Virtual or human influencers as endorsers? Behavioral and EEG evidence of how influencer type affects purchase intention of new products

Abstract

Virtual influencers are increasingly popular, drawing attention to their endorsement effectiveness. This study explores the effectiveness of virtual versus human influencers in endorsing new products, specifically Really New Products (RNPs) and Incrementally New Products (INPs) via two online and one EEG experiment. Behavioral results indicate that virtual influencers increase purchase intentions for RNPs, while human influencers do so for INPs. We propose a moderated mediation model positioning perceived risk as a mediator and new product type as a moderator. EEG data show that, at the early subconscious stage, virtual influencer endorsements (vs. human) for INPs required higher cognitive control (larger N200). At the later conscious stage, virtual influencer endorsement (vs. human) elicited lower emotional engagement (smaller P300) for INPs but higher emotional engagement (larger P300) for RNPs. A positive link between purchase intention and P300 amplitude indicates that heightened emotional engagement boosts purchase intentions. Using Construal Level Theory (CLT), we explain that virtual influencers, promoting an abstract mindset, effectively reduce perceived risk for RNPs, while human influencers, inducing a concrete mindset, lower perceived risk for INPs. This alignment of the construal levels between influencer type and product type reduces cognitive control and increases emotional engagement. CLT hasn't explained influencer type effects across product newness; this study extends CLT to virtual endorsements, confirming psychological-distance principles apply equally. These findings underscore the importance of aligning influencer type with new product type, deepening our understanding of virtual influencer endorsement and offering practical insights into their use for promoting new products.