More Twitter research gives us an insight

Last week I blogged about research by Meeyoung Cha, from Max Planck Institute for Software Systems in Germany, and her colleagues that showed on Twitter, the number of followers you have doesn’t correlate to the influence you have.

Corroborating that is research from Haewoon Kwak, Changhyun Lee, Hosung Park, and Sue Moon from the Department of Computer Science at the Korea Advanced Institute of Science and Technology. According to Shiv Singh, in this second piece of research:

The researchers also analyzed the influence of Twitter users and found that there’s a discrepancy in the relationship between the number of followers and the popularity of someone’s tweets. This basically means that the number of followers is not the only measure of someone’s value.

Singh draw out seven points of interest from the research, some of which are interesting and some of which are blindingly obvious to anyone who’s spent any time on Twitter:

  1. Twitter users have 4.12 degrees of separation on average
  2. The reTweet is powerful
  3. 75% of reTweets happen within an hour of the original Tweet
  4. Followers != influence
  5. Trending topics are mainly news headlines or ‘persistent news’
  6. Only a minority of users have reciprocal relationships, and there are a lot of observers
  7. ReTweets spread quickly

Read the whole post for the Singh’s full analysis.

It’s good to see researchers digging into the nuts and bolts of social media. As I said about Cha’s work, those of us who’ve been in this area for a while have built up through experience and observation a set of instincts about how things work. We use heuristics to get a sense of how the whole system functions, but like any assumption built from personal experience there are risks that we are wrong. So it’s very valuable to have those assumptions tested by research which can then ground us in evidence rather than gut feeling.