A comparative analysis of followers' engagements on bilingual tweets using regression-text mining approach. A case of Tanzanian-based airlines

Abstract

Business entities utilize multiple languages on social media for marketing, promotions, and communication. The impact of utilizing one language over the other on engagements has not been well explored. This study applied a regression-text mining approach to over 3000 Tanzanian-based airlines' tweets posted between 2018 and 2020 to explore the influence of English and Swahili languages in communication and marketing. By defining engagement as the retweets and favorites, the study found that English tweets had relatively higher engagements than Swahili tweets. However, Swahili tweets with photos and videos were likely to have more engagements than English tweets. Conversely, hashtags attracted higher engagements for English tweets. Time of the day revealed mixed findings. Further, several key patterns were observed when favorites and retweets were considered separately. The practical applications of the study were also discussed. It is expected that the study findings will benefit bilingual business entities on a global scale.