Early last year I wrote a post about online influence measurement service Klout, I ran a few simple tests and came to a conclusion: Klout doesn’t work.
Based on several recent articles supporting the service, I thought it was worth taking another look at Klout and restating a few points about the value of online influence and measurement.
At the time I wrote the original post, a lot of people loved Klout. Many still do, but there has been something of an online backlash as well.
Lots of commentators have had a proper moan since then about the inaccuracies of the measurement algorithm Klout uses, and rightly so.
While trawling the internet over the past few days however, I’ve noticed a raft of articles popping up telling me that I’m completely wrong.
That Klout does matter.
And I agree, kinda.
For the record, I’ve never felt that Klout was useless. If anything, it’s the expectations of users that are the problem, rather than the service itself.
If you’re a PR and need to hunt out a few bloggers to promote your client’s latest doodad, then it’s an excellent place to get started. Go to Klout, type in ‘Doohickey Bloggers’ and pull a list.
Just don’t forget to go through each and every one’s content carefully before you send them all free rocket boots.
As I say, we’ve heard this before. It’s slowly been improving, but is still far from perfect. Despite this, several people have recently been making statements to the contrary:
Over on Convince and Convert, Jay Baer recently wrote that critics of Klout are missing the big picture.
I have immense respect for Jay, but here I believe his argument is flawed. His post claims that many are opposed to Klout because it fails to identify offline influence.
As an example, he reiterates Paul Gillin’s claim that NetScape founder Marc Andreessen has a low Klout score, because he doesn’t spend a lot of time online.
And why would he? He’s too busy making massive piles of cash. It doesn’t mean he’s not influential.
That’s a perfectly good point but I really don’t believe that this is the problem. No-one I know cares whether or not Klout registers offline influence. Klout never claimed they could.
The problem is that Klout measures online influence innaccurately.
As an example, I recently took a week off. During that time, I significantly reduced my online presence. I checked in to Foursquare once, tweeted twice, and possibly sent an Instagram.
My score increased by three points.
During the same period, the @Econsultancy Twitter account gained 8,500 new followers (Hello new followers!). Yet the company score dropped by two points, despite more followers, more content and increases in engagement, @replies, ReTweets, conversion and CTR…
You can see where I’m going with this.
Over at Grow, Mark Schaefer also believes Klout is important, although he has a slightly different take on things. He believes that anyone who can create and move content can be an influencer.
I’d go one further and say that you don’t even really need to create, just be actively disseminating information.
I post a fair bit on my Twitter account about social media, because that’s what I do all day, but I also post links to music, stupid gifs, and on one memorable occasion a picture of me being chased by a giant wooden elephant (oddly, that last one made several people’s paper.li dailies. Proof that totally automated curation isn’t all it’s cracked up to be).
It’s not targeted or optimised, it’s just “stuff I like”.
I wouldn’t say this makes me an influencer; rather, I have just enough followers who are vaguely interested to grant me a little broadcast capability.
It’s useful, but true influence, in my opinion, is a narrowcast discipline. Occasionally someone will ask my help directly, and that’s where I can influence. I can personally recommend a product or service.
Again, it’s all down to context.
Last week Eloqua’s Joel Rothman tried to determine if a higher Klout score affected conversion rates by mapping users with high scores and comparing their ability to spread a message compared to their low-scoring counterparts.
His experiment showed that, yes, if you have a million followers then it’s a bit more likely someone will click on a link in your tweets. If I advertise sofas on national TV, more people will see that ad than if I put it on a card in my local newsagent’s window.
Will they purchase? Only if they’re looking for a sofa.
So what’s the problem?
(Or if you prefer: Get on with it Matt, for God’s sake.)
The problem is that Klout adds up all this data and stick an arbitrary number on it.
If you’re more active or have a larger network then yes, your score will increase, and you can claim your free jetpack. Whether you're a human, a twitterbot or a dog on the internet doesn't matter.
That’s fine on an individual level, but companies need to stop behaving as though it’s an accurate rating for... well, anything at all.
It is (just) possible to use Klout as a vague benchmark against your competitors, if you are running very, very similar campaigns.
Unfortunately a lack of knowledge or deep analysis means businesses are still viewing these figures as deliverables.
I’ve seen a number of recruitment notices (no, I’m not looking…) where 'Increasing the company Klout score’ is a job function.
This is utterly ridiculous. Why? So your company can get free Stephen King novels?
I could work for a hardware store, there’s no reason I can’t post funny cat videos on the company Twitter account all day and get a high Klout score. I still couldn’t sell a hammer.
Similarly, it’s easy to game a system like this, especially if you want a few freebies.
Content is king, but relevance is key
In essence, this is an old argument. And I’m not saying that Klout and similar services aren’t important.
I believe that anything can be important if enough people decide it is. Money is important because we all decide to believe it is.
Last night’s football scores are important, but they don’t have the slightest effect on your daily life (unless of course, you bet money on them).
Klout scores are important, if you want to kid yourself that a large arbitrary figure has anything to do with how relevant or successful you are.
They aren’t important as a measurable business metric, and they certainly aren’t a strategic measure or a deliverable.
I am convinced that it is possible to measure an individual’s importance to your business online using a mathematical formula.We just haven't finished writing that formula yet.
Once the algorithms are less muddy, scoring will become an important method of segmenting different audience sectors for business.
I’m thoroughly in favour of giving major purchasers and powerful PR bloggers a bit of extra love, and I also believe in the importance of nurturing low-level customers and possible leads carefully over the long term.
I’m not in favour of randomly sending out free stuff to people who have no relevance to my customer base.
Again and again I've seen this argument pop up. I believe that Klout is on the right track, but users at all levels need to be realistic about how and why they use the service.
Are you benchmarking? Are you after a free cocktail? Or are you genuinely looking for relevant influencers?
If it's the latter, there's still no substitute for checking them out in person.
- Big numbers don't matter.
- Having a higher Klout score than your competitors doesn't matter
- Quality content and relevancy matters.