Virtual Influencing: Uncharted Frontier in the Uncanny Valley

Abstract: Title: Virtual Influencing: Uncharted Frontier in the Uncanny Valley Date of the Seminar: 4th June 2020 Course: BUSN39 Degree Project in Global Marketing Authors: Mary Caroline Creasey and Adrián Vázquez Anido Supervisor: Burak Tunca Keywords: influencer marketing, virtual influencers, social media marketing, opinion leadership, brand spokespeople, brand spokes-characters, source credibility, uncanny valley Thesis purpose: In exploring if and to what extent reality level affects sentiment and skepticism, this study aims to provide valuable and actionable insights to both guide academics in further studies and help managers better understand nuances of influencer perceptions. Methodology: To analyze whether perception differs across influencers’ reality levels (human or virtual), this study collected comment data from Instagram to assess levels of sentiment and skepticism. Additionally, researchers collected data for longevity, engagement rate, following size, influencer gender, and influencer race to examine effects on perception. Theoretical perspective: This study combines Ohanian’s (1990) Source-Credibility Model with Mori’s (1970) Uncanny Valley Effect, slightly modifying each to develop a simple means of evaluating perceptions towards influencers in a social media context. The findings are analyzed under the scope of opinion leadership, branded spokespeople, social media influencers, emotional branding, and para-social interaction. Empirical data: This study is based on secondary data collection. Researchers retrieved publicly available account metrics, influencer demographic information, and post comments from the Instagram accounts of all human and virtual influencers in the sample. Findings/conclusions: The findings reveal that an influencer’s reality level has a strong impact on the level of sentiment and skepticism expressed by their followers. Results also uncover main effects for following size, race, and gender and interaction effects for race. This research contributes to existing literature by providing a new layer to the Source-Credibility Model that enables testing sentiment and skepticism using data from social media. It also confirms the Uncanny Valley Effect’s continued relevance in application to virtual influencers. Practical implications: This paper provides marketing professionals and managers with a broader understanding on the novel concept of virtual influencers. Although leveraging virtual influencers may help brands avoid some of the risks associated with human influencers, virtual influencers may prove problematic as a marketing tactic since the Uncanny Valley Effect can negatively impact how social media users perceive some virtual influencers.