Purpose Consumer engagement behavior (CEB) is considered to be closely related to the sales performance of e-commerce live-streaming and the high-level CEB is likely to bring revenues to enterprises. The impact of CEB on live streaming sales performance is often considered to be unidirectional and static. However, CEB is an ever-changing process. Few studies have considered how CEB affects the sales performance in live-streaming from a dynamic perspective. This article examines the role of CEB on sales performance in live-streaming. Design/methodology/approach We collected the minute-level data from TikTok to verify the possibility of like behavior, comment behavior, sales performance and online viewers influencing each other and estimated by the panel vector auto regression model. Our original data set is obtained by downloading live videos on TikTok platform through a web crawler after processing, which contains 3274 live streams in the top 100 from August to September 2023. After data collation, the final data set included 6,945,013 data groups from 2,876 live streams. Findings First, consumer engagement promotes sales performance and we proved that the two parties are mutually influencing for the first time. Second, we find that sales performance is not only affected by the current period of consumer engagement, but also by the lag effect. This means that the relationship between two variables is dynamic. Third, we find that there is also a bidirectional effect between different behaviors. And “comments” have a more significant impact on “likes” and last longer. Finally, the fans number, the popularity and the influencer host can promote the mutual promotion between CEB and sales performance. Originality/value First, we propose that there is a dynamic and bidirectional relationship between CEB and sales performance, revealing the process of achieving high-level sales performance. Second, we discuss causal relationships between different consumer engagement behaviors. We treat multiple behaviors as independent variables and discuss their combined impact on another variable. Third, we demonstrate that the impact and duration of high-interaction-intensity are significantly higher than low-interaction-intensity behaviors. Finally, our results prove the importance of live-streaming characteristics in influencing CEB, such as influencer hosts, high fans and popularity, providing new insights into the promotion of consumer participation in different live rooms.
Bolun et al. (2026) studied this question.