"Re-adding this as it appears to have gone missing: This is very similar thinking to the post I wrote yesterday. I have said for a while that influence cannot be accurately measured externally to a service and that the host (FB, Twitter, Google, etc.) is best placed to calculate that due to (theoretically) having access to all data. This, however, means that any influence score or rating will be "service specific" - perhaps that's not such a bad thing. Trying to use one score to rate out influence across the entire social web is crazy. We all know that follower numbers is a rubbish stat. Klout tries to negate to negate its effect by actually measuring the ratio of engagement to followers to see what fraction of your audience is actually interacting with you. It's part of the way there but not entirely so. If 40% of users on Twitter only consume then that's a lot of "engagement" unaccounted for - no replies, no retweets, no favourites. Wrapping everything in t.co addresses, however,..." - Colin Walker