While trying to get a Spotify invite today, I revisited Klout (@klout), which calculates a reputation score based on your social networks.
If you read here, you might remember I’ve been waiting for someone to reputation and influence scoring for a few years. In fact, back in 2009 before we joined WebCenter (@oraclewebcenter) development, I wrote a functional spec for a reputation engine that we were thinking about building.
As with many of our ideas, we never got around to doing it, but I still think reputation is a huge area for innovation and the inherent game mechanics of scoring make a fascinating subject. This is one of those areas where the enterprise can do better than consumer, thanks to trust. Companies like Spigit (@spigit) are doing cool stuff in this area.
Anyway, I haven’t checked in on Klout since they launched as a Twitter reputation scoring system. At the time, I recall thinking this was a good start, but only supporting Twitter was limiting. Well, apparently they agreed because now they score based on Twitter, Facebook, LinkedIn and Foursquare, and they’ve built out their product quite a bit, including Klout Style and achievements.
The inclusion of LinkedIn is very interesting to me. If Klout can gain some credibility, they could really do some cool things with LinkedIn.
I do wonder how long social networks will allow use of their APIs for Klout though, since these networks themselves stand to benefit from applying their own scoring algorithms to their members.
Things will really get interesting when Google+ releases their APIs, and I hope Klout has plans to add other services, Disqus specifically.
Anyway, they’ve evolved nicely into what could become the standard for influence scoring. Check out Klout and score yourself.
Was the reputation engine that you designed much like the one at http://www.ShareRep.com? They seem to have a different approach on the idea of personal online reputations. More about quality than celebrity.
Probably not since it was an internal tool. Signal to noise isn’t as much an issue inside the corporate firewall, and it’s the first easy metric to measure. Start there, then evolve over time. Participation and generating interest are bigger concerns for enterprises.
That’s interesting. I guess there is a risk in providing too much incentive in a reputation system as it can become more important than the content. Have you seen http://meta.stackoverflow.com/questions/42481/the-problem-with-extrinsic-motivation, quite an interesting discussion.
Interesting thread. I totally agree that extrinsic motivations can cloud participation, but again, enterprise is different bc you’re getting paid. Money is the biggest overriding extrinsic motivator. So, you don’t have to worry about why people are helping as much, but it can get sticky if your intrinsic desire to assist reaches outside your normal duties.
This can be alleviated by creating compensation and recognition that encourages both extrinsic and intrinsic participation.
Just to chime
in, I agree with Jake that ‘Participation and generating interest are bigger
concerns for enterprises.’ There is a really great article on motivation and
using rewards and recognition in order to just get people involved here: http://bit.ly/ea125I Reputation
certainly is another interesting factor in the overall equation, but not
something that will solve all the sorting, filtering, and identifying value all
that much easier for the enterprise.
Reputation can really help though. I’m a fan of throwing together a number (or some other aggregate reputation identifier) and pushing it out there for people to chew on vs. spending a bunch of time on complexity and anti-gaming. Just getting people interested is such a hurdle, the other stuff can come later.
I do think reputation (and game mechanics) can help w sorting and filtering information by creating incentives for the mavens in an enterprise.