Virtual Influencers (VIs) are computer-generated characters that imitate human behavior. Especially highly realistic human-like VIs blur the line between human and non-human appearances. Many VIs secured advertising partnerships with major brands due to advantages like control, risk-free brand-consistent storytelling, scalability, and cost-efficiency. This study investigates the impact of different types of Instagram postings by VIs on user sentiment and engagement. Twelve VIs were selected based on pre-defined criteria. The analysis comprised three main steps: First, posts were categorized into content themes. Second, we assessed agency and identity using Llama3.1-405B. User reactions were then analyzed for engagement and sentiment distributions. Emotion classification was conducted by assigning comments to Ekman’s six basic emotions. Keyword frequency analysis identified prominent terms associated with specific sentiments. The results of the sentiment analysis indicate that visually engaging content themes, such as fashion and travel-related posts, tend to elicit more positive reactions. In contrast, personality-driven content themes, particularly posts related to family, are more likely to evoke lower sentiment and negative emotions. Moreover, posts emphasizing agency or identity do not necessarily enhance user engagement or sentiment. The study lays the groundwork for examining how digital personas influence consumer behavior. Our findings have implications for digital marketing strategies, highlighting the effectiveness of visually compelling content while underscoring the need for a cautious approach to personality-driven themes.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Content Themes and Self-presentation of Virtual Influencers - Exploring the Impact on User Engagement and Sentiment

  • Laura Tölle,
  • Torben Rehmer,
  • Fynn Ruppel,
  • Johann Müller,
  • Matthias Trier

摘要

Virtual Influencers (VIs) are computer-generated characters that imitate human behavior. Especially highly realistic human-like VIs blur the line between human and non-human appearances. Many VIs secured advertising partnerships with major brands due to advantages like control, risk-free brand-consistent storytelling, scalability, and cost-efficiency. This study investigates the impact of different types of Instagram postings by VIs on user sentiment and engagement. Twelve VIs were selected based on pre-defined criteria. The analysis comprised three main steps: First, posts were categorized into content themes. Second, we assessed agency and identity using Llama3.1-405B. User reactions were then analyzed for engagement and sentiment distributions. Emotion classification was conducted by assigning comments to Ekman’s six basic emotions. Keyword frequency analysis identified prominent terms associated with specific sentiments. The results of the sentiment analysis indicate that visually engaging content themes, such as fashion and travel-related posts, tend to elicit more positive reactions. In contrast, personality-driven content themes, particularly posts related to family, are more likely to evoke lower sentiment and negative emotions. Moreover, posts emphasizing agency or identity do not necessarily enhance user engagement or sentiment. The study lays the groundwork for examining how digital personas influence consumer behavior. Our findings have implications for digital marketing strategies, highlighting the effectiveness of visually compelling content while underscoring the need for a cautious approach to personality-driven themes.