Scoping Library Website Analytics
I had a chat earlier today with a couple of folks from the OU Library about what sorts of questions we could ask about user behaviour on the OU Library website if something like Google Analytics was running on that site. I didn't take any notes at the time, but here are some of the take home points I remember about low hanging fruit and potential segmentation of analytics according to perceived user groups:
- What are the most popular areas of the site? Are they reflected by the top level and front page navigation schemes used on the site?
- How would a site overlay of click-thru behaviour on the front page compare with the anticipated/expected/desired behaviour?
- Would analytics reveal any periodicity in activity on the website, on a daily or weekly basis? Would this mirror activity in calls to the Libray support desk, e.g. via the "Ask a Librarian" service?
- Would it be possible to identify course related activity? For example, where courses include a particular library website activity, would it be possible to define analytics filters to capture evidence that this activity was being completed by students, and to what extent? Could a goal page relating to a goal state/page within the library be defined for one or more course related library activities (such as an exercise in the info skills course TU120 Beyond Google;-) allowing user behaviour leading up to that goal page being captured? What sorts of analytics might a course team be interested in receiving from the library regarding library resources linked to from a course (e.g. on a course page in Moodle?)
- Library users can potentially be segmented in various ways - for example, by IP range (which would identify on-campus or regional users). Would it be possible to use filters to report on these user groups, and what use might be expected? Using an OTS analytics tool such as Google analytics, tracking users against authentication would not be possible. However, would it be possible to segment users based on their behaviour? Or could landing pages/entry pages be identified for different users - on-campus researchers, for example, as opposed to on-campus course developers and/or students, or off-site students?
- What sort of search terms deliver users to the OU Library website from organic search listings and what sort of activity does those users engage in?
- Would it be possible to reconcile exit page analytics (e.g. from particular database listing pages) with reports from accesses to external, licensed services?
The above list is just a first pass, of course...
If anyone out there is using, I'd be interested in hearing what your favourite report views are, and why... ;-)
If anyone out there is thinking of using Google Analytics to explore user behaviour on library websites, what sort of questions would you want to use analytics to help you try to answer?
How to define library website goals is another interesting exercise... If the site was Amazon, where the aim is to sell goods, a relevant goal page would be a "Thanks for the cash - the goods will be with you in a day or two" page. What is the range of useful, successful transactions on a Library website? And waht are the corresponding 'goal pages'?
[Comments can only be easily added to this page from a comment link for this post on the front page of the blog (this cuts down on the spam! Feel free to mail me with comments - a dot j dot hirst _AT_ open.ac.uk, and I'll append them to the post...]
Blogged with Flock
Tags: library, analytics
Posted by ajh59 at November 2, 2007 12:28 AM