Day: February 5, 2003

Rogers Cadenhead does some snooping in Alexa and discovers (along with graphical comparisons of blog traffic) some interesting things about some of our favorite blog hosters:

• Five weblogs receive one-third of all visits to EditThisPage.Com: Familie Berg (13 percent), Hack the Planet (8 percent), Phil Wolff’s Dijest (5 percent), Cyberdaily (4 percent), and Flangy News (4 percent).
• Top five on ManilaSites.Com: Caversham Booksellers (17 percent), J.D. Lasica (14 percent), CHAINguardian (8 percent), B.A.’s Weblog (4 percent), and Heroes and Villains (4 percent).

Adam Vandenberg (aka Flangy) thinks his prominence is less due to merit than the fact that a lot of people are moving off EditThisPage. Au contraire. (Although there are certainly people moving off EditThisPage.)

And Dave wonders why his Slashdotting looks like 5000 reads, while Joel Spolsky’s looks like 400,000. Well, I can’t explain two orders of magnitude, Dave, but one order might just be a difference in methodology. The stats in Manila and Radio (and SiteMeter) present summaries of visits and page views, and don’t count things like graphics loads as hits (though Manila, at least, counts RSS downloads). I don’t know about Joel’s software, but if he’s looking at more raw web traffic data, he’ll see a hit for every image on his page in addition to the page view itself.

I have meant to blogroll Ross Mayfield’s weblog for a while, but it took his linking to me a second time (after I jokingly compared him to Gibson’s Gentry) to get around to it. Ross’s site is the best around for pointers to publicly accessible data about the shape of the Internet. These days if I have questions about how fast the Net is growing, how its users are distributed, what its mechanism of growth looks like I look to Ross first. This morning, for instance he points to a paper by NEC researchers that says that the old “power law” explanation for the popularity of web sites (the most popular attract more links than the newer entries) doesn’t hold across all categories. He goes on to suggest that the falling barrier to entry for web publishing may also be a factor.