An interesting thread at Cre8asiteforums, titled “When lots of your visitors go straight to search? discusses a member’s curiosity about navigation patterns after noticing that a significant percentage of his visitors — 25% — go directly to search after arriving at his site.

It’s an interesting element of site navigation to investigate, and the thread raises a pretty significant number of additional questions worth analyzing.

The navigation path of any given user will be fundamentally unique. However, when taken in aggregate, navigation paths begin to suggest a lot about your navigation structures. The percentage of visitors who immediately jump into a site search, however, suggests a very different thought process.

On sites which I visit frequently, for example, I generally have a very set system for finding information. If the site has a good search, then I may use the search. If they have a very clear navigation, I may use that, instead. If they have neither, it’s unlikely that I visit the site frequently…but if I do, I generally have separate bookmarks to the individual features which I actually use.

And that raises a separate question — if a site has difficult navigation and inferior search, what would drive you to actually visit it frequently? For me, the site has to offer some specific information or functionality that I simply can’t get elsewhere. Otherwise, there’s simply no justification for the challenge of using the site.

You can learn a lot about the effectiveness of site navigation by following analytics data. Knowing that most users who use your search feature fail to find what they’re looking for, for example, should suggest that this is a feature of your site which needs work. Finding that users frequently enter several sections of your site before finding the right information can be significant as well — it suggests that you need to rethink the way your site categories/sections are organized.

So…important question, then: HOW do you follow this? Where do you get this information?

You’re not going to get meaningful user information from standard statistics packages like AWStats or Webalizer. You need to use some tool which will provide a means to track the path of specific users. This can be parsed from your server logs. A high-end statistics package such as Clicktracks will give you user path data. There are a number of other services which can provide this information (and, if you know what you’re doing, you’ve already got the information).

I’m not really an expert on analytics packages, of course. If you want a lot of detailed information about web analytics and analytics packages, here are a few resources: