Targeting Opinion Leaders in the Twittersphere

In less than a decade, social media (SNS) have become relevant platforms in electoral campaigns, by providing the means to develop new communication strategies, enhancing political engagement, fundraising, or fueling political discussion. Although the literature has portrayed most of these platforms as communication spaces in which candidates can establish horizontal interactions with voters, maintaining a credible conversation with an electorate of several thousand people would be too costly and time-consuming for politicians. Evidence shows that candidates follow, on average, a considerably smaller number of users on their SNS than followers they account. Their replies to followers' mentions are even fewer. Hence, the relationship is asymmetrical and candidates select whom they interact with. Following two-step communication flow theories, we argue that candidates will target users able to enhance the spread of their message beyond the audience in their own networks. We employ a novel dataset of 25M tweets from the 2016 Primaries Elections in the U.S. to identify and categorize online opinion leaders, and analyze how US presidential candidates targeted them in the 2016 Primaries.