How not to explain Klout

Take a look at this video, (came to me via Mashable)

This video is a great example of using Xtranormal, an incredibly easy to use service that lets you make cartoon videos from your own scripts. I’ve experimented with the free version a while back, and kept meaning to go back to it, seeing this video reminds me to look into it again.

Kout moved down this week

The trigger for this video seems to have been the drop in many Klout scores this week as “they” refined their algorithm. My personal score seemed pretty unaffected so I can take a disinterested view. My score’s  here. As I write this I’m at 47: 47.15, 485, 18, 19 to give my full vital statistics.

But I recall a few months back, an earlier algorithm change dumped me, and I felt quite down.

Measures that define themselves

Klout is one of those ideal businesses. They’re measuring something which is important to many people: how important they are in social media. But which is not easily to cross-check. Its measured in a way which they make little attempt to explain in detail. It’s like a modern day AC Nielsen rating or market share stat. Hard to live without, and hard to check. It’s a bit as though the beauty pageant judges also took all the vital statistics (do people still use those words? apparently they do: Google shows up 6m plus results; with 200 plus in News), in a way that no-one could check.

I preferred the old-style Klout of a few months back. That showed in more detail the people they claimed influenced me. I have no idea at all how they measured this. But it gave a useful repertoire of new people to follow. I simply drilled down in Klout to find the people that influenced the people they claimed influenced me. They tended to be journalists or broadcasters, mainly in UK, who had higher Klout scores than me. And mostly they followed me back.

This resource is much depleted now. Klout now focuses on people I influence, with tags suggesting I invite them to Klout. I have a nagging feeling that if I did my Klout score would go up.

Klout Perks

Then they’ve added Klout Perks. These are offers, most of which carry the small print “Sorry you’re not eligible” and a “How to be eligible” roll over. For me eligibility mainly requires that I tweet about different topics, and move to the US, or a specific part thereof.

These perks are an outgrowth of earlier celebrated marketing use of Klout scores. A couple that I recall: last June (2010), Virgin America offered free flights to high Klout scores to launch its new route out of Toronto. Then Las Vegas Palms Hotel and Casino set up a special Klout Klub last September (2010): jacket not required, but Klout score essential. The Virgin application seems more appropriate as if high Klout scores are really influentials, and if the flight service is good, it should get a deal of social media buzz.

Basically I think Klout Perks are hoping to bring Klout promotions within the reach of smaller, local, businesses. Buzz word of the year: localization. Easy do-it-yourself promotions that get lots of views, and a few actions: with the actions only available to the influential, defined within a locality.

Who guards the guards?

It makes me feel a little uneasy that the fox is assessing the solidity of the henhouse. Should the people whose magic gives you the score, also be making money from it? So those beauty pageant judges, decide the contest, measure the contestants, sign the sponsorship deals, AND take the sponsorship funds.

Competition

Now there is a Klout competitor: TweetLevel from Edelman. Now they provide a lot of (apparent) information on how their scores are derived. And one of those laughable formulas that looks as if it came from the mad scientist’s blackboard in a horror movie.

TweetLevel's Methodology is a mad scientist formula

 

One of my thesis supervisors long, long ago, at London Business School, the late Professor Andrew Ehrenberg, coined the word “sonking” for this sort of thing. Stands for the Scientification of Non-Knowledge.

He would often say “I SONK therefore I am”. Needless to say he greatly discouraged us PhD students from doing it. It is a very bad habit.

I’m amazed to find that I can even find a hyperlink for SONKING. That source aptly describes Professor Andrew Ehrenberg as iconoclastic. That he was, and a brilliant man, and magnificent teacher.

My TweetLevel score right now is 67.8: my complete vital stats? 63.3, 51.7, 43.3. I see my lowest score is for “trust”. This is cross-referenced to Edelmans’ Trust Barometer.

 

 

The blurb says that its basically the rate at I which my tweets are re-tweeted. Not sure that’s what I would mean by “trust”. More like it’s interest. newsworthiness, or the intriguingness of the headline.

Refinement

But these are the days of alchemy in digital marketing. Different recipes are bandied about and hailed as elixirs. Over time the least trustworthy (omg, according to TweetLevel, that’s me!) will be officially designated as snake-oil. And there will emerge reputable measures that marketers find really do measure influence and reputation.

It will be a while before we get there. And Hey! if my scores go down, I’ll say it’s because they didn’t like this piece…

Be Sociable, Share!