By Patrick Hunt, President and CEO of Strategux Consulting, Inc. (Published Jun. 21, 2004)

In the mid-1980s, when Late Night with David Letterman was on NBC, Dave occasionally stared into a tight close-up with big, buggy eyes and proclaimed, “I’ve been hypnotized.” That was the entirety of the skit, and it was really funny at the time to this then undergrad.

Today, there’s something not so funny going on with website users of all backgrounds and levels of experience with the web. The popularity of Google and Yahoo—along with other search sites MSN, AOL, and Ask Jeeves—is strongly influencing user expectations for the search results of other sites, to the extent that major departures from the designs of the two most trafficked sites cause usability problems. To paraphrase Mr. Letterman, “We’ve been Google-ized.”

In this column, I’ll relate some usability testing experiences that illustrate this phenomenon, and make a few recommendations for how to address it.

In the last year, I facilitated, observed, and analyzed many rounds of usability tests for large corporations, non-profit organizations, and government agencies. Several tests were specifically designed to gather user preferences for search results, while others were more broad-based studies in which participants often used search tools voluntarily.

During one quick round of testing, we attempted to gauge whether a redesign of a search results page was better or worse than the design in production at the time. Not surprisingly, the qualitative sessions revealed that users relied heavily on Google and Yahoo as their primary web search engines, a fact supported by official web traffic and utilization statistics. Based on their choices, we probed to learn more about what they wanted to see in search results, and concluded that Google or Yahoo-inspired designs would be most effective.

Luckily, my client wanted to validate these conclusions through another round of tests. In this iterative usability study, we attempted to gauge preferences among four design candidates, including the preferred version from the earlier sessions, Google and Yahoo-inspired designs, and a fourth layout. After a series of qualitative questions about each design, we asked participants to rank the four and tell us what one thing they would change about their favorite.

In every case, the Google and Yahoo-inspired designs were ranked first or second, and user comments suggested that the other two designs were not even close. In most cases, users identified changes to the Google version that would make it look more like the Yahoo-inspired design, and vice versa. Ultimately, users appeared to want results that looked like Google’s—with a clear emphasis on the search results and supporting material at the top and right—with obvious markers for the page sections. (On Google, the page sections that accompany the results are sponsored links and text ads; on other sites, these supporting materials could include the top requested documents, as an example.)

After reaching these conclusions for the specific site in question, I analyzed results of usability tests for other clients that featured interaction with or questions about search. I also observed user behavior during subsequent tests on other sites. The commonalities were striking. Virtually every user expected search results to appear and function like Google or Yahoo’s, and even minor deviations caused problems for most users. It appears that the designs of the popular search engines are significantly influencing user expectations for search results on other sites, including Internet sites, intranets and extranets.

What to do about it

The limitations of this medium prevent me from covering much of the detailed analysis, so some of these recommendations may seem like a big leap. But they are grounded in my deeper assessment of many usability tests. Here are a few things you can do to ensure your search results meet user expectations. Some may seem obvious, but it is easy to overlook them when dealing with the many competing factors in managing your website.

1. Place the emphasis on the results. If your search results page includes anything other than the results themselves, ensure that the results comprise the most visually prominent section of the page.
2. Create a clear hierarchy for the entry detail. The design of the results entry should emphasize the main document title as the link, and include a brief description, file type/size, and URL as well. Users should be able to quickly scan the titles on the results page, and read or click those that are of interest.
3. Rank results in order of relevance. Put the most relevant result at the top of the list unless users specify another sort order.
4. Conduct your own testing. There’s no substitute for getting your own users to provide feedback on your unique search results page.

Conclusion

Website visitors use search engines for a very specific reason: they want what they want, without a lot of hassle. Create a simple, clean search results page that is easy to scan and places the result most likely to match user needs at the top of the list. Google, Yahoo and others do it, and your users expect it. They will thank you for it, and you’ll likely improve the performance of your website.

Patrick Hunt is President and CEO of Strategux Consulting, Inc., a boutique Internet consulting firm specializing in user research, information architecture, and usability testing for corporate, government, and non-profit clients. He can be reached at patrick@strategux.com.

Back to top
Current Digitalharboronline Columns Page