By Gary Schirr, Professor at Radford University
Like many of you, I sign up for the free influence-measuring sites. I am skeptical of all of them but check out my scores and those of online friends. I am well aware that I am no Justin Bieber (Klout score 100) or Barack Obama (Klout score 87) in online influence. But I take comfort if I am near the scores of tweeters I admire, such as @CKBurgess, @KentHuffman, @ChuckMartin1, @DavidAaker, and @WareMalcombCMO.
I find early efforts to measure influence interesting, but I am concerned that parties are taking these fledging measures seriously and making decisions based on them. Some tweeters choose who to follow by their influence score. A recent Wall Street Journal article discussed the importance that businesses are assigning to Klout scores: marketing freebies and even job offers may be tied to Klout scores.
Although I am not concerned that my tenure committee will pass me over for Justin Bieber or Britney Spears, I worry whether the influence of the influence measurers may impact SM communities. If tweeters believe that their online status, likelihood of being followed, and even employability may be affected by these measures, they will adjust how they act. A well-known management dictum notes that nothing can be managed until it is measured. The relevant corollary is that what is measured will affect behavior: bad metrics can lead to bad behavior!
Will the clout of Klout cause loutish behavior in social media? @Econsultancy summarized the flaws of Klout’s measures in a recent blog post. And in one study by @Adriaan_Pelzer, a bot achieved a 51 Klout score in 80 days simply by 1) tweeting gibberish once a minute and 2) not following back new followers. Senior officers at Klout responded to that study (and similar ones) by noting that they were working on algorithms to spot bots. Klout proposes to attack the symptom head on (who wants to be fooled by a bot?) but downplay issues of the metrics themselves.
What kind of “community” will Twitter be if everyone follows these “winning” behaviors? Do you judge tweeters by quantity of tweets? Is it optimal to have 10 times as many followers as people followed? Phillip Hotchkiss, Chief Product Officer at Klout and a serial start-up entrepreneur, commented on an earlier post I wrote on this issue. His most interesting observations are that Klout’s metrics are always being modified to measure influence, and that Klout is trying to differentiate between “bad bots” and “good bots.” It is good to know that Klout is constantly evolving in pursuit of good influence measurement. Phillip’s “good bots” are a little scary. Maybe there is a good science fiction story there!
Be wary of even well-intentioned measurers. For example, many researchers believe that U.S. News & World Report’s college rankings hinder educational innovation by focusing on reputation and resources that please faculty, rather than cost/benefit measures that apply to most services. Similarly, focusing on factors such as tweets that generate action (regardless of content) and on follower/followed ratios could impair Twitter’s evolution.
Be wary of these developing measures. Do not let the measures affect your behaviors or enjoyment of social media. Do not make hiring decisions based on Klout, unless you honestly believe that Justin Bieber is the perfect 100 and the most influential person online! And please, please don’t start tweeting every three minutes!