From human to virtual: Unmasking consumer switching intentions to virtual influencers by an integrated fsQCA and NCA method

The growing prominence of Human-Computer Interaction and Artificial Intelligence has propelled the popularity of virtual influencers in social media. Nevertheless, the widespread adoption of the trend hinges on a thorough comprehension of consumers' switching intention from traditional human influencers to virtual influencers as viable alternatives. Given that past studies have primarily concentrated on assessing the overall impact of determinants without delving into the intricacies of behaviours, this study endeavours to develop and substantiate an asymmetric framework incorporating AI technology-like (here, hedonic motivation, expected performance and privacy disclosure concern), human-like (here, anthropomorphism and perceived intelligence), social attributes (here, social influence) and consumer personalities (here, extroversion, introversion and open-mindedness) shedding light on consumers' switching intention to virtual influencers. An online questionnaire was conducted in the United Kingdom and was analysed using an integrated approach that combines fuzzy-set Qualitative Comparative Analysis (fsQCA) and Necessary Conditions Analysis. The results unveiled six configurations of arrangements, each characterised by a unique combination of causation. This study contributes value to virtual influencers and endorsement literature by developing an innovative research method for uncovering the intricacies associated with the switching intention. Furthermore, it provides practical recommendations for social media practitioners on how to optimise the balance between AI technology-like, human-like and social attributes based on consumer personalities, thus enhancing the successful implementation of virtual influencers.