Designing chatbots for E-commerce engagement: A multi-study approach

This study investigates how anthropomorphism design cues in shopping chatbots influence consumers’ perceptions and usage intentions on online shopping platforms. Using a sequential mixed-methods design, qualitative interviews identified two key dimensions of chatbot anthropomorphism: interactive personalization cues and conversational emotional cues. Three controlled experiments with Vietnamese online shoppers tested their effects on perceived shopping security, perceived recommendation quality, shopping enjoyment, and subsequent usage intentions. Findings reveal that emotional message cues significantly enhance consumer perceptions and intentions, while personalization cues alone have a limited impact. Moreover, the interaction between the two cue types was insignificant. Mediation analysis confirmed that perceived security, recommendation quality, and enjoyment transmit the effects of emotional cues on adoption intention. These results highlight the importance of designing emotionally expressive chatbots to foster trust, satisfaction, and continued engagement in online shopping contexts.