A few weeks ago, I saw a fairly steep drop in my Klout score and wondered about it for a few minutes. I chalked it up to an algorithm change – something they are not shy about doing.
However, I’ve been live-Tweeting the 2011 Inbound Marketing Summit for the past few days and was curious to see how it affected my score.
Obviously, the score was going to go up due to the unusually high number of updates and re-Tweets. But then, looking at the time graph it occurred to me to ask, “When exactly was WordCamp Boston?” That happened to be the last live-Tweeting event I attended.
As it turns out, the steep drop in my Klout score occurred exactly 30 days after WordCamp Boston:
It’s not terribly meaningful, but it does provide some insight into the Klout algorithm. Apparently, there the Network Influence and Amplification Probability calculations are performed over a 30 day moving window. It further seems to be a fairly simple calculation that is not using any sort of moving average, but just a basic summation.