Drivers of Social Media Rebroadcasting: Investigating the Role of Message Content and Influencers

Abstract

We investigate the drivers of social media rebroadcasting behavior by jointly examining the effects of message content, user heterogeneity (both observed and unobserved), the fit between message content and individual users, and the influence of prior rebroadcasting activity. We focus on the Twitter postings from the top 10 business schools as identified by Businessweek and model the subsequent rebroadcasts (or retweets) of these messages as a split hazard process. Our results show that not only does rebroadcasting activity vary with the content of the original message but also that individuals are more likely to rebroadcast content that closely fits with their own interests. We then use a series of simulations, we show how the effects of different seeding strategies, where specific users are targeted, impact the rate with which a social media message spreads. We show that tailoring message content to match the interests of influential users can increase the rebroadcasting of the message, but such a strategy has its limits as the tastes of influential users may not necessarily match those of the broader user base.

Keywords: Social Media, Hazard Model, Text Analysis