David Sifry has posted another State of the Blogosphere article on his blog. There is some interesting stuff in there, for sure, but there are some misleading statements as well. Firstly, the count of the size of the blogosphere makes no sense. When one looks at population statistics, one doesn't count all the dead people. Why do the same for blogs? The size of the blogosphere should include a clear description of active blogs. This issue leads to the second - claims about the growth of the blogosphere. The rate of growth has to be moderated to remove the dead material otherwise it is just an accumulation of corpses. Finally, Sifry never makes the distinction between the size of the Technorati index and the actual size of the blogosphere. It would certainly be interesting to do some research to estimate the size of the blogosphere, but looking at the size of Technorati's index would only be part of that process.
Perhaps more disturbing that the lack of clarity in the post is the manner in which the information is lapped up and repeated, by both the blogosphere and other channels.
Thank you. Thank you. Thank you.
You are spot on about counting dead blogs. I also think that Sifry is seriously underestimating the number of splogs out there. (Try running a search on technorati.com for "hewlett packard". Easily 40% of the hits are spam.)
The Technorati-reported numbers are repeated as gospel by the media because reporters like easy facts to add to their articles. 50 million blogs? Fine. Next.
Blog Pulse reports its tally at around 29 million. I think that, too, is seriously inflated if what you are trying to count is "active" bloggers.
The data that Factiva has puts the number of "active" blogs (those which have a post in the past 30 days) at only 1.9 million. The rest of the long tail (be it 27.1 million or 48.1 million) is making hardly any noise.
Posted by: Glenn Fannick | August 10, 2006 at 05:00 PM
The first problem would be to decide what a dead blog is, and whether or not it can come back to life. I mean, to decide that in a way that can be tested. For example, on my blog, I have a javascript function at the top that causes the page to change every time you load it (it displays a quote), so that, depending on how you look at it, it might or might not show an update. I concur that a better way of measuring blogs would be useful, and excluding cobwebs would be a step in that direction.
Of course, part of the problem might just be in defining a blog. Is Lileks' Bleat a blog? I would say yes. Lileks has said no in the past.
Posted by: Jeff Medcalf | August 11, 2006 at 10:12 PM
On personal opinion, I find this very helpful.
Guys, I have also posted some more relevant info further on this, not sure if you find it useful: http://www.bidmaxhost.com/forum/
Posted by: ocnsss | March 22, 2007 at 03:13 AM