Be smart, human and empathic: understanding the role of AI chatbots in fostering brand preference

Purpose The study develops and tests a framework to investigate the artificial intelligence (AI) chatbots’ communication qualities and human-like characteristics on perceived empathy, which translates into service authenticity and empowerment, and eventually brand preference. Design/methodology/approach In study 1, a combined Importance–Performance Map Analysis (IPMA) and fuzzy set qualitative comparative analysis (fsQCA) was used to analyse the data. Study 2 is a follow-up qualitative study using in-depth interviews. Findings The findings reveal that perceived anthropomorphism and perceived intelligence are the two strongest predictors of perceived empathy. Perceived empathy positively influences service authenticity and consumer empowerment. The cIPMA results further highlight a performance gap in perceived consumer empowerment and empathy. The fsQCA revealed five distinct configurations that foster chatbot empathy. The qualitative findings validate the quantitative results and reveal additional themes that warrant consideration. Originality/value This study provides insight into how key AI chatbot attributes influence brand preference, advancing our understanding of the consumer behaviour mechanisms that drive perceived empathy and overall interaction effectiveness in human–AI communication.