Talking to a bot or a wall? How chatbots vs. human agents affect anticipated communication quality

Chatbots have been applied to computer-mediated communication to replace human agents due to their high efficiency and cost-effectiveness. However, their outcomes are not always desirable, and limited guidance exists on how chatbots impact users' perceptions of the communication process. Drawing on cue-filtered-out theories and multiple resource theory, this study investigated the impacts of the communicating agent (chatbot vs. human agent) on anticipated communication quality and the underlying mechanism. Two experimental studies revealed that users' anticipated communication quality was lower with chatbots than with human agents due to the serial mediating role of self-focused attention and user empathy. Moreover, moderation analyses found that a multiple-choice communication strategy for chatbots can enhance users' anticipated communication quality by impacting self-focused attention compared to an open-ended strategy. The findings contribute to the knowledge of the processes driving users’ anticipated communication quality toward chatbot applications and offer managerial implications for chatbots and communication strategies.