Today I read on the Backtype blog that they were adding profiles to their service to identify influencers on Twitter. I got excited because I like stats and I’m intrigued by how services are being created to try and solve this. I recently wrote about Klout and Peerindex. That post discussed how a users blog and commenting should be factored into influence. It also discussed a scoring problem when a blog is mentioned on Twitter not referencing the user who created it which doesn’t contribute to the calculation. These are several pieces of data that Backtype does collect, so I was excited to hear about their announcement.
Unfortunately, after reading their blog post and visiting my profile, I was pretty disappointed that their influence score and metrics only take Twitter data into account. This has already been done and they have the potential to offer so much more. What a bummer. Hopefully they will incorporate additional data points they are already collecting to modify the scoring algorithm.
In any case it’s still worth checking out your profile on their site as they do offer one interesting graph that provides the percentages of tweets broken down by updates, @replies, retweets, and links. Looking at user profiles and analyzing those 4 data points is pretty interesting. I personally like following many people on Twitter because of the content they share so the links percentage is a good metric for me to determine whether I want to follow new people. Also, after doing some a/b comparison testing on a few people to see their scores on Backtype and Klout, there is definitely scoring differences so I’m curious to what they are. I’ll have to dig in some more.
In any case the influence scoring sector is getting hot and I’m sure more will be joining this area. It should be nice to watch it mature and incorporate more data. Let’s see if Backtype can push the needle further quickly as the race is on.