Published: Aug 12, 2008
"Why do you think your site map needs a redesign?" I asked the web manager of a mid-sized government agency.
"Our logs show that one out of six visitors is getting to the site map page. It's something we haven't paid any attention to and we think that it could be much better than it is."
One out of six users is a lot, especially for a site that garners more than a million visitors every month. It's natural the team felt concerned about the 166,000+ visitors the page is attracting and would want to make it the best they could. The only problem is that our research shows that effort could be wasted, since we've realized working on site map improvements is a design cop-out.
Don't get me wrong: I think it's a noble goal to make every aspect of the design as perfect as it could be. That's all this manager was asking to do. In addition, working on the symptoms often requires substantially less effort than the research, followed by trial-and-error process, that you'll need to fix the source problem. There are good reasons to fix symptoms.
Yet, a team has limited time and effort to put towards design changes. Spending those constrained resources on the symptoms only delays the inevitable. A design cop-out usually needs continual updating, whereas fixing the root problem can nip it in the bud and release long-term resources.
A couple of years ago, we had a participant come into our lab with a newly discovered medical issue: chemical sensitivity. They'd been suffering from increasingly uncomfortable skin rashes and a friend told them that Dove had a nice line of products that could help.
Therefore, we decided to visit Dove.com. At the time, the only substantial links on the home page were generic, awkward marketing contributions, such as Your Body, Your Hair, Your Skin, and Real Beauty. The participant needed all these things, so it was hard to know where to start. Because of her ambivalence, she ended up clicking on the link labeled Site Map.
The site map was quite different from the home page. There were close to 50 links, mostly names of products. These were categorized into groups, with an entire group labeled "Sensitive Skin." Our guest clicked into that group and found what she'd been seeking.
At first glance, it might seem this was a successful task. After all, the user found what they wanted. However, the home page practically failed them -- with the site map being the only hope. (It's interesting to note that Dove.com has since undergone a major redesign, rendering the site map similarly vague. All the specific product links are now gone, replaced by generic links like Hair and Face.)
These days, there are two predominate ways that users get to a web site. Either they type the URL into the address bar, bringing them to the site's home page, or they come to the site through an aggregator or referral, such as Google, often taking them to a specific page within the site. For some sites, the home page is the most popular route, but increasingly, users link deep into the site.
What's interesting is that neither method is likely to bring them straight to the site map. Like our Dove user above, they'll hunt around on another page, searching for the trigger words to their desired content.
By itself, "site map" doesn't give off scent -- the clues that tell the user if the link will lead them to their desired content. It's only in the absence of anything else that gives off scent that users start to think it's a likely help. Therefore, the real problem is the pages that lead to the site map are missing important scent. Fixing the scent issues on those pages will eliminate the need for the site map. However, deciding to improve the site map doesn't fix the scent problem -- it's only a cop-out.
Site maps have their place and we're not suggesting that you dump them entirely. Search engine spiders crawl through the site map to find the pages that are hard to find otherwise on your site, so you wouldn't want to eliminate them unless those pages are now accessible some other way. (Of course, if the spiders can only find them thru the site map, that probably goes for the users too, and therefore may be indicative of a bigger problem.) We're only suggesting you work hard to make them redundant.
If you're seeing a large number of visitors going to your site map, that's indicative of a scent problem. Over the years, we've recommended putting the site map links on the home page. In many cases, creating a link-rich home page can solve much of the problem, reducing demand on (and thereby the need to maintain) the site map.
Looking at your site analytics can help you figure out where the traffic to the site map originates. If it isn't coming from the home page, you'll need to locate the pages that are generating the visits. By tracking the user's path from the source page through the site map to the destination page, you can start to piece together clues about the scent that's missing from the specific pages. Working diligently to eliminate site map traversals will improve the user's experience.
Ideally, you're dedicating some of your time to watching users on the site. Watch enough users and, if you have a problem with scent, you'll see them jump to the site map. When we see this, we take that moment to explore what they're looking for and how we could change the source page to provide a direct link.
If your users ask for site map (or to improve your existing map), it's only because they don't know to ask for better scent on the source pages. In the hundreds of usability test sessions I've now personally observed, I never once had a participant turn to me and say, "I loved this site. I found what I needed and it worked great, except for that site map page, which is why I'll never use the site again." The site map is there only to save it from total failure. Fix the problem and the site map is no longer necessary. Don't fall for the cop-out.
Have you done something special with your gallery pages? How have users responded? We'd love to hear what you've been doing. Leave a comment on the UIE Brain Sparks Blog.
Read related articles: