Wednesday, June 25, 2003

n/nr: LiveJournal is 213th most popular site on the Internet.

Nielsen/NetRatings says that last month, May 2003, LiveJournal was the 650th most popular site on the Internet by unique audience. The 184k people visiting about every ten days (3.13 visits per person during the period) were so active that LiveJournal was number 213 by pages viewed. When people came, they spent 22 minutes at the site, hitting the back button about 26% of the time.

So we now have three measurement methods.

  • Server logs, as reported by weblog hosts
  • Census, as reported by web crawlers, and
  • statistical sampling of user behavior or network packets.
Each reports different counts of LiveJournal users. Why? When should we use some of the numbers vs. others?

At this stage, I'm most comfortable using server log analysis for counts, crawlers for understanding interconnectedness, and sampling for share of blogging vs other high-traffic services.