How do the attributes of AI-generated digital humans relate to online customer experience? Evidence from SEM and fsQCA

Abstract

This study investigates how AI-generated digital human attributes relate to online customer experience, identifies the configurational pathways linked to shared online customer experience, extends the research boundaries of generative AI in marketing, and expands the body of literature concerning online customer experience. A quantitative online questionnaire was distributed to consumers with AI digital human interaction experience, collecting 607 valid responses for subsequent statistical analysis. This study draws on affordance theory to conceptualise four key attributes of AI-generated digital humans. A mixed-analytical design integrating structural equation modelling (SEM) and fuzzy-set qualitative comparative analysis (fsQCA) is adopted to examine variable associations, mediating linkages, and multi-factor configurational pathways. SEM results indicate that intelligent service capability and social influence of AI-generated digital humans are positively associated with both solo and shared online customer experiences, while sensory anthropomorphism is positively associated only with solo online customer experience. Furthermore, fsQCA results reveal three equifinal configurational pathways associated with high shared online customer experience. Solo online customer experience, intelligent service capability and social influence all serve as core conditions, while sensory anthropomorphism functions merely as a peripheral condition. This study investigates the associations between attributes of AI-generated digital humans and online customer experience outcomes, deepening the application of affordance theory in intelligent marketing contexts. The analysis delivers rigorous empirical evidence and actionable managerial insights to guide firms in refining AI-generated digital human design and deployment strategies and enhancing online customer experience quality.