I’ve been fascinated by the Klout score for weeks; I don’t put too terribly much stock in it as a valid measuring tool at this point, but it’s becoming the Next Big Thing in social media and, as such, it warrants some attention. I’m more fascinated at this point by how the score is calculated — what formula Klout is using behind the scenes to come up with the actual number, and what sort of factors influence changes in the score.
My own score has fluctuated wildly. One week ago today, on Sept. 7, I hit an all-time high of 59 (58.96, rounded up), and today, I have hit what I think is an all-time low at 54.79 (although I’ve been at the rounded-to score of 55 on several previous occasions – the 54.79 raw score I think is a low since I started keeping track) . That’s a swing of seven and a half percent in one week, and my activities aren’t appreciably different. So I wondered what led to the changes.
Here’s my graph from klout.com, showing changes in my score from Aug. 17 (on the far left) to today, that 54.79 on the far right:
I identified what I think were the two leading factors to the big changes at points A and B.
For the huge spike upward at Point A, leading to my all-time high score of 59, in the previous day I had picked up around 25 new followers in one day – an enormous number for me. As of this posting I have 225 total followers, so that was a one-day boost of more than 10 percent in my follower count. What Klout didn’t pick up on, and couldn’t *possibly* pick up on, since it’s an automated process, is that 23 of those “new followers” were different specialized accounts from MinnesotaJobs.com — accounts listing admin jobs, engineering jobs, manufacturing jobs, etc. All splinter accounts from essentially one new follower, and therefore basically useless as a measure of “influence.” Still, the follower increase led to a huge spike. And then, since that wasn’t sustainable given my previous history, since I didn’t have a similar boost the next day my score plummeted three points, back to 56 (which is where I seem to be leveling out)
For today’s drop at Point B, where I lost almost an entire point in one day, two of my three metrics are up – “Network Influence” has climbed from 54 to 55 (rounded values) and my “Amplification Probability” is up about a quarter of a point, although the displayed value is still 30. However, for some reason my “True Reach” score took a four-point nosedive, my biggest single-day drop recorded on the graph (so, dating back about a month), falling from 216 to 212. So it seems like a safe bet that the four-point drop in reach led to today’s drop.
Again, it’s hard to know how much stock to put in these Klout scores. But I sure am fascinated by trying to figure out how the formula works.