Visitor-determined website improvement: combining survey and log results on Cochrane.org

Article type
Authors
Booker D, Horstmann J, Saunders G, Royle N
Abstract
Background: The hundreds of thousands of people visiting cochrane.org each year are anonymous. This presents challenges as we attempt to tailor content and features to our actual, as opposed to intended, audience.

Objectives: To prioritize development of Cochrane.org so that features are targeted accurately to the needs and goals of website visitors.

Methods: Analysis of logs of access to cochrane.org was performed with AWStats, a program which extracts geographic, hardware/software, and session analyses about website visitors from server logs. An online survey, developed outside the Cochrane Web Team, collected information about visitors' goals and experiences with the site. For a two-week period in March 2006, visitors to the site were presented with an invitation to participate in the survey. Participants needed about 15 minutes to answer about 40 free-text or multiple-choice questions about the site.

Results: Data from nearly 2,000 survey participants revealed:

- nearly half were first-time visitors;

- more than 80% described themselves as healthcare professionals;

- 68% visit the site to find Cochrane reviews;

- 40% visit the site to get to The Cochrane Library;

- 25% visit the site to read the news, itself a new feature on the site;

- navigating through the site is a common difficulty.


Analysis of server logs revealed:

- many visitors use Google or Yahoo! to find specific pages on the site, and then stay only for a short time;
- visits were made from nearly 100 countries;
- self-reported behaviour of the survey was confirmed.


Conclusions: Combined log and survey data make it possible to pin-point parts of the site which are: important to visitors; already helpful; still difficult to use; or deserving further development and visibility due to their popularity. Such data also highlight the international usability needed of such websites.