Mapping Categories of Influence in the Blogosphere

If we take a graph of the blogosphere, we can visualize who is linked to whom. If we select a specific blog, we can highlight the edges in the graph that link blogs which are a minimum distance from the selected blog. In the following examples, we can see that some bloggers have a very broad reach in either the socio-political subsphere or the technology subsphere. More interesting than this is the case where a blogger acts as a bridge between these to topical areas.

 

	Om Malik's gigaom blog has clear reach in the technology subsphere. The graph highlights edges that traverse the blogosphere graph to a maximum depth of 3.
	Malkin's blog has very broad coverage in the socio-political subsphere.
	Jeff Jarvis is positioned to reach both the technology space and the socio-political space.

The insight here is that influence is not simply a matter of numbers - the number of people that you can reach and inform, but also the disperssion in to topical areas. If, like Jarvis, you are positioned to inform both of the two major subspheres of the core, that puts you in a different role than if you are purely of topical influence.

There is an interesting notion here which can be contrasted with one of the themes in Chris Anderson's book 'The Long Tail' (review to come soon). Anderson makes the point that top N lists are pointless as they aggregate counts across genres. In other words, the top 100 singles don't compare apples with apples as they mix different types of music. The correct thing to do, he argues, is to have ranks by genre if at all. We could apply this to the blogosphere and categorize bloggers by topic/genre/intent. However, analysis such as the above suggests that there are bloggers whose appeal truly is comprehensive and thus there really is an A list which transcends topical niches. Once could argue, then, that Jarvis is on this list.

Comments

[Michal Migurski]

How can this be squared with the "Obliteration Phenomenon?" (http://www.garfield.library.upenn.edu/essays/v2p396y1974-76.pdf)

In that essay, Garfield describes how citations in academic papers drop off when ideas become so widely influential that they're no longer considered worth citing, e.g. Einstein's original papers on relativity. A more current example might be Michael Goldhaber's ideas on attention economics - he doesn't get cited often, but his few papers on the topic from the past decade influence and support huge swaths of business plans being tossed around right now. These graphs feel as though they are measuring second-order influence, e.g. people who are pundits popularizing ideas worked out by others. Who is the actual influencer here?

[Ian Stiles]

Very interesting.

Would you be so kind as to indicate what you used to gather the data and then view it?

[Flemming]

Interesting. Thanks for sharing.
And you are right - influence is not just about numbers; that's called popularity.

While those who are very popular usually have some influence you don't need to be popular to have influence.