Sincere or Skeptical? The Role of Chatbot Social Presence in Driving Propensity to Purchase via Chatbot

AI-powered conversational agents (chatbots) have become pivotal for shaping user trust and effective human-AI interaction. However, a key barrier to adoption is user skepticism toward AI. Grounded in social presence theory, this study investigates the interplay between a chatbot’s level of social presence and user skepticism in driving behavioral intentions during online interactions. Through a multi-study experimental design (Total N = 661; Study 1 N = 162, Study 2 N = 195, Study 3 N = 304), we measured participants’ experiences via recalled interactions with a conversational agent across three product categories representing involvement levels. Results confirm that chatbot social presence and perceived sincerity drive purchase propensity. Crucially, the findings reveal a nuanced “double-edged sword” effect: while social presence generally enhances the experience, it can reduce behavioral intention when users exhibit high AI skepticism. Theoretically, this study highlights a key boundary condition for social presence theory. This study provides actionable guidance for designers of conversational agents, arguing against a one-size-fits-all approach.