This post discussed some aspects of C&I volume 5 (2005), based on statistics for visits beginning 1/1/2005 and running through the end of the year. I noted that I was waiting for a cumulative set of stats, and might have more to say.
That cumulative set is ready, covering December 18, 2002 (when Cites & Insights moved to the Boise State platform at the start of volume 3) through January 6, 2006.
I won’t go into nearly as much detail. A few overall items are sort of remarkable: visits from 87,336 unique IP addresses during that three-year period, including visitors (real or imaginary) from 177 country-equivalents, 109 of which had 10 or more visitors, 62 with 100 or more visitors (Bahrain with 103, Serbia and Montenegro just below with 99), 15 with more than 1000 (US, UK, Canada, Germany, France, Norway, Australia, China, Netherlands, Japan, Italy, Spain, India, Sweden, Korea–Taiwan, Israel, Belgium, New Zealand come next, with 670 to 974 visitors).
I combined numbers for the 50 most popular HTML pages (45 articles plus the 5 site pages) and 50 most frequent PDF downloads, then did the same calculation as for the single year, figuring one reader per HTML and 1.5 per PDF. I added PDF counts for 2005 alone for the four volume 5 issues that weren’t among the 50 most popular overall (excluding separate HTML downloads).
Highest readership overall: No contest. 3:9, the CIPA special, more than 11,000 estimated readers. Second highest and highest article: Also no contest, 5:10b, Investigating the Biblioblogosphere, just shy of 10,000 estimated readers.
Then it gets weird–or maybe not: the next four, all between 7800 and 8000 readers (and the only ones between 7000 and 9500) are one of the Offtopic old-movie roundups, a Copyright article, a Wikipedia perspective, and a Products roundup–but all four were in 4:12, and it was mostly full-issue downloads.
In general, it appears that copyright is popular, as are some scholarly access pieces–but so, to a lesser degree, is almost everything else. The range between 5,000 and 7,000 estimated readers includes 8 articles and 12 pre-volume 4 issues (almost all of volume 3 and three volume 4 issues that didn’t have highly popular article downloads–in one case, the glossary 4:2, because there was no HTML version). Two of the “less popular” issues of v.4 also didn’t have HTML (the Broadcast Flag and Copyright specials).
Only three issues since the move to Boise State don’t show at least 3,000 estimated readers, at least for part of the issue–and two of those three (5:12 and 5:14) are too recent for a good count, since it’s clear that readership for old issues continues at a reasonable pace. No issue failed to hit the 2,000-reader mark.
The lowest readership in the past three years? Volume 5, Issue 9, the July/August 2005 issue. Still more than 2,000, but just barely. Still way more readers than I ever expected for this ejournal, to be sure. What about that issue? Well, it came out during the summer (I notice that other relatively-lower issues are also summertime issues). It lacked any really hot issues or perspectives; the longest essay, on the Grokster decision, was basically a “move along, there’s nothing to see here” essay. Apparently my lumpy little perspective on “Predicting the Future of Academic Libraries” was not a hot item. Last summer was a fallow time. Those times also have their uses.