Behind Chatbot Dialogues: The Power of Message Length Similarity and Its Effect on Attitudes Toward Chatbot

This research explores how message length similarity (MLS) between customers and chatbots influences user attitudes, social presence, and interaction dynamics. Grounded in communication accommodation theory (CAT), we hypothesize that MLS serves as a human-like cue, fostering social presence and improving attitudes toward chatbots. Across two experimental studies (N = 267), we find that MLS enhances attitudes, with social presence mediating this effect. Additionally, MLS is particularly effective in fostering positive attitudes under heightened customer anger, demonstrating its potential in emotionally charged interactions. These findings highlight MLS as a practical and impactful design principle for chatbots, mimicking human conversational norms to improve psychological and attitudinal outcomes. The results provide actionable insights for developing adaptive, emotionally intelligent chatbots capable of optimizing user experiences across diverse emotional and situational contexts. By emphasizing conversational alignment, this work bridges theoretical understanding and practical applications, offering a foundation for future innovations in chatbot design.