Influence without Klout

All the attention given to "social influence measurement" tools like Klout and sort-of-Australian newcomer Kred is hilarious, because as far as I can tell it's all complete bullshit.

Klout, founded in (where else?) San Francisco in 2008 and billing itself as "The Standard for Influence", is the current fave. The company reckons its Klout Score "measures influence based on your ability to drive action".

"Every time you create content or engage you influence others," they write. The Klout Score measures how many people you influence (what they call "true reach"), how much you influence them ("amplification"), and then in turn the influence of your network of contacts ("network"). Each is rated out of 100, and the three numbers combine — somehow — to become the Klout Score.

Now the core idea kinda makes sense.

We've always been influenced by word-of-mouth recommendations, trusting close friends more than strangers and anonymous marketers. Some people are more influential than others. Marketers have always wanted to reach those people. The visibility of our online activities means that we might be able to identify those influencers more easily.

But strip off the social media expert guru (SMEG) jargon, take off the SMEG goggles, and take a deep breath of fresh air. Because there's at least three problems here.

One, Klout takes no account of your offline activity. All Klout and friends measure is whether what you say and do online (what SMEGs call "creating content") is mentioned, passed on or responded to by others ("engaging with the content").

In Klout's online bubble-world, you could be on TV every night and in the newspapers every morning, but unless you're tweeting or posting on Facebook in your own name, and interacting with people, you don't exist.

Greg Combet may be an ALP power-broker and government minister, but as far as Klout is concerned he's a nobody.

Related to this is the rather naive assumption that sheer volume is influential.

"Klout, for instance, assumes that if you talk a lot about a particular subject on Twitter, and receive numerous retweets and replies, then you’re more likely to be influential about it," wrote Max Reyner, head of insight at UK marketing firm Protein. "You could be just mentioning that subject as a critic, or asking a lot of questions about it as a novice."

Two, where's the research connecting all this pseudo-scientific measurement of online chatter and re-chatter to actual, real-world behaviour such as changing a political vote, buying a different model of car this year or giving up smoking?

There isn't any.

In fact Mark Earls, author of Herd: How To Change Mass Behaviour by Harnessing Our True Nature, reckons these influence-measuring tools are based on outdated heirarchical broadcast-style models of influence — the idea that people of lower status pay attention to those of higher status, who then give them a message.

"That's not how social influence works," Earls is quoted as saying by Reyner. "In reality it's much messier. It's not one or two interactions involving a recommendation. It's several interactions with a number of people that add up to influence."

Three, there's no evidence that targeting these so-called influencers is any more effective that pumping out your message at random. In fact, research indicates the exact opposite.

Australian sociologist Duncan Watts is now director of Human Social Dynamics group at Yahoo! Research. In 2006 he was professor of sociology at Columbia University where he co-authored the paper Influentials, Networks, and Public Opinion Formation.

Based on computer modelling, it's pretty maths-heavy. But here's the money-shot:

"Large-scale changes in public opinion are not driven by highly influential people who influence everyone else but rather by easily influenced people influencing other easily influenced people."

In other words, it's not about the influencers. It's all about whether people are willing to be influenced.

More recent research from Hewlett-Packard's Social Computing Lab showed something similar with Twitter's trending topics.

"User activity and number of followers do not contribute strongly to trend creation and its propagation. In fact, we find that the resonance of the content with the users of the social network plays a major role in causing trends."

As Naked Communications' Adam Ferrier wrote in AdNews, Klout is really for narcissistic, insecure attention seekers.

"Klout will contribute to people comparing themselves with others. It will increase peoples' desire to be famous and see their self worth tied up not in who they are, but rather in who they are seen to be. You're  measuring yourself, and unless you're Justin Bieber you won't measure up."

And speaking of Justin Bieber, the very fact that he scores a Klout perfect 100 should tell you something about how accurately this measures true influence.

One of the messages of Alfred Crosby's 1997 book The Measure of Reality: Quantification in Western Europe, 1250-1600 is that once we can measure something it becomes important and shapes our worldview.

Klout attaches a number to our social interactions online. The number seems to be bigger for people who are more active online, so we imagine that it's a useful measure. Something very human makes us want to make our own number bigger — presumably by thinking about our social interactions as "creating content" and "engaging".

But us humans will chase any number. Klout could just as well have handed out thermometers so we could compare and compete with our rectal temperatures. Which, when you think about it, is precisely what they have done.

Whether Kred addresses any of these issues remains to be seen. Its main differentiator seems to be transparency. Perhaps it's just a clearer view of the bullshit.

But in the end none of this really matters. Punters get a brag-number to stick on their Facebook profiles. Advertisers get another vast tranche of personal data. And that's what the internet is for, right?