Should a chatbot disclose itself? Implications for an online conversational retailer

Today many consumers prefer interactions with companies via chat and instant messaging, however, although in most cases it is now a virtual agent to handle the interactions, many of them feel it would be eerie if a chatbot pretended to be human. The present study aims at disentangling this sort of ambivalence people have for chatbots through an investigation on how the explicit disclosure of the chatbot identity, before the interaction, influences consumers’ perceptions. Specifically, this study compares the effects that the explicit disclosure of the chatbot identity has on social presence trust and users’ attitudes toward the online retailer. Findings from an online experiment with 160 participants show that interacting with the chatbot whose identity has been primed through a disclosure leads to less perceived social presence, trust, and attitude toward the online retailer, compared to interacting with the chatbot whose identity has not been disclosed before the interaction. The study further analyses a causal chain among the variables, proving that social presence and trust mediate the relationship between the chatbot identity disclosure and the attitude toward the online retailer.