Mapping the Blogosphere
Being a new, fast growing medium, the Blogosphere still hasn't been properly mapped out.
And by this, I mean in all it's dimensions: the economics, the socio-behavioural characteristics, the key companies, the key individuals, any geographic clusters, and so on.
Recently there appears to be a race on to create the best Top 100 list of Blogs (whatever that means), spurred on by Jason Calacanis. But many people believe this is a very "old media" way of measuring influence because the readership of blogs is so distributed and spread over so many topics.
The reason I bring this up is that I've noticed some interesting "experiments" done by Fernando Plaza here and here (posts in Spanish).
Basically, he's been playing around with two measurements:
1- the amount of overlap between blogs. I think this is very important because in the Old Media world, few people would subscribe to both Time and Newsweek, because you're paying twice the money for essentially the same news. But in the blogosphere, where it's trivial to add subscriptions via RSS, the same gadget freak might be reading both Engadget and Gizmodo. What are the implications for advertising? Does this dilute or amplify a single blog's influence?
2- Fernando also explores the "Slashdot" effect and theorizes that a blog's real power is not how many readers it has but rather how many of those readers it can redirect to content it links to.
This last point raises the question of whether it might be more relevant to measure referrer traffic as a way to guage the strength of the referrer, particularly as a share of the referrer's total traffic, rather than simple inbound links.