The issues with determining “influence”
Most social media “listening” systems have a way to evaluate the influence of the various sources in your social media ecosystem. They do that by looking at a variety of factors, such as number of incoming links, number of comments, number of unique commenters, Google Pagerank, etc.
The problem I have with these systems is that an influential person for one company may not be so influential for another. For example, a blogger could score really high on all the above metrics yet only influence other social media pundits. That might be a good influencer for one company, but not for a company who is trying to influence traditional press people or a business audience. They would need a blogger who appeals mostly to those audiences.
Does that mean that the only way to determine influence is to look at those influencers manually?
You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
December 11th, 2008 at 1:19 pm
I think it does not matter “that” much which audience a blogger appeals to. When ideas are valuable they will spread, and it will be noticed & picked up by the rest of the blogosphere.
Though influence is important and is definitely an asset, it must not be overrated. Eventually it is the message that will determine whether the information will spread or not.
Interesting (academic) research on influence can be found here.. (bit heavy material :p)
http://ebiquity.umbc.edu/paper/html/id/359/A-Framework-for-Modeling-Influence-Opinions-and-Structure-in-Social-Media
http://twitter.com/Adgenius
December 11th, 2008 at 3:08 pm
I’ve picked up a few ways to focus the analysis of influence in my conversations with the “listening” companies (tracking that market is my specialty). First is the observation that influence is contextual–it operates within a topic or field. So, for example, TechCrunch is influential for web startups but probably not for fashion. On-topic posts and links contribute to identifying relevant sites.
Second, influence flows through networks–influencers can influence other influencers. So a blogger might reach other bloggers (or other media) who reach the intended audience. Influence analysis needs to go beyond one degree of separation.
Third, some companies consider audience composition in determining on-topic influence. This may be more of a services story than a tools story so far, but M&A activity by Nielsen and TNS suggest that we’ll see more value from combining metrics soon.
Mix, stir, repeat. There’s a lot more to influence than a Technorati number. But eventually, yes. All of these tools put the information in front of a person who makes the ultimate decision.
December 11th, 2008 at 3:21 pm
Francois:
I agree with Nathan above – that influence in contextual. In rec.windsurfing I am a maven (and influential) but in the digital TV forums I am just another newbie asking for help.
Also, I think there is a huge mistake in chasing influencers. It is like staking your video campaign on viral videos. Better to do the right thing all the time than to try to kiss the influencer’s asses.
TO’B