Usability studies of WWW sites: heuristic evaluation vs. laboratory testing

This paper describes the strengths and weaknesses of two usability assessment methods frequently applied to web sites. It uses case histories of WWW usability studies conducted by the authors to illustrate issues of special interest to designers of web sites. The discussion not only compares the two methods, but also discusses how an effective usability process can combine them, applying the methods at different times during site development. PREREQUISITES FOR ASSESSING WEB SITE USABILITY The two methods discussed in this paper for assessing the usability of web sites both require the usability specialist to have three vital pieces of background information: the purpose of the web site; profiles of its intended users; and typical scenarios for users accessing the site. These elements are equally important in evaluating the usability of any product or service, but here is how they apply especially to Web site evaluation. • When discussing the purpose of a web site, it’s helpful to consider three categories. Web sites that supply descriptions of companies (or other organizations) and their products, services, informational offerings, or events can be described as informational sites. Web sites that provide explicit links to extensive databases are called search sites. Web sites that behave like products, where users perform other tasks in addition to reading or retrieving information, are referred to as transactional sites. Multi-purpose sites blur these boundaries. • Unfortunately, the definition of user in our increasingly Web-centric environment is becoming more vague, because “anyone can access the site.” However, we must keep in mind which site visitors are the most likely—or the most welcome—and focus usability efforts on those subgroups. • Finally, the scenario for accessing a web site might be a straightforward URL to a home page, or a more roundabout path through a link in search-engine results to a page deep in the bowels of a site. Evaluators should keep in mind that any web page might be the user’s door to that web site. Although users may perform more complex tasks in transactional sites, the free-form nature of navigation in any type of web site makes ensuring (and measuring) success more complex in the Web environment. Home