Examining Platform Strategy for Influencer Marketing Using Text Mining

Social media influencer marketing is widely accepted as an effective approach for companies and brands to reach target consumers and allows advertisers to gather consumers' feedback in real time. There is limited research on the investigation of the efficacy of influencer marketing based on platform strategy, as most literature attributes the success of influencer marketing to that of the source (influencers) and message (content). To fill this gap in research, this study utilizes natural language processing to examine social media users' responses to influencers' advertisements by mining their textual comments on three different major social media platforms: Facebook, YouTube, and X (formerly Twitter). By comparatively analyzing the nature of social media user responses on three different platforms, specifically the evaluation of the advertising messages, varying insights can be gleaned on the efficacy of social media platforms for influencer marketing. The results of sentiment analysis and topic modeling indicate that Facebook yields the most positive responses to advertisements in influencer-generated content as social media users display strong fandom behavior. Moreover, social media users tend to indicate purchase intentions and leave post-purchase reviews on X while forming discussions around the contents of the advertisement on YouTube.