(adapted from my answer in Quora)
I see them fulfilling uses cases such as:
– I search for “Rush” to find what Twitterers are the most authoritative about this awesome Canadian band;
– Rush management wants to promote their new tour and need to find who are the super fans to spread the word.
When people talk about PageRank; about figuring out that Page A is important because loads of other pages are linked to it (I know other signals are considered); they take for granted that Google knows what topic Page A covers. So when you search for “Rush” you get a result of the most authoritative page about the band.
In the page-centric web world, where the node is the page, we’ve all been trained to write super topic-focussed, use key words and provide meta-tags. Even if the page refreshes a bit, which it should to preserve it’s SEO ranking, we will keep it on topic.
When it comes to the people-centric world, where the node can be a Twitter account (usually a person) we are a mess. It’s easy to determine that an account has more followers (and other signals of popularity) but it’s nigh impossible, so far, to figure out why. The regular account will not only tweet about a topic they are authoritative but also about sports results, good movies, funny YouTube clips and many other subject matters that interest them. The same applies to Facebook updates, Google+ postings, etc.
PeerIndex is trying to figure what a Twitter account is authoritative about by analysing the content of their tweets (and other social media content). In my personal experience it doesn’t work (yet)–what they think of me is completely incoherent.
Klout appears to have opted for a crowd sourced solution with +K. They identify what an account might be about but then let loyal registered users (those that login at least once a day) to vote up their best guesses.
Prediction: I think these two players will be snapped up by one of the big boys: Twitter, Microsoft, Google or LinkedIn within two years.