Following on from my post about Klout on Monday, a certain West Ham fan forwarded me this excellent critique of the tool by Jeremiah Owyang, that includes a great illustration of why Klout shouldn’t be relied upon to give a full picture of online influence.
After Kenneth Cole’s PR disaster at the start of the month (where he piggybacked on the uprising in Cairo to advertise his new spring collection), any sane person would expect his online influence to have dropped... right? Not so according to Klout, where his influence score increased by 46%.
Based on Klout’s measurement criteria (that is "highly correlated to clicks, comments and retweets" but crucially does not factor in the sentiment of these reactions) Kenneth Cole suddenly became a lot more influential. But from a quick scan of Twitter it was clear that the vast majority of reaction was negative, and Klout’s single metric fails to tell this story.
It’s worth heading to Jeremiah’s post to read the full article, but to get you started here are the six insufficiencies he highlights in Klout:
- Alienating your mainstream customers in desire to serve influentials. Careful when using Klout to segment customers priority, while high scoring Klout users may appreciate the ego boost –anyone with less than the ideal number of points may quickly fall out of love with your brand if you display elitist behavior. No one likes a ‘better than thou’ unless they are the ‘thou’.
- Consumers will game the system –reducing validity of metric. Expect many people to start gaming the Klout systems, in fact I see some ‘influential types’ tweeting over 200 times a day to try to hopefully raise their Klout scores, which just ends up annoying their followers.
- Klout is not representative of a customer’s real influence. Currently, as I understand it, Klout only siphons in content from Twitter and Facebook if the user allows for FB connect.
- Without sentiment of the influencer –the gauge is incomplete. Klout lacks sentiment analysis, so true opinions of what’s being said about the person may be ill-informed, see Kenneth Cole example above.
- Relying on this single metric alone is dangerous. as Frank Eliason of Citi (formerly Comcast Cares) indicates the “sleeping Comcast technician” was uploaded by someone who had practically zero prior online influence.
- Influence is not a gauge of true buying potential. Perhaps the most important point is that influence scores don’t necessarily impact the revenue or customer satisfaction of your brand to your core set of customers. While we trust Scoble for technology choices, relying on him for the latest in landscaping design may be a mis direction worth avoiding.