Leveraging Political Ideology and Brand Type: Assessing Consumer Engagement with DEI Initiatives on Social Media with Deep Learning

In recent years, brands and companies are renewing their commitments to diversity, equity, and inclusion (DEI) in response to changing societal demands. Although previous research has shown that brand-generated content on social media plays a significant role in marketing excellence, what constitutes the “right content” remains unknown, especially regarding the impact of DEI-related content on Twitter. Drawing on the congruence theory, this project examines how communicating DEI-related content influences the consumers’ digital engagement with brands’ Twitter accounts. We constructed a dataset consisting of 311 brands and 809,335 tweets collected weekly over the course of one year and employed deep learning algorithms to detect DEI content from tweets and political ideology from brands’ Twitter followers. The results indicate that DEI-related content generally benefits brands’ digital engagement on Twitter. However, this effect is moderated by the political spectrum of their followers and their brand type, defined as goods or service providers. The findings develop brand activism theory and congruence theory in an interactive context and could enlighten marketing practitioners.