Are There Influencers in Your Database?

A few months ago, on a flight from Sweden to The Netherlands, KLM upgraded me unexpectedly to business class. I happily tweeted about this at the time. And only later I wondered: did KLM do this intentionally because they found out about my higher-than-average number of followers on Twitter? Did they assume I was an influencer?

There has been a flurry of blog posts around the concept of influencers lately, all inspired by FastCompany’s controversial Influence Project, which, some people say, is all about reach and popularity, and not about influence.

If the FastCompany project fails in determining influencers among their readers, then is it possible to identify the influencers in your database? Is it enough to run your database against a social search tool to find out which customers have the most friends and followers on social networks?

Well, let’s see what research says about this:

Michael Wu, Principal Scientist of Analytics, has identified the 6 Factors of Social Media Influence. He writes that someone’s power to influence other people depends on 2 factors: credibility and bandwidth. (And that’s only the first part of the issue, because the likelihood of someone actually being influenced depends on relevance, timing, alignment, and confidence.)

Karen Stephenson, a corporate anthropologist, identifies three archetypes of communicators in social networks: Hubs (people who draw information to themselves and then broadcast it all around them, Gatekeepers (experts at carefully managing information flows), and Pulsetakers (great observers of people and trends).

Apparently, the chance of someone being able to influence other people depends in part on the bandwidth between this person and others, and their power to draw information to themselves and broadcast it all around them. But it is only one part of a complex issue, and the relative importance of the other criteria may be bigger!

But wait, there’s more…

The NewScientist article Why Facebook friends are worth keeping describes the work of researcher Michael Kearns, of the University of Pennsylvania. Kearn’s social experiments indicate that being better connected can give an individual apparently disproportionate influence. It seems, according to this research, that there is a correlation between a person’s influence and the number of his/her social connections.

And in their book The Hidden Power of Social Networks authors Rob Cross and Andrew Parker describe their research, which indicates that a person’s individual excellence or expertise is not a good indicator of his/her performance. Instead, a person’s level of connectedness in a social network is a much better indicator of his/her performance. This might indicate that, in Wu’s model of influence, bandwidth (connectedness) has a bigger impact on influence in the social network than credibility (expertise)!

Now compare this with two opposing views from experts:

Esteban Kolsky, CRM expert, writes that a person with 50 followers can have more influence than a person with 10,000 followers. Strictly speaking, this is true. Just like a person with 1 follower can, in theory, still have more influence than Ashton Kutcher. Such a message appeals to people who claim that the number of followers on Twitter means nothing. But this kind of reasoning is flawed when we’re talking of the likelihood of people having influence, which, as we’ve seen from research, correlates (to some extent) with bandwidth and connectedness. Therefore, by pure logic, the chance that a person with only 50 followers has more followers than a person with 10,000 followers is not 50%. It is less than that! And the number of really influential people among a top 10 of best-connected people is higher than among the 10 least-connected ones.

In an interview with FastCompany, Guy Kawasaki, famous blogger, says “It’s hard to name a person who is unpopular who has influence.” Of course, it is possible to name influencers with low connectedness in social networks. It’s just more work finding them than it is when trying to find them among well-connected people.

And that’s why it makes sense to find well-connected people in your database, and why it would make sense for KLM to find well-connected people among their passengers. It doesn’t mean these people are better influencers. Maybe they are not! And they may get it totally wrong. But it does mean the chance that they are real influencers is at least (somewhat) higher. And if you have just one free business class seat available, you can only make the bet once…