Web Usage Reporting
Originally, DKV was asked to report on usage of Columbia Interactive (CI). Specifically, we needed to provide statistics on unique visitors, page views, and other metrics. This information would be used to determine the success of various promotional and marketing efforts, the popularity of different resources (to see which areas we should be developing further), and the priority for design or navigational changes. After some research, we concluded that WebHound was the program we wanted to use to obtain these figures. In strong collaboration with AcIS, we obtained a license for WebHound, installed it, and began using it. As we did so, we found other uses for WebHound within DKV and across the university. At this point, EPIC began developing their own set of reports going back several years. In general, we aim to serve the University's needs by sharing our knowledge of WebHound and its capabilities with staff members at other University web publishing initiatives.
To determine the success of Columbia Interactive, DKV wanted to get a baseline on usage of the site and all of its areas and then track usage over time. We also wanted to make changes and improvements and run promotions and then see if these initiatives led to increased interest. Specifically, we were interested in the following:
- Total visitors
- Unique visitors
- Total page views
- Average session duration
- Top 10 course description pages (i.e., descriptions of our e-seminars)
- Top 10 referring URLs
- Top directories accessed
We also wanted to take advantage of the dynamic reports that WebHound provides, such as:
- Visualization: to see a visual representation of common paths through the site
- Drillable: reports that allow someone to view data by more than one variable (such as entry points by referring domains)
Some of these reports were available out of the box but many involved custom development.
For EPIC, these reports were developed as part of a larger project to determine the success of EPIC's publications (see EPIC Online Use and Cost Evaluation). Some of the customization for this purpose involved generating reports by semester, and producing reports for multiple websites at the same time.
We have reports on CI that date back to January 2002, and we have been running reports on a daily basis since April 2002. EPIC has reports that date back several years. In parallel, AcIS is using WebHound to generate reports for CU as a whole. It is DKV's intention to obtain reports for CAHO, CERO, and future DKV knowledge centers. DKV is also in the process of building reports that provide information on a particular subscribing institution to our knowledge centers on that institution's users.
There are many possibilities for future uses of WebHound. Some development will be done within DKV and EPIC, but we will also provide assistance and advice to other CU staff members to leverage this software for their online initiatives.
Carol Kassel (DKV)
Susan Maloney (EPIC)