When Klout changed their algorithm for social influence scoring at the tail-end of October, a cry went out around the internet as many people saw their Klout scores nose-dive and their egos take a bruising. Some people even said the changes cost them their jobs.

Klout’s new model represents the “biggest step forward in accuracy, transparency and our technology in Klout’s history,” said Ash Rust, Director of Ranking at Klout on Klout’s official blog, but many influencers begged to differ. Was there much point in influencers jumping to the defence of their previous scores or was it all sound and noise, signifying nothing? Are social influence metrics doing influential users justice and are they providing a true reflection of a person’s influence?

I did a bit of crowdsourcing on Twitter and the consensus wasn’t positive. Most respondents said they only use social influence aggregators to track their own score, while those who use it professionally say they only utilise it to give them a basic idea of someone’s possible influence before looking into their online behaviours manually.

@pandorasinbox It’s interesting, but Klout doesn’t make itself transparent enough for me to believe it’s in any way useful.

— Oz Kills (@ozmills) November 9, 2011

@pandorasinbox I’m not convinced, the algorithms seem limited and one said I was influential about cereal.

— Nicola Balkind (@robotnic) November 9, 2011

@pavedwithgold1 @pandorasinbox I think it’s easier to judge by visual evidence like numbers of RTs & @replies over a period of time.

— Nicola Balkind (@robotnic) November 9, 2011

@robotnic @pandorasinbox I agree with Nicola. Klout etc. I tend to use for my own amusement, but for others I prefer a more manual approach.

— Kaye Symington (@pavedwithgold1) November 9, 2011

@pandorasinbox It will never be a perfect system… Some manual work is always important! You get a clue on activity level…

— Michael Kristiansen (@Mikkasen) November 9, 2011

@pandorasinbox I look at them in combination with many other factors but wouldn’t make a decision based solely on e.g. klout

— Annie Macfarlane (@anniefiddle) November 9, 2011

@pandorasinbox Largely, yes.Practical applications depend on what tangible benefits there are from having a ‘reputation’ in first place.

— Iiiiaaaaaaarrrrrrghn (@iainmhepburn) November 9, 2011

So what’s the problem?

To understand one of the major reasons for this we have to strip social media back down to its core – the fact is that just because someone isn’t measured as influential on social media doesn’t mean they’re not influential elsewhere, perhaps even in the offline world. These tools are also not contextual and a user involved in a scandal may be deigned to be influential just for having more followers than they’re following or receiving the attention of other influencers.

Social influence scores are currently negatively impacted if you interact highly with other users with low influence scores, which goes against what social media is all about. It’s a very elitist approach that ultimately encourages users to only interact with people who can buff their social influence score.

The value of tools is also restricted by how much you let them in to your online world. With an entirely “friends only” Facebook account, my social measurements there can’t be effectively read to contribute to my score, even though I am extremely active on Facebook. My own Klout (63) and Peerindex (44) scores are based almost entirely on my activity on Twitter and my engagement with other influencers there. Bearing this in mind, does it seem as if these scores are giving a true reflection of a user’s level of influence across the board?

But are they valuable?

All of this isn’t to say that social influence metrics won’t be hugely beneficial in the future. Social influence measurement tools such as Klout, Peerindex and PROskore are still in their infancy and each evolution of their algorithms is enabling them to make more accurate estimations of influence. However, for the time being, social influence scores need to be taken with a massive pinch of salt – although they can be valuable tools in conjunction with your own research when trying to determine a user’s influence, they still have a long way to go before they become accurate enough to be our main tool.

How useful do you find social influence aggregators? Do you find them useful in your personal and professional life? Have you ever tried to actively cultivate a high score? We’d love to hear your thoughts!

Check out how wildly my influence scores vary from site to site:

Klout: 63
Peerindex: 44
PROskore: 39
TweetGrader: 95.5

Caliberi - Nov 11, 2011 | Blog, Crowdsourcing, Social Media