Author(s): LI, Z., GUO, Z.1, WANG, Y.2, Institution: TSINGHUA UNIVERSITY, Country: CHINA, Abstract-ID: 291

Introduction
The scale of live-streaming users reached 765 million, and the transaction scale in 2022 reached 3.5 trillion in China. Live streaming has offered new opportunities for retailers to increase their sales (Zheng et al., 2023). Many sporting goods brands began to take advantage of live-streaming platforms to improve product sales volume. Previous research focused on the anchors’ features how to impact the purchase intention; however, the underlying mechanism by which scene feature influences consumers’ purchase intention is still unknown. Thus, this study aims to explore the relationship between purchase intention (PI) and scene features，which include visual appeal (VA), presence (PR), scene-product matching (SM) in the sporting goods field. Flow experience (FE) and sport identification (SI) as mediating and moderating variables respectively.
Methods
Data were collected through snowballing referrals on social software (i.e., WeChat and QQ) and several anchor’s fan groups of sporting goods e-commerce live streaming via an online questionnaire from March 10 to 20, 2023. 340 valid questionnaires were gathered. Using structural equation modeling to analyze the path influence relationship between the variables and test hypotheses through SPSS26.0 and AMOS 26.0.
Results
Firstly, VA (β = 0.24, p = 0.008), PR (β = 0.31, p < 0.001) and SM (β = 0.39, p < 0.001) positively impact FE. From the perspective of the path that scene features affect PI, VA (β = 0.31, p = 0.009), PR (β = -0.03, p = 0.751), and SM (β = 0.32, p = 0.003). Moreover, FE contributes positively to PI (β = 0.26, p = 0.036). The variances explained for FE and PI are 0.67 and 0.82, respectively. Secondly, FE has a significant mediating effect between scene features (i.e., VA, PR, SM) and PI; the mediating effect accounts are 45.3%, 72%, and 44.4%, respectively. Thirdly, the regression coefficient of interaction term (FE×SI) is 0.001 (t = 0.16, p = 0.987), indicating that the model’s explanatory power is weak.
Discussion
This study totally proposed 11 hypotheses while two were refused. We suggested that sporting goods vendor create an attractive and matching scene to enhance PI. An authentic outdoor environment as the scene can give consumers a more immersive feeling than using a virtual store. PR has no significant effect on PI, which echoes previous conclusion. The consumer is relatively rational and objective, and needs more factors to stimulate the purchase intention. Moreover, a strong sales atmosphere makes consumers who watch e-commerce live streaming to get pleasure may have antipathy. Contrary to prior research in the United States. SI has no significant moderating effect between FE and PI. Consumers with low SI can buy sporting goods for fitness activities. In addition, the main crowd is young people, who buy it for liking sporting trend culture or impulse buying.
References
Zheng, S. et al., (2023). J. Retail. https://doi.org/10.1016/j.jretconser.2022.103240