June 4, 2001

Website navigation is useful

Summary: although users tend to navigate websites by search mechanisms or by links embedded in actual content, website navigation serves useful purposes.

In one of his Alertbox columns, web usability consultant Jakob Nielsen asks the question, "is navigation useful?" In other words, does a "menu" of some kind make sense for web users? Nielsen concludes that navigation may be useful, but that there are too many other problems with the web to make it useful most of the time.

Nielsen's articulations are very pessimistic, and he makes some sweeping suggestions that, without further qualification, are terrible ideas. For one, Nielsen says: "do not link to all sections of the site from all pages". Well, what is a "section"? Why not show users what's on the site? Isn't there a situation in which that can be useful? Of course! Nielsen backs up his argument with a ridiculous example: "what is the probability that a user will go from looking at hairdryers to looking at grunge music?" Not likely, admittedly, but few websites offer such a choice. A more reasonable example: what is the probability that a user will go from looking at "hardware" to "software" or "support" or "contact information"?

How users get around websites

Users can arrive at any page within a website by a number of means: a link, a search engine, a directory, a software program's "about box", typing a web address directly, and so on. There is often no way to predict how a user will end up at any website. Therefore, web pages can't be isolated from the rest of the site if the site is to be successful and have lots of users. Visitors need some way of getting around the website:

There are other ways of getting around a website, including constantly visible site maps, 3D visualizations, and context-sensitive relations, all of which remain fairly uncommon today. Navigation is at issue here, particularly because Nielsen, one of the most-read web usability "gurus", might be giving designers the wrong idea.

A search mechanism is a must for fairly large websites, or for those that can't be assigned an easily-understood architecture (in other words, a simple site map). Many users head straight to search in order to find what they're looking for, especially if the users perceive that the website has a complex structure that is difficult to navigate. Sites with lots of text and links, but no clear navigation that indicates the scope of the website, are likely to be perceived as complex.

Hypertext navigation is what the web was originally all about: providing content authors a way to connect to other content, by letting users click on key words or phrases. Links are still a key way of getting around; in fact, links are the most important and most common way of users getting around a website. However, links don't provide an indication of the breadth of a website, nor do they necessarily show users all that may be relevant.

Nielsen's own website doesn't really have navigation. Instead, it relies mostly on well-implemented hypertext linking, and users can get around most of the website simply by clicking on links in the content. Nielsen's website (which is surprisingly complex, and its home page difficult to decipher) gets away with basically no navigation, because Nielsen understands hypertext linking. In Nielsen's ideal world wide web, sites wouldn't really need any navigation other than a "contact information" link on every page. In the real world wide web, sites need navigation. Hypertext is under used, sites aren't ultra-specific, and search is unreliable.

What is "navigation"?

Navigation is basically a menu mechanism: something that lets the user jump from one part of a website to another. On the merges.net website, every page has navigation: a menu that users can take advantage of for quick movement around the site. Everything on the site is divided into a few content areas, and the navigational menu gives users access to each content area, but doesn't overwhelm them with choices.

Essentially, navigation can be any mechanism that constantly gives users access to certain sections of a website. It's very difficult to incorporate navigation into a paragraph of text and to use hypertext links, but it's easy to create a "menu" like the one on this site.

Why navigation is useful

Navigation, which can be made too complex if not carefully managed and restricted, is tremendously useful, even though Nielsen claims that users don't really notice it. Here are a few reasons why navigation is useful:

When designers are creating a website, developing navigation is a task that should be carefully organized, and the results tested according to the purposes of navigation. If navigation is reasonably designed, people will use it. If navigation tries to show too much, people won't use it, because it will contribute to their perception of the website as "complex".

On this website, merges.net, there are no text links to the home page, client extranet, or portfolio sections of the website on any other page. Therefore, the only way to get to those sections of the website are via the navigation menu. It turns out that most users of merges.net enter the site at a specific theory article, then click on the "home" button in the navigation menu, or on the logo in the top left corner. A large number of users also click on the "portfolio" button. Navigation is useful, because it offers users a way of getting to pertinent information without having to wade through content that they may consider irrelevant. Where would I put a link to the portfolio on this page?

Presenting scope

Navigation should show users what's on the website, but not highlight specific content. A complex home page with headings and text that describes everything on the website is no good; users won't sit at their screens and read all of that. Plus, users that find their way into the site on some page other than the home page won't be able to get an idea of what's on the site unless there is some kind of navigation available. A simple home page with some scope-indicating text is fine, but it should also have navigation, like every other page on the site, that clearly shows what's on the site.

In the next revision of merges.net, the navigation will have to be improved to more clearly explain the scope of the site. Right now, some of the menu titles are ambiguous. What kind of services does merges offer?

A fallback mechanism

All search services are inadequate, though some are better than others, and hypertext is frequently incorrectly used or insufficient for every user. When the user can't click on links or do a search to find what they're looking for, they need something else. Navigation is, if not a first-choice for many users, a fallback for all users. Navigation can help users decide whether or not they are likely to find what they're looking for, whereas hypertext and search don't act as such a knowledge compass.

When all else fails, users can click on "contact info" and ask a question, because that link is clear and part of the navigation.

Reducing cognitive overhead

Navigation tends to be very simple to comprehend: few words combined with decent pliant response are generally a recipe for easy understanding. Hypertext links, which are embedded in content, and search mechanisms, which produce lists of hypertext links embedded in content, require lots of cognitive processing. If there is no navigation on a website, users have to read or scan pages, and pick out keywords or phrases that may be different from page to page. That's complicated, and since each page can be quite different from another, finding patterns can be tough.


Navigation saves the user from some frustration. Certainly, websites cannot rely strictly on navigation and expect to be successful (it is still used less than regular links, and possibly less than search mechanisms), but it is necessary to help users who find the other methods of getting around the website inadequate. Additionally, navigation helps users understand the website as a whole: what's on the website, and how it's organized. Carefully plan navigation—make sure it accomplishes what search and hypertext can't—and it will save users again and again.

Adam Baker is a user experience designer who's worked at Google, Apple, BlackBerry, and Marketcircle, and mentored startups in Vancouver.