Linguistic Strategies and Customer Engagement in Social Media Marketing: A Corpus-Based Analysis

Abstract

This study examines how in modern age the use of digital platforms has manipulated the way of communication between organizations and the consumer (Firat et al., n.d.). Though copious research on the online marketing has been paid a meager attention how the customer engagement via these online platforms. In this paper we investigate linguistic strategies employed in social media marketing and how they affect the involvement of customers is examined. The research involves the analysis of linguistic elements like emotional language, inclusive pronouns, imperative sentences, storytelling, humor, emoji, hashtag and call-to-action expressions, based on theories in persuasive communication and digital discourse (McGuire, 2025). A quantitative corpus-based content analysis is applied which yields authentic samples of social media posts from various leading brands on important platforms such as Instagram, Facebook and X. Linguistic characteristics are systematically coded and quantitatively analyzed with customer engagement characteristics such as likes, comments and shares. Based on statistical analyses such as descriptive, correlation, and multiple regression, the relationship between linguistic strategies and customer engagement was determined. Results have shown that linguistic persuasive elements evoke high involvement thereof significantly; the effect of the individual persuasive elements differs between the measures of involvement. The study also makes a significant contribution to applied linguistics, as it expands the knowledge of how language used in the field of digital marketing, and can prove relevant for digital marketers, content writers and communication practitioners looking into improving the efficiency of their social media marketing efforts. It further serves as a ground for doing further research on marketing content created by artificial intelligence and multilingual communication for digital media (Roy et al., 2015).