Talk to me: how interactive AI chatbots orchestrate customer decision satisfaction through conversational human-likeness and cognitive comfort

Purpose This study investigates the interplay of artificial intelligence (AI) chatbot interactivity, conversational human-likeness, cognitive comfort and customer decision satisfaction in an emerging B2C e-commerce segment. Design/methodology/approach A field survey questionnaire was used to collect data from 315 customers who use e-commerce platforms for online shopping, employing a nonprobability purposive sampling technique. The data was analyzed through the partial least squares structural equation modeling technique using Smart PLS 4.1.1.6. Findings Results show that AI chatbot interactivity positively influences conversational human-likeness, and conversational human-likeness positively influences customer decision satisfaction. Conversational human-likeness fully mediates the association between AI chatbot interactivity and customer decision satisfaction, with a strong moderating effect of cognitive comfort on the relationship between conversational human-likeness and customer decision satisfaction. Originality/value The research adds contextual value by compiling a distinctive dataset from customers of B2C e-commerce platforms in a developing nation. Theoretically, the model has been conceptualized through stimulus-organism response theory by endorsing conversational human-likeness as an intervening mechanism and cognitive comfort as a boundary condition.