To the front pageThe Interaction Designer's Coffee Break - Weekly postings and quarterly articles about interaction design  
  To the front pageSign inTo the frontpageSearch in GUUUI postingsAbout GUUUI  

ISSUE 08 - Q4 2003

Balancing visual and structural complexity in interaction design
How visual simplicity can harm usability

Usability is based on principles such as "Less is more" and "Keep it simple, stupid". But there is more to simplicity than meets the eye. By reducing visual complexity at the cost of structural simplicity, you will give your users a hard time understanding and navigating the content of a web site.

In their effort to make things simple and user friendly, designers often try to reduce the visual complexity of web pages. They cut down the number of menu items, hide them away in dropdown menus, move related content and details to other pages, and split articles into multiple pages. The reason is often an aesthetic one, but designers find support in widely held beliefs about users, such as:

  • Users can only manage a certain amount of information at a time
  • Users don't want to wait for things to download
  • Users don't want to scroll

While such statements hold a grain of truth, they are oversimplified, undifferentiated, and at worst misleading. Taking them too literally will most likely do harm to usability. It's true that the more simple a page looks, the easier users can find information on it. But reducing visual complexity to make things pleasing to the eye by hiding critical information from users will inevitably increase structural complexity, and make it difficult for users to grasp and navigate the site.

In the following I will show some common pitfalls, where studies have proven that what appears to be simple isn't always what is easy to use.

7 +/- 2 items

One of the most misleading arguments used in favour of reducing visual complexity is the rule of 7 +/- 2. The rule states that the human brain can't handle more than 7 +/- 2 items at a time. If you apply the rule to visual design, it would mean that things such as lists of menu items or items in a bulleted list can be no more than nine.

The trouble with this rule is that the psychologist George Miller who formulated it was studying the limitations of short-term memory - not limitations of what people can perceive visually at a time. Humans can only retain 7 +/- 2 items in the immediate memory, but have no problem in dealing with great amounts of information in the field of vision. As long as you have information present for continuous reference, immediate memory plays no significant role in your perception.

The rule of 7 +/- 2 can be quite harmful when applied to navigation. On the surface it might seem reasonable that reducing the number of menu items of each web page will make it easier for people to navigate. But this is not true. Reducing the number of menu items will make the site hierarchy deeper and thereby increase structural complexity. Research has shown that users generally find information faster in broad and shallow menu architectures than narrow and deep ones. Roughly 16 top level links leading into 2-3 subsequent menus seems to be the most efficient and least error prone.

Download time

Another argument for reducing visual complexity is download time. The more content on a page, the longer it will take to download it. It's true that people don't want to waste their time, but fact is that perception of download time and the actual time it takes aren't strictly related.

A study by the usability consultancy UIE on how people perceived download time at 10 different sites showed that there was no correlation between the actual download time and the perceived speed reported by users. Instead, UIE found a strong correlation between perceived download time and whether the test participants successfully completed their task. Thus focusing on supporting users in finding what they are looking for is a better strategy than reducing visual complexity in order to make pages look clean and load fast.


Users and clients tell us that they don't like to scroll, and many designers will go far to make their pages fit nicely into the browser window. Their view has been supported by Jakob Nielsen's "Top Ten Mistakes in Web Design" since 1996, where he reported that only 10% will scroll beyond the information that is visible on the screen.

But according to Jakob himself this is no longer true. Studies of his from 1997 have shown that most users are perfectly willing to scroll even long pages, and Jakob has declared that scrolling is now allowed and "no longer a usability disaster."

More recent studies from UIE and the usability research laboratory SURL support his findings. According to UIE users are willing to scroll if the pages give them strong clues that it might help them find what they are looking for. And research from SURL has shown that scrolling through articles is significant faster than reading articles split into multiple pages. The test participants stated that they found such articles to be "too broken up", and where frustrated by having to go back and forth to search for information.

The prejudice against scrolling is that information is hidden from users. But as UIE rightly states, "Short pages may avoid this potential problem by showing more (or all) of an individual page, but the information is still hidden — on other pages." By hiding information on other pages, you will inevitable increase structural complexity.

Product lists

Hiding critical information isn't just annoying for users. It may also have an impact on the bottom-line of e-commerce sites. Another study by UIE has shown that when product lists provide insufficient product details, people will buy less.

In their study, UIE observed that when product lists provided enough information for the test participants to make selections they where five times more likely to add items to their shopping carts, than when they had to click back and forth between product lists and product description pages to make up their mind. They also found that the participants who didn't find enough information in the product lists where one-third more likely to quit shopping and had lower opinions of the sites tested.

Striking the right balance

Usability has become an important aspect of many web projects. For people with little experience in interaction design it's tempting to rely on what you see and what people tell you. Clients, co-worker, managers, and prospective users will want simplicity and will complain about information overload if your screens look crowded. But as we have seen, there is more between heaven and earth than meets the eye.

For the interaction designer it's a question of striking the right balance between visual and structural complexity. People will complain about a visually complex page at the sight of it. But they will also complain if the information they need isn't immediately available to them when they start using the site. It's the good old story about the difference between what people say and what they do. If a site isn't solely for visual pleasure, you should rely on user behaviour and not on what people tell you. Use findings such as those mentioned in this article as guidelines and test your designs with prospective users.

Text: Henrik Olsen

"People will complain about a visually complex page at the sight of it. But they will also complain if the information they need isn't immediately available to them when they start using the site."

Posting added before you last visit at GUUUI Product lists' impact on sales
Posting added before you last visit at GUUUI First Rule of Usability? Don't Listen to Users
Posting added before you last visit at GUUUI Breadth vs. depth in menu design
Posting added before you last visit at GUUUI The truth about download time
Posting added before you last visit at GUUUI Scrolling is faster that paging
Posting added before you last visit at GUUUI Scrolling may be the best approach for users
Posting added before you last visit at GUUUI What users say vs. what they actually do
  To the front pageSign inTo the frontpageSearch in GUUUI postingsAbout GUUUI