This study investigates how the linguistic style of CEO digital communication influences audience engagement. Using an NLP pipeline with a panel regression model on a data set of 19,566 tweets from CEOs, this study reveals that linguistic clarity and an on-platform focus are the most robust predictors of engagement; syntactic complexity and the inclusion of external URLs consistently deter engagement metrics. The effects of stylistic choices like emojis and hashtags are less consistent and depend on the type of engagement being measured. These results offer an expanded understanding of digital communication for CEOs and provide direct implications for business communication pedagogy.