRA explains the relationship between attitudes, intentions, and behaviors, providing a foundation for understanding purchase and repurchase intentions. ECT, focusing on post-purchase evaluation, examines how the comparison of perceived performance with pre-purchase expectations shapes satisfaction and repurchase intentions. Previous studies using TRA have examined factors influencing purchase intentions for various products, such as organic food and luxury goods. Research employing ECT has investigated repurchase intention in e-commerce and the continuous use of information systems. This study utilizes ECT to analyze the impact of perceived experience on repurchase intention in the context of LSS, an area with limited prior research focusing on post-purchase behavior.

To test the proposed research model, the study employed structural equation modeling (SEM). Data was collected through an online questionnaire survey administered in China between July 6, 2022, and September 16, 2022, using the Questionnaire Star platform. Two pre-surveys were conducted to refine the questionnaire. The sample consisted of 507 respondents who had prior experience with LSS. The questionnaire measured several key variables: perceived interactivity, perceived quality, perceived discounts, perceived value, confirmation, satisfaction, and repurchase intention. The reliability and validity of the measurement scales were assessed using Cronbach's alpha, composite reliability (CR), average variance extracted (AVE), and the Fornell-Larcker criterion and Heterotrait-Monotrait ratio (HTMT) for discriminant validity. Common method bias was examined using both a single-factor test and a confirmatory factor analysis (CFA) comparing models with and without a common latent factor. Non-response bias was also checked. SEM was then used to test the hypothesized relationships between the variables, followed by mediation analysis to examine the mediating role of confirmation.

The SEM analysis supported most of the hypotheses. Perceived interactivity, perceived quality, perceived discounts, and perceived value all significantly and positively influenced confirmation. Perceived discounts showed the strongest influence on confirmation (β = 0.325, p < 0.001), while perceived quality had the weakest (β = 0.149, p < 0.01). Confirmation significantly and positively impacted satisfaction (β = 0.300, p < 0.001), which, in turn, significantly and positively impacted repurchase intention (β = 0.757, p < 0.001). Perceived interactivity, perceived quality, and perceived value also directly and significantly influenced satisfaction. However, perceived discounts did not directly influence satisfaction (β = 0.015, p > 0.05). Mediation analysis using the bootstrapping method revealed that confirmation completely mediated the relationship between perceived discounts and satisfaction, and partially mediated the relationships between perceived interactivity, perceived quality, and perceived value and satisfaction. Of the control variables, only gender significantly impacted repurchase intention. The model explained 59.5% of the variance in repurchase intention, 60.6% of the variance in satisfaction, and 41.6% of the variance in confirmation.

The findings demonstrate that self-perceived experience, mediated by confirmation, plays a crucial role in shaping LSS users' satisfaction and repurchase intention. The strong influence of confirmation highlights the importance of managing user expectations and ensuring that the actual LSS experience aligns with these expectations. The significant, direct effect of perceived quality, interactivity, and value on satisfaction underscores the need to provide high-quality products and an engaging and convenient shopping experience. The complete mediating role of confirmation in the relationship between perceived discounts and satisfaction suggests that simply offering discounts is insufficient to improve satisfaction; consumers must also perceive these discounts as genuine and valuable. The gender effect highlights the need for targeted marketing strategies to cater to the specific preferences and needs of different user segments. These findings offer valuable contributions to both the existing literature on consumer behavior and the practical applications for improving LSS platforms.

This study contributes to the understanding of LSS users' repurchase intention by focusing on the post-purchase experience and the mediating role of confirmation. It emphasizes the importance of managing expectations, providing high-quality products and engaging experiences, and employing gender-sensitive marketing. Future research could explore the reasons behind the low repurchase rates, integrate dual-processing theory to better understand information processing during LSS, utilize eye-tracking data for objective measures of user experience, and examine actual purchase behavior using transaction data.

The study's limitations include its reliance on self-reported data from experienced LSS users, which may not fully represent the broader consumer population. The cross-sectional nature of the study limits causal inferences. Future research should address these limitations by including a wider range of participants, using longitudinal data, and incorporating objective measures of user experience.