Virtual influencers refer to digital agents that imitate the appearance and content of human influencers. While scholars have increasingly examined this novel phenomenon, limited research has assessed consumers’ parasocial interaction with virtual influencers. To the best of our knowledge, no research has examined whether parasocial interaction differs for sponsored and non-sponsored content. Guided by the persuasion knowledge model (PKM), the present study addresses these gaps by examining consumers’ parasocial interactions with Miquela (@lilmiquela), a prominent virtual influencer with three million Instagram followers. Using computational methods, we extracted 276 posts from Miquela’s Instagram profile and 16,876 English-language user comments. Sentiment analyses of text and emojis indicated a positivity bias for Miquela’s Instagram posts and user comments, regardless of sponsorship disclosure. However, the sponsored content possessed significantly more positive sentiment than the non-sponsored content across Miquela’s Instagram posts and user comments. Topic modeling revealed that consumers were confused about Miquela’s identity as a robot or human. Consumers also provided polarizing emotional responses: Some proclaimed their love for Miquela and complimented her beauty. Yet, others reacted with fear and psychological reactance. Overall, our findings offer valuable insights for brands that intend to harness the power of virtual influencers.
|Published - 24 Mar 2023
|American Academy of Advertising 2023 Annual Conference - Curtis Hotel, Denver, United States
Duration: 23 Mar 2023 → 26 Mar 2023
|American Academy of Advertising 2023 Annual Conference
|23/03/23 → 26/03/23