Sort of a captive audience.
For one of the sites for which I've viewed statistics there's a very legitimate
reason to collect the information. A central goal of the site is to make information
available to readers, and if they're not coming to the site, changes could most
certainly be called for. (Though that's true, there's also a good deal of logic
in the claim that many of the changes that might be made should be focused more
on helping people find the site than on staying on board once they do). The other
site, however, is a collection of online classes where the "classrooms"
of the registered students are located. We try to design an interesting and engaging
online experience, but even if all we had to offer was texts that the students
have to read and assignments to be performed and then submitted, it's unclear
that statistics on use could influence anything. We know how many students should
be logging in to each class, and if they're submitting their assignments (something
that doesn't need web statistics to be determined) it really doesn't matter how
much time they devote to each page. We might be able to learn what assignments
keep students on a particular page longer than others,
but this should be a function of what's being taught, rather than of statistical
analysis of page hits.
Go to: Still running it up the flagpole