I concur with James' expansions. There are many proxies for visitor measures (particularly when interested in how the number varies over time, or between different parts of the same website). The difficult part (expanded below) is when you need an absolute count to compare with other kinds of visit statistics (e.g. turnstiles).
A key factor omitted is consideration of the length of a visit, or, more importantly the time that must have elapsed between a users two consecutive "hits" for them to be considered distinct visits. We are working with this definition of a virtual visit for public libraries websites: A Virtual Visit is a sequence of visits to an identified collection of web pages covering all aspects of a library's web presence. The sequence begins at the first visit from a particular web browser session. The sequence ends after an interval of 30 minutes elapses with no further visits. Each sequence of visits counts as a single virtual visit. [After Bertot et al (Bertot, John Carlo, Charles R. McClure, Joe Ryan. Statistics and Performance Measures for Public Library Networked Services. 2001.)] But I digress - as special logging techniques are required to efficiently derive these measures. Adrian At 12:12 PM 9/24/02 -0500, you wrote: >While Analog doesn't offer a visitor count per se, if your site assigns cookies and >your log files record them, Analog will produce a User report based on those >cookies, and a visitor count can easily be derived from this report. > >To do this, just include the line USER ON in your configuration file, and set >USERFLOOR for a high number of requests e.g. USERFLOOR 1000r. You'll get a report >that looks like the following, in which the last line includes a "user count" of >sorts "[not listed: 3,218 users]" in this particular report. Of course, this >excludes all users who refuse cookies, multi-counts users who access your site from >more than one computer, undercounts users who share computers (families, libraries, >etc.), and presents other problems as well. But it is a number you can give to a >boss who insists on a user count. > >If you don't assign cookies, you can also use the "Distinct hosts served" line in >Analog's summary. This has even more flaws than cookie based user counts, but it's >there. In this particular report for a single day the cookie user/host user ratio >was 3,218/3,235, which suggests to me that these count methods are not quite as >horrendously inaccurate as claimed by the "religion" Adrian refers to. Your mileage >may vary. > > >User Report >(Go To: Top: General Summary: Daily Report: Daily Summary: Hourly Summary: Domain >Report: Organisation Report: User Report: Referrer Report: Referring Site Report: >Search Query Report: Browser Summary: Operating System Report: Virtual Host Report: >Status Code Report: File Size Report: Internal Search Query Report: Request Report) > > >This report lists the users who requested files, if users have been authenticated >or can be identified by cookies. > >Listing users with at least 1000 requests, sorted by the number of requests. > > reqs: %reqs: first time: last time: user >-----: ------: ---------------: ---------------: ---- >22177: 44.12%: 23/Sep/02 04:59: 24/Sep/02 04:59: >aspsessionidggqqgxeo=ebpkmbjbnjdcoacfpdacpmfn >28083: 55.88%: 23/Sep/02 05:00: 24/Sep/02 04:59: [not listed: 3,218 users] > > >============================================ >James Riemermann >MN Office of Tourism >651/297-2077 >[EMAIL PROTECTED] Adrian Dolling, Systems Consultant Public Library Services Branch Ministry of Community, Aboriginal and Women's Services PO Box 9490 Stn Prov Govt (mailing address) Victoria BC V8W 9N7 800 Johnson St, Victoria, BC (street address) mailto:[EMAIL PROTECTED] http://www.mcaws.gov.bc.ca/LGD/public_libraries/ Tel: 250-387-4043 or 1-800-663-7051 FAX: 250-953-3225 +------------------------------------------------------------------------ | This is the analog-help mailing list. To unsubscribe from this | mailing list, go to | http://lists.isite.net/listgate/analog-help/unsubscribe.html | | List archives are available at | http://www.mail-archive.com/analog-help@lists.isite.net/ | http://lists.isite.net/listgate/analog-help/archives/ | http://www.tallylist.com/archives/index.cfm/mlist.7 +------------------------------------------------------------------------
