WHEN VOICE MEETS AVATAR: HOW CONVERSATIONAL STYLE AND AVATAR TYPE SHAPE CUSTOMER RESPONSES IN SMART COMMERCE

Abstract

Purpose of the study: AI voice avatar assistants are emerging as socially interactive service interfaces in smart commerce. Drawing on the stereotype content model, this research investigates how conversational style and avatar type shape consumers’ willingness to use AI devices through perceived warmth and perceived competence. Methodology: To comprehensively examine the effects of AI voice avatar assistants, two studies were conducted. We recruited participants from Credamo to conduct two experiments (Study 1=259; Study 2=250) were conducted using an online survey and to test a moderated mediated model using SPSS PROCESS macro-models for analysis. Main Findings: Study 1 shows that task-oriented conversational style increases consumers’ willingness to use AI devices, and that this effect is mediated by both perceived warmth and perceived competence. Study 2 reveals a significant interaction between conversational style and avatar type: social-oriented language enhances willingness to use AI devices when paired with realistic AI avatars, whereas this effect does not emerge for cartoon avatars. The moderated mediation results further demonstrate that perceived warmth and perceived competence explain this interaction effect. Applications of this study: These findings contribute to the understanding of customer acceptance of AI voice avatar assistants by highlighting the influence of conversational style and avatar type, and offer practical implications for companies to enhance the effectiveness of AI voice avatar assistants through conversational style-matching strategies. Novelty/Originality of this study: This research contributes to AI service and smart commerce literature by identifying warmth and competence as key psychological mechanisms through which consumers evaluate AI voice avatar assistants.