How streamer’s follower count affect consumers’ immediate purchase behavior: the moderating effect of context collapse

Purpose This study aims to identify how streamer follower count influences consumers’ immediate purchase behavior in live commerce and to clarify the role of streamer conduct by introducing the concept of context collapse. Design/methodology/approach We first conducted grounded theory interviews to develop a conceptual framework linking follower count, predictive expectation and immediate purchase behavior. We then executed two scenario-based online experiments using Taobao Live simulations to test hypotheses regarding (1) nonlinear effects of follower count, (2) mediation by predictive expectation and (3) moderation by perceived context collapse (collusion vs conflict). Findings Results demonstrate an inverted U-shaped relationship between follower count and both immediate purchase behavior and predictive expectation; predictive expectation significantly mediates the former. Moreover, when streamer behavior induces perceived context conflict rather than collusion, these inverted U-shaped curves become markedly steeper. Medium streamers (around one million followers) maximize impulse-buying responses. Originality/value By integrating social impact theory with context collapse, this research advances understanding of nonlinear effects in live commerce partnerships and illuminates the differential impacts of streamer conduct. The findings inform e-commerce companies’ streamer selection strategies and extend the theoretical application of context collapse to digital shopping contexts.