Klout Confesses Interesting Falsehood

klout jeansIt has been said that data tortured long enough will confess to anything. Recently, courtesy of Klout, I believe this was confirmed.

I signed up for Klout a few years ago and have watched the rating rise and fall based on content I posted. Over time, Klout recognized me as an “expert” in a few areas. Since these are areas in which I work and write about, it seemed logical and accurate.

A few weeks ago, I posted the following sentence on social media: “I remember the days when I could get something out of my jeans pockets without having to almost fully recline.” TMI for this blog, maybe, but let’s face it, things (like my waistline) are different now than they were 20 years ago.

I was merely pointing out something obvious, and something I found humorous, to my Facebook friends. Some thought it was entertaining, it captured a few comments and likes from my contemporaries. And that was the end of that, or so I thought, until I received an email from Klout that I had been recognized as an expert in a new topic…

Klout Confesses Falsehood

Seriously? I am now an expert in “Jeans”? Really? Klout, there is something messed up with your influence calculation algorithm. I’m not even an expert on pants much less something so specific. I mean, I know how to put them on, but my expertise pretty much ends with the basics. I walked about in abject fear, anxious that I would be found to be a mere novice in the land of jeans experts.

If there is any expertise to be found here it is in persisting in the belief that I can wear the same size jeans as I could before I entered my fifties, my metabolism slowed to a crawl, and I didn’t mind making fun of myself. I may post that and see if Klout awards me expertise in Self-Deprecation.

One comment meant to bring a laugh to a few folks had the consequence of discovering a flaw in whatever algorithm Klout is using. A flaw in that it takes whatever information it can obtain and analyzes it until it can generalize the content for some tracked topic. And therein lies a problem with automating anything related to massive amounts of data.

At some point, some intelligent analysis and thorough review is needed to figure out if it the results being produced make any sense at all. Some checks and balances. A bit of quality assurance.

Hey, even my jeans were QA’d by Employee #37.


Add Comment