Abstract
Understanding the microeconomic mechanisms through which digital communication alters consumer demand for environmentally sustainable goods is critical for contemporary market design and sustainability policy. This study investigates the causal impact of social media influencer (SMI) endorsements on consumer preferences, willingness to pay, and aggregate sales for sustainable products. Combining large-scale natural language processing (NLP) and text mining of over 1.2 million influencer posts across visual and microblogging platforms with a high-frequency consumer transaction panel covering fast-moving consumer goods (FMCG) and apparel, we deploy a staggered difference-in-differences econometric framework augmented by instrumental variables. Our text mining pipeline decomposes influencer content into discrete semantic dimensions, including informational transparency, emotional valence, and perceived authenticity. Empirical findings reveal that influencer endorsements induce an average causal sales lift of 8.4% for eco-certified products, with the magnitude heavily moderated by content framing. Specifically, technical transparency and supply chain traceability narratives generate statistically significant reductions in price sensitivity, whereas vague ecological claims yield negligible or negative effects. Furthermore, micro-influencers exhibit higher conversion efficiencies per follower unit compared to macro-celebrities due to perceived source credibility. These insights bridge marketing strategy and behavioral economics, providing empirical guidance for green brand positioning, platform governance, and anti-greenwashing regulatory interventions.
