Tom Webster, writing and speaking

Should Klout Scores Be "Stickier?"

Added on by Tom Webster.

After reading my friend Mark Schaefer's latest provocative post about Klout, "The Making of a Social Media Slut," I had a good, long look at my Klout score. It's not bad. I'm no Jay Baer, but my score is higher than that of former Family Ties star and childhood crush Justine Bateman, so I'm doing OK, I guess. While looking at some of the various components of my score, though, I was reminded of something that baseball scouts know as gospel - once you demonstrate a skill, you "own" that skill. In other words, if you hit 25 home runs in a season, people can say it is a fluke, but they cannot say you aren't capable of hitting 25 home runs. After all, you did. You own that skill, and forecasters can't treat yesterdays fluke as tomorrow's impossibility. So, performance indicators tend to be a little "sticky" - even if the previous performance is considered to be a fluke, the projections won't fall too precipitously far from the bar that the player has, indeed, demonstrably set.

So, with that potentially irrelevant preamble aside, I pose this question: Why do Klout scores decline in the short term? In other words, consider my graph:


This represents my Klout score from one month ago (June 21st, 2011) to today. I started off at 66, and hung there for a while. By July 9th, I had fallen to 57 - a decline of 14%. Not an insignificant decline. Turns out, I was on vacation prior to that - I spent the week of July 4th seriously off the grid, and unleashed my withering torrent of tweets to a lake full of unreceptive fish rather than the Internet. So, I dropped, but I didn't drop because my ability to influence eroded. It dropped because I wasn't active.

It's now back to 61. I'm trying, folks, I'm trying. But if Oprah stopped tweeting for a year, and then uncorked a single "click here to find out about my new TV show" link, I think you'd see that her actual influence would not have eroded one bit, even if her Klout score had declined precipitously.

What this really comes down to is this: inherent in this algorithm is an assumption - a human assumption, and not a piece of math - that out-of-sight = out-of-mind on social networks. To some extent, that has to be true, surely. But just one month ago I "owned" a 66 - this was my demonstrable Klout for well over a month, I believe. Did my week in the hinterlands of Maine somehow erode my powers of persuasion? Now, that's not to say that over time, inactivity shouldn't cause a corresponding drop in Klout - but how much activity, and over how much time - these are human assumptions.

So, my point here is this: if, as Mark Schaefer has suggested several times in the past, the Klout score is the new social "credit score," then I surely hope potential employers, conference directors and even prospective dates do what competent banks do - look at a range of scores over time, and not just a snapshot. Because, as I've noted here in the past several times, the problem isn't with Klout then - they are getting better and better at what they do - the problem is with the people using Klout irresponsibly for decision support.

Anyway, I'm going to tweet Justine Bateman for a date now. At last, I think I have a chance.