Effects of Consumer Loneliness on Chatbot Recommendation Intention: Moderating Effects of Chatbot Conversation Style

Abstract

Anthropomorphizing chatbots can produce more effective consumer interactions owing to the human warmth provided by the chatbot. This study expands the scope of previous studies by examining the influence of loneliness, a personal consumer characteristic, on the acceptance intention of a chatbot recommendation. Additionally, the moderating effect of chatbot conversation style (warmth vs. competence) on this relationship was examined. To verify this hypothesis, we conducted two pretests and one laboratory experiment. The results of the analysis indicate that the higher the degree of loneliness, the stronger the tendency to accept chatbot recommendations (Hypothesis 1). Particularly, when the chatbot interacts in a warm manner, consumers who are experiencing loneliness are more likely to accept chatbot recommendations. However, when the chatbot interacts competently, the degree of loneliness does not affect the acceptance intention of the chatbot recommendation (Hypothesis 2). This finding contributes to the literature on chatbot anthropomorphism and provides practical information for chatbot conversation decisions in relation to consumer characteristics.