Klout is Broken. Here’s Why.

Last week, Klout began creating new user profiles and scores based on data pulled from Facebook.  This means that, if you have your Facebook account linked to your Klout profile, you will start seeing your Facebook friends and family (including kids as young as 13) appear in your Klout influence network, with Klout scores assigned to them–something which has raised major privacy concerns.

But there’s another issue, and one that is very serious for any company that is using Klout scores to inform business decisions (like hotel perks, customer service triage, blogger outreach, hiring or grading decisions decisions (video at 3:17)):  the new system is creating duplicate accounts for the same individual — with different Klout scores.

For example, here’s the Klout profile for Liz Gumbinner, who runs the popular Mom-101 blog and has been “cited as a digital influencer and thought leader by Nielsen, Forbes and CNN.”  Liz has activated her account and chosen to link her LinkedIn profile, but note that she has not linked her Facebook profile.  She has 19,811 followers on Twitter, and a Klout score of 54.

Klout is broken -- creating duplicate profiles and scores for some users.

But, wait — here’s another Liz Gumbinner!  This one has a profile and score created only based on her Facebook data; she has not activated her Klout profile or linked other accounts.  She shows up as being in Shelly Kramer’s influence network, and has been assigned a Klout score of 60.

Klout is creating duplicate accounts from data scraped from Facebook.

I have found two other examples of users–people I know personally so I know for a fact that they are the same person–with duplicate Klout profiles.  In each case I found, the score created from the Facebook profile was higher than the score created by the Twitter-based profile, with a delta as high as 10 points between the two scores.

I am not going to try and analyze why a personal Facebook page would yield a higher Klout score than a public and influential Twitter + LinkedIn profile.  But if you’re using Klout scores for “fast sorting of a bulk of people online in lightweight circomstances,” how will you know which Liz is the one to pay attention to if all you’re seeing is the score displayed in some plug-in or app?  For a company that takes its role “as the standard of influence incredibly seriously,” this does not inspire confidence and trust.

For the record, I believe that influence metrics are here to stay, and, used appropriately, can play a valuable role in managing online conversations.  But that will not happen if privacy concerns begin to stifle those conversations.  And the companies who are at the forefront of this still-developing science, especially if they are well-funded, will not help develop trust–or influence–unless the content they deliver passes a basic common-sense test.

In the meantime, I would recommend that any companies who use Klout scores as a data point to help make business decisions proceed with extreme caution until these issues are addressed.