Web Analytics: Page Tagging

Have you ever heard of a program called Google Analytics?  It’s offered by some little online search company for absolutely no cost, just out of the kindness of their hearts…Or maybe the web giant is providing the service to help sell more ads, who can tell, either way it’s FREE.
Google Analytics uses a method called page tagging.  Little pieces of code are inserted on each page of website.  This code registers the actions of the website visitor, sending them off to an external server.  The analyst then just opens up their GA account to see the results.
Unlike logfile analysis, page tagging is user focused.  Since the code transmits whenever the user accesses a page, it doesn’t matter if they are viewing the page from their browser cache or downloading it from the server.  Non human visitors are also not likely to be picked up by a page tagging service, since bots do not tend to run the code.  Unique users are determined with the use of cookies, so individuals sharing an IP will register as unique visitors.  With page tagging you also have the ability to collect some pieces of information not registered on a standard logfile.
Ok, now for the negatives.  People sometimes use multiple browsers, clear their cookies, or choose not to collect cookies in the first place meaning it’s easy to get duplicate counts.  The longer the time period, the more likely you will see the duplication, i.e. unique users in a month vs unique users in a day.  Data is stored on a third party server, one that you don’t control, which can be an issue for some users, although it’s increasingly becoming harder to do everything on one’s own server.  You also have to make sure the codes are properly set on each web page, and that they are registering properly on the analytic’s server (note: it could take a couple days for everything to start registering properly).
All in all, for researchers and evaluators page tagging is likely going to be your tool of choice.  There is nothing keeping you from using both though, if you want the comparison.  Just understand there are quirks with both methods and report accordingly.