Web Site Usability: The Big Picture
Spool worked with a team of software designers who attempted to come up with an objective standard by which Web site design could be assessed for effectiveness, and tested a variety of different sites: news sites, game sites, shopping sites, financial information sites, magazine sites, travel information, etc.
The tests were "scavenger hunt" tests in which subjects were asked a variety of questions for which they would need to use a Web site to find answers. For example:
- Simple facts - How much does an item cost?
- Comparisons - Which of these two items is cheaper?
- Simple Judgment - Is that item worth its price?
- Complex Judgment - Which of these items is the best value?
Subjects were observed, and the researchers noted the time and effort it took to answer questions such as the ones above - or in some cases, for users to get "lost" and give up their pursuit.
A bias in the study, which is common to many such studies, is that the test subjects were likely college students, so the sample is biased toward that demographic.
Regardless of the question and the site, they arrived at six major implications:
Implication 1: Graphic Design Neither Helps Nor Hurts
There was no evidence that the visual design of the site had an effect, either way, on the ability of users to find the content they needed. A text-based site with minimal decoration scores as well or better than sites with eye-catching professional designs.
Implication 2: Text Links Are Vital
Most users sought out (and clicked) text links more often than image links, and the more descriptive a link was (in that it indicated where it would take users), the quicker users were able to identify and follow the route to the information they sought.
Implication 3: Navigation and Content Are Inseparable
The "shell strategy" of site development is to develop the structure of the site first and shovel in content later. The problem with doing so is that navigation categories are vague and ill-defined, and often incongruous with the content filed underneath them.
It's also stated (assumed) that a shell-site evolves clumsily - new information that doesn't fit the navigation is "stuffed" into places it doesn't really belong, and where users can't find it.
Implication 4: Information Retrieval Is Different than Surfing
The concept of "surfing" the Web, with users nonchalantly clicking things at random to find bits of interesting content, is overused, and it's obvious that designers pander to it by attempting to "grab" the interest of a user who's just surfing along.
A user who is actively seeking information is much more focused and deliberate, and the things done to attract "surfers" are distracting and obstructive to information-gatherers.
Implication 5: Web Sites Aren't Like Software
Since Web sites are experienced on a computer, some firms design and test them in the same way they design software. It's clearly different.
In software testing, the results generally indicate that users choose a product that gives them the broadest array of capabilities for doing a variety of tasks, and will try harder to "learn" to software if they encounter difficulties.
With Web sites, users prefer those that provide them the quickest path to accomplish a very specific task, and will give up and abandon their task a lot quicker than they will with software.