Study—Everyone’s an Influencer: Quantifying Influence on Twitter

At the recent IPR Measurement Commission meeting Duncan Watts (see below) presented results of a study of influence among Twitter users by Jake Hofman, Winter Mason, Duncan Watts, and Eytan Bakshy called “Everyone’s an Influencer: Quantifying Influence on Twitter.” The study investigates 90 million posts by 1.6 million users.  Download it here.

From the abstract:

In this paper we investigate the attributes and relative influence of 1.6M Twitter users by tracking 74 million diffusion events... 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... 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... 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.

Also, Duncan Watts has a new book on sale March 29, Everything Is Obvious: *Once You Know the Answer.  Here is what Alan Alda says about it: “You have to take notice when common sense, the bedrock thing we’ve always counted on, is challenged brilliantly. Especially when something better than common sense is suggested.”

--WTP