Cites & Insights 2006: A few “popularity” notes

A year ago I did this commentary on the reach and popularity of Cites & Insights volume 5 (2005). Here’s a similar breakdown for volume 6–but with a modest amount of confusion, since volume 6 is split across two domains….using two different log stats systems.

For the old site (where the final issue was C&I 6:8, July 2006), statistics cover the period 12/19/2005 (the day C&I 6:1 was posted) through 12/18/2006–exactly one year. For the new site, which includes all the old issues and began on July 10, 2006, statistics cover 7/10/2006 through 12/18/2006.

Sustaining interest: One clear fact is that readership continues long after an issue has been posted, with much of that readership going directly to issues and essays, not the home page. Strongest indications: Although hits per month at dropped from an average of 56,753 per month for February through June [January 2006 was abnormally high thanks to the Library 2.0 issue] to 32,364 per month for July through November–a drop of 43%–visitors per month only dropped from an average of 17,805 January-June to an average of 16,766 July-November (a drop of 6%). Basically, the daily visitors graph since the site change, entirely to old issues, looks about the same as it did before the site change, but lopping off the spikes that occur just after each issue is loaded.

Overall readership: C&I was visited from 50,818 unique IP addresses on the old site and 11,374 unique IP addresses on the new site. Combining the two, there were some 663,000 total hits (up about 56% from 2005), with an average of 578 visitors per day on the old site, 172 sessions per day (the closest comparison) on the new.

Geographic distribution: Noting that some of these may be spambots, total countries are about the same as last year (166 vs. 167) on the old site with exactly the same number (143) showing more than one visit. The new site shows 92 for the half year, 72 with more than one session. For the old site, 72 countries show 20 or more visits, 61 show 50 or more, and 49 (same as last year) show 100 or more.

I didn’t look at browsers, OS, and spiders in any detail; I believe Firefox is running about 20% of browsers, Mac OS about 3% of OS–and Yahoo! Slurp continues to be the most hyperactive spider by far.

Popularity: It’s tough to make overall judgments for two reasons: The split between the two sites (with only the 50 most visited pages available for the old one) and the clear sense that a pretty substantial portion of an article’s readership comes some time after it was published–with a significant portion coming more than six months later. It would make more sense to look at popularity for the August-December issues no earlier than next July.

For what I’m willing to conjecture, I’m using the same algorithm as last year: 1.5 readers per PDF download and one reader per HTML page visit. Using that metric, here’s what I can say with moderate assurance:

  • The full-issue essay Library 2.0 and “Library 2.0” was by far the most widely read essay and issue–nearly 28,000, almost four times the readership of the second highest, and a whole bunch more than 2005’s top piece (Investigating the Blogosphere)
  • Looking at Liblogs: The great middle (6:10) came next, with more than 7,000 readers. This essay was in the August issue, which makes its high readership particularly noteworthy.
  • Folksonomy and Dichotomy (6:4) had around 6,600 readers; Beyond Library 2.0 and (C)2: What NC Means to Me (both 6:3) round out the top five, with more than 6,000 readers each. The July Bibs & Blather (asking for help on liblogs)also had just over 6,000.
  • Six other essays had more than 5,500 readers. In descending order: (C)2: Will Fair Use Survive? (6:1), my commentary on OCLC’s Perceptions report (6:3); Library Stuff from March (6:4); the August Bibs & Blather [meaningless: those are just PDF numbers for people reading Looking at Liblogs]; (C)4: Analog hole and broadcast flag (6:3), and (C)1: Term and Extent (6:4).

More significant, I think, is that readership was strong across the board–every essay prior to September (except the PDF-only 75th issue) shows at least 3,700 readers, and every issue first posted on the new site has already been downloaded as a PDF at least 1,100 times–including 6:14, which hasn’t been out all that long.

I won’t draw conclusions as to popular and unpopular types of articles; it’s not that clear–except, to be sure, that Library 2.0 and liblogging are big draws. Heck, even the silly 75th anniversary issue had close to 1,800 downloads…

Comments are closed.