Post language and user engagement in online content communities

Publication Type:
Journal Article
European Journal of Marketing, 2016, 50 (5-6), pp. 695 - 723
Issue Date:
Full metadata record
Files in This Item:
Filename Description Size
EJM-12-2014-0785.pdfPublished Version292.39 kB
Adobe PDF
© 2016, © Emerald Group Publishing Limited. PurposeThis study aims to uncover relationships between content communities post language, such as parts of speech, and user engagement. Design/methodology/approachAnalyses of almost 12,000 posts from the content community Reddit are undertaken. First, posts’ titles are subjected to electronic classification and subsequent counting of main parts of speech and other language elements. Then, statistical models are built to examine the relationships between these elements and user engagement, controlling for variables identified in previous research. FindingsThe number of adjectives and nouns, adverbs, pronouns, punctuation (exclamation marks, quotation marks and ellipses), question marks, advisory words (should, shall, must and have to) and complexity indicators that appear in content community posts’ titles relate to post popularity (scores: number of favourable minus unfavourable votes) and number of comments. However, these relationships vary according to the category, for example, text-based categories (e.g. Politics and World News) vs image-based ones (e.g. Pictures). Research limitations/implicationsWhile the relationships uncovered are appealing, this research is correlational, so causality cannot be implied. Practical implicationsAmong other implications, companies may tailor their own content community post titles to match the types of language related to higher user engagement in a particular category. Companies may also provide advice to brand ambassadors on how to make better use of language to increase user engagement. Originality/valueThis paper shows that language features add explained variance to models of online engagement variables, providing significant contribution to both language and social media researchers and practitioners.
Please use this identifier to cite or link to this item: