Beyond human streamers: how AI-avatar engagement strategies influence consumer acceptance

Purpose This paper aims to investigate the effects of various engagement strategies employed by Ai-avatar live streamer (ALS) on consumer acceptance intentions. Design/methodology/approach Using survey data from 3,066 respondents in China, this paper employs correlation analysis and hierarchical regression techniques to quantify the moderating effects of ALS’ engagement strategies – visual impact, narrative framing (NF) and interactive engagement. Findings The results indicate that visual attractiveness plays a notable but adverse moderating role in acceptance intentions via subjective norms (SN). This indicates that consumers’ acceptance of the ALS is driven more by personal attraction to the avatar’s beauty than by social norms or group expectations. Conversely, visual influence positively moderates purchasing intentions through perceived behavioral control, suggesting that appealing appearances enhance consumers’ sense of ease and comfort in engaging with ALS. Moreover, the NF of ALS negatively moderates SN, suggesting that even if the social environment encourages engagement with AI-avatar live streamers, the current unrealistic or inauthentic narrative prevents consumers from accepting it. Originality/value Firstly, this paper provides a quantitative and comparative evaluation of mainstream ALS tactics on consumer acceptance intentions, which addresses a gap in existing literature. Secondly, this paper addresses a significant gap in existing studies by exploring how big-data recommendations implemented by ALS affect the consumer behavior. Thirdly, this study extends the theory of planned behavior (TPB) to the artificial intelligence live commerce context.