The Problem with Concentric Circles
Ana hits on something that’s been bothering me for awhile: All these graphs about social media influence that show a bunch of concentric circles are wrong. She wrote: “Aside of the fact that this kind of thinking is oversimplified and wrong for the obvious reason (a person who’s a flickr follower is also a twitter follower is also a facebook friend, so these circles are hardly ever concentric), there is a more important one. Interpersonal influence is not broadcast. Simply put: how influential you are going to be does not only depend on you, it depends on how influential your friends are. That is, your influence is the outcome of their network even more so than yours. That applies to plain scale (how many people are in their network), but also on scope (who they are). So, hard as it may be on the ego, you are not at the center.” (I added in the comments that the subject matter of the communication had a lot to do with influence as well.)
I wrote a bit about this back in 2006, specifically pointing the confusion in these types of graphs between reach and influence.
Hi, I'm 
Love this post Noah. Reminds me of something I was struggling with when one of the brand managers I was working with was passionately interested in working with a WOM vendor. The vendor was throwing our ridiculous numbers in terms of exposures.
The most likely outcome if you get highly influential people to care about your content is that they share it with a bunch of people whose own influence is far less so the rate of sharing steeply declines along the way. There may be more value in simple mass reach vs. focusing on finding influencers and the idea that they’re more valuable. Influencer power wanes across the network.
Thanks for sharing.
re: the subject matter of communication.
while the most obvious answer is that not everything is interesting to everyone but that a similar dynamics of info sharing can be applied to different subject matters.
message about food and not marketing (to your comment on my blog) will spread because of my friends’ networks, not because of my own reach. As long as someone with the big network of foodies picks it up.
which brings me to something more interesting: why people choose to share. So far it is known that they share stuff that they either like a lot or DISLIKE a lot (remember mccain & palin’s stuff?; reactions to Crispin’s Whopper Virgins?, etc.) Liking/disliking, i.e. positive/negative “valence” of content may be a better predictor if it’s going to spread than its subject matter.