A list of the most tweets from Digital Capital Week is making the rounds today.
The list, generated by the Bivings Group and “powered by TwitterSlurp,” does seem to accurately record the volume of tweets authored by individuals, as well as the number of @mentions generated by those tweets.
Over the course of the 10 day tech festival in Washington, there were 12,916 tweets by 2,425 people about Digital Capital Week or on the #DCWeek hashtag.
Well and good.
Unfortunately, these kinds of lists are akin to the measuring the influence of people on Twitter by the number of followers they have.
As Anil Dash put it earlier this year, no one has a million followers on Twitter. The “million follower fallacy” has since been validated by research, confirming the common sense understanding of many long-term observers of Twitter.
Instead of measuring tweet volume, looking at influence as measured by retweets, @mentions and click throughs is useful, along with trickier offline analysis that might include catalyzing people to do things offline. Charlene Li’s tweet that she was heading over to a keynote on open leadership, for instance, motivated some people to come see her speak.
To get a sense of influence, it might be useful to parse the list of “top #DC Week” Twitter accounts through TweetReach.
A rough “back of the envelope calculation” might compare the ratio of tweets to mentions. Pulling from #DCWeek stats and using that ratio, it’s possible to generate a better list of the folks who had social capital during D.C. Week.
Andy Carvin (@acarvin), for instance, “only” tweeted 52 times but had 209 mentions.
Here are some other notable high ratios:
@frankgruber: 115 tweets, 259 mentions
@Jillfoster: 40 tweets, 104 mentions
@dcweek: 234 tweets, 767 mentions
@corbett3000: 96 tweets, 410 mentions
@digitalsista: 31 tweets, 82 mentions
@darthcheeta: 29 tweets, 82 mentions
@mikeschaffer: 33 tweets, 62 mentions
@noreaster: 46 tweets, 137 mentions
That ratio is confounded by the reach of an account, like @jeffpulver. 36 tweets, 462 mentions, but to more than 360,000 followers.
If you took that ratio and factored in reach of the user, it might come closer to reflecting a “top Twitterer” from a given event or #hashtag chat.
Have at it, math geeks.
The bottom line is that we don’t have terrific technological tools to assess the “best tweets” or top Twitterers after the fact, though tools like Twazzup.com can help in the moment.
For those who think it’s all silly, fine. But measuring audience sentiment and journalists’ coverage at events is likely to be something of interest to politicians, businesses and media alike. Here’s hoping that the analysis relies upon more than volume.
5 responses to “Twinfluence: A better measure of social capital at #DCWeek”
Pingback: Twinfluence: A better measure of social capital at #DCWeek (via digiphile) « Chicago Mac/PC Support
Thanks for your write up on this- We originally designed Twitterslurp to be displayed / projected live in the background at events so that attendees (or anyone interested) could follow a live stream of Tweets relating to specific hash-tags or phrases as well as view an archive of all Tweets about an event without maxing out on API calls. Obviously a secondary objective is to encourage a competitive aspect to covering events on Twitter and be recognized on the leaderboard while also giving panels the ability to crowdsource questions and solicit feedback during their presentations.
I agree with your take on influence and that for post event analysis, the focus should be generally be on quality over quantity. For future iterations of Twitterslurp, beyond counting the number of mentions, we are developing customizable metrics to display information about a users’ follower / following ratio, number of times re-tweeted, @replies, use of links, inclusion on lists and so on.
Thanks for the great coverage on technology and government!
Pingback: Twinfluence: A better measure of social capital at #DCWeek « digiphile | D@J
Pingback: How to Improve Your Twinfluence and Twitter Grade | TwidBits
Pingback: You Can Calculate the Value of a Customer Tweet | PrintMediaCentr.com