Everyone’s an influencer: quantifying influence on twitter
Proceedings of the fourth ACM international conference on Web search and data mining |
Organized by ACM
In this paper we investigate the attributes and relative inﬂuence of 1.6M Twitter users by tracking 74 million diﬀusion events that took place on the Twitter follower graph over a two month interval in 2009. Unsurprisingly, we ﬁnd that the largest cascades tend to be generated by users who have been inﬂuential in the past and who have a large number of followers. We also ﬁnd that URLs that were rated more interesting and/or elicited more positive feelings by workers on Mechanical Turk were more likely to spread. In spite of these intuitive results, however, we ﬁnd that predictions of which particular user or URL will generate large cascades are relatively unreliable. We conclude, therefore, that word-of-mouth diﬀusion can only be harnessed reliably by targeting large numbers of potential inﬂuencers, thereby capturing average eﬀects. Finally, we consider a family of hypothetical marketing strategies, deﬁned by the relative cost of identifying versus compensating potential “inﬂuencers.” We ﬁnd that although under some circumstances, the most inﬂuential users are also the most cost-eﬀective, under a wide range of plausible assumptions the most cost-eﬀective performance can be realized using “ordinary inﬂuencers”— individuals who exert average or even less-than-average inﬂuence.