Are we who we follow? Computationally analyzing human personality and brand following on Twitter

In this study, we present a computational method for analyzing the congruence between personality of a brand's Twitter account and the personality of their followers. We investigated attachment to brands on Twitter by computing personalities through a machine-learned computational analysis of Twitter postings rather than traditional personality tests. By studying three different brands, results revealed that on average, brand followers have personalities that are more congruent with the personality of brands they follow compared to users that do not follow those brands. Taking these findings into consideration, we discuss some considerations for advertising researchers and practitioners, as well as provide a new tool, the Brand Analytics Environment (BAE), to allow individuals without computer programming backgrounds to conduct this method themselves.