The rapid growth of online shopping, facilitated by widespread mobile device usage and advanced internet technology, has transformed consumer behavior. Traditional e-commerce, however, suffers from limitations such as a lack of social interaction and visual appeal, resulting in a one-sided information flow. In contrast, live-stream commerce in China has experienced phenomenal success, combining influencer marketing with real-time interaction and dynamic product displays. By the end of 2020, approximately 703.4 million users in China actively engaged with live-streaming platforms, signifying a significant market penetration. This study aims to understand the reasons behind live-stream commerce's dominance, focusing on its unique communication style enabled by live broadcast technology. Existing research predominantly analyzes either individual psychological influences (motivation, personality, perception) or situational factors (social environment, infrastructure) on e-commerce adoption. A critical gap exists in understanding the combined impact of both individual and situational factors on this innovative business model. To address this gap, this study proposes an integrated model that considers both aspects, utilizing information processing theory (IPT) and flow theory within the Stimulus-Organism-Response (S-O-R) framework.

The study draws upon information processing theory (IPT), which explains how individuals encode, store, and retrieve information, affecting their cognitive and affective responses to online stimuli. The authors integrate flow theory, emphasizing the role of complete engagement in intrinsically rewarding activities in producing optimal experiences characterized by focused attention and exclusion of extraneous stimuli. This theoretical framework is used to understand how individual and situational factors impact the flow experience within the context of live-stream commerce. Prior research on e-commerce adoption has either isolated individual psychological factors or situational variables. This study bridges this gap by proposing an integrated model that combines both to predict the drivers and constraints of live-stream commerce. The model specifically examines the mediating roles of immersive experience and social interaction between three live-streaming attributes (dynamic characteristics, atmosphere clues, perceived interactivity) and online purchase intention.

The research employed a two-stage sampling process involving a pre-test and a formal online survey targeting Chinese consumers with live-streaming shopping experience in 2020. A total of 408 valid responses were collected, exceeding the recommended sample size for exploratory factor analysis and structural equation modeling. The questionnaire comprised three sections: a filter question on live-stream shopping experience, items measuring six constructs (dynamic characteristics, atmosphere clues, perceived interactivity, immersive experience, social interaction, and purchase intention) using a 5-point Likert scale, and demographic questions (gender, age, education, income). The measures were adapted from existing literature, with Cronbach's alpha values indicating acceptable internal consistency (ranging from 0.74 to 0.91 for individual constructs, and 0.96 for the overall scale). Confirmatory factor analysis (CFA) was conducted using AMOS 22.0 to assess the measurement model before proceeding with structural equation modeling (SEM) to test the structural model. The analysis evaluated construct reliability and validity (composite reliability and average variance extracted) and assessed model fit using various goodness-of-fit indices (χ²/df, GFI, CFI, IFI, SRMR, RMSEA). Simultaneous equations modeling was used to test the hypothesized relationships, examining partial and full mediation effects.

The CFA results indicated good construct reliability and validity for all six constructs. The SEM analysis revealed that the proposed model demonstrated acceptable model fit. The results supported the mediating roles of immersive experience and social interaction on the relationship between live-streaming attributes and purchase intention. Specifically: * Dynamic characteristics positively predicted immersive experience, which in turn positively predicted purchase intention (partially mediating the direct effect of dynamic characteristics on purchase intention). * Atmosphere clues positively influenced both immersive experience and social interaction, both of which positively predicted purchase intention (partially mediating the direct effect of atmosphere clues on purchase intention). * Perceived interactivity positively influenced immersive experience, which completely mediated its effect on purchase intention. Social interaction also completely mediated the effect of perceived interactivity on purchase intention. Social interaction did not mediate the relationship between dynamic characteristics and purchase intention. The model showed substantial explanatory power for the endogenous variables (R² values above 0.25 for immersive experience, social interaction, and purchase intention).

The findings demonstrate that live-stream commerce leverages a distinct communication style compared to traditional e-commerce, driving purchase intentions through immersive experiences and social interaction. The study's integrated model successfully identifies six pathways linking three key live-streaming attributes to purchase intention. The results highlight the critical role of both individual and situational factors in shaping consumer behavior in this context. The significant mediation effects suggest that fostering immersive experiences and social interaction is crucial for maximizing purchase intentions. While dynamic characteristics enhance the immersive experience, they do not necessarily translate to improved social interaction, suggesting a need for further development of digital infrastructure to strengthen consumer-streamer and consumer-consumer interaction. The findings are largely consistent with existing literature highlighting the positive impact of immersive and interactive communication on online shopping experiences and purchase intentions.

This study provides valuable theoretical and managerial insights into the success of live-stream commerce in China. The integrated model contributes to a deeper understanding of the mediating roles of immersive experience and social interaction in driving purchase intention. The findings emphasize the importance of enhancing both immersive experiences and social interaction in live-streaming platforms to increase purchase intentions. Future research should explore the model's generalizability across different e-commerce platforms and countries, investigate the impact of immersive technologies (AR, VR, AI) on presence and empathy, and examine how consumer's sense of control or trust affects the flow experience in live-streaming environments.

The cross-sectional nature of this study using data from a specific context (live-stream commerce in China) limits the generalizability of the findings. Future research should replicate the study in other cultural contexts and with different e-commerce platforms. The reliance on self-reported data may also introduce potential biases. The study did not consider various other factors that influence consumer behavior, such as product category, brand reputation and price. Further research could include these factors to provide a more comprehensive analysis.