I’m currently involved in a project to support Hill & Knowlton’s sponsorship of the COP15 United Nations Climate Change Conference next month, which involves analysing the different media, influencers and topics driving the conversation.
One of the things we are doing is measuring the most prominent tweeters (the COP15 Twitterati – latest table here and Twitter list here). We’re doing this purely by volume of tweets mentioning specific keywords (translated into about eight languages), so there’s no magic pixie dust until we get to the job of semantic analysis to extract the people and topics being referenced.
However, I thought it might also be interesting to look at the correlation between the volume of tweets from our top 15 and their respective influence, as measured by Edelman’s TweetLevel methodology. The scatter plot appears below (click for the readable version), but in summary:
- There’s a positive correlation between volume and influence at 0.33, albeit a weak one
- The r2 value is 0.11, suggesting that around 11% of influence is attributable to volume
I intend to explore this further. I’ve long thought that online influence measures are fairly ropey, and it would be interesting to see exactly how much volume really is the only useful measure. We keep banging on about quality being more important than quantity online, but is that really so?