Media as a Double-Edged Sword: How ESG Discourse Shapes Corporate Image in Western Markets with a Case Study of Unilever

Abstract

Amid growing global emphasis on sustainable development, ESG (Environmental, Social, and Governance) has become a focal point in both business and public discourse. This study employs corpus-based discourse analysis to investigate how Anglo-American media discourse shapes the Western public’s perceptions of corporate ESG strategies, based on the case study of Unilever in Western markets. Specifically, it adopts a diachronic approach to examine Unilever’s corporate image during three critical phases of its ESG strategy implementation: the baseline period (2008-2010), implementation period (2011-2020), and current period (2024-2025). The research methodology involves constructing two specialized corpora - a media corpus drawn from English-language sources including LexisNexis news reports and a public perception corpus comprising Google Trends search data and Trustpilot consumer reviews, providing a lens into the dynamics of ESG communication in major Western markets. This dual-corpus design enables systematic analysis of three key dimensions: (1) media’s role in amplifying ESG adoption, (2) agenda-setting effects on public attention, and (3) public’s attitudinal polarization towards Unilever brand. Findings reveal that media initially promotes ESG initiatives but later scrutinizes implementation gaps, creating cyclical trust-building and erosion patterns. A medium correlation between media coverage and public searches confirm agenda-setting effects, while polarized consumer ratings demonstrate how evaluative language influences credibility judgments. The study enriches ESG communication literature by identifying distinct phases of media influence—from amplification to skepticism—and highlights the evolving dynamics between corporate messaging and stakeholder expectations. Practical insights are offered for improving ESG reporting to better align with public perceptions and mitigate reputational risks.