Web Site Expert巻頭レポート(英語)

Keep it Simple: A Conversation with Usability Expert Jacob Nielsen

The first thing you notice about Jakob Nielsen’s website, www.useit.com, is that there are no pictures: just text and hyperlinks. He does use a bit of color?the right hand column is a pale yellow, the left a pale blue. But nobody is going to give the site any design awards, and that’s exactly Nielsen’s point: in Website design, being pretty is not the same as being good. Not that Nielsen recommends his austere approach to everyone. ⁠Its an extreme example of what I talk about,⁠⁠ he says.

What Nielsen talks about is the value of keeping Web design simple in order to make websites more user-friendly. He doesn’t object if a site looks attractive, but too often, an aesthetic sensibility overrides usability?and that, says Nielsen, is the cardinal sin of web design. The message behind his own website is this: in choosing a balance between simplicity and aesthetics, err on the side of simplicity, every time.

Nielsen isn’t just making this up. He has adopted a research technique called eye tracking that shows, literally, where people are looking. From that, he can tell which parts of the screen receive a lot of attention, a passing glance, or aren’t seen at all. The tests can be a humbling experience for Web designers when they discover, for example, that the lovely graphic they inserted on the page is completely ignored, while the viewer scans the hyperlinks in search of something else.

Nielsen’s background is as unusual as his approach. He was once a Distinguished Engineer and Sun Microsystems, a position that gave him a lot of room to follow his interests. That was back in the early 1990s, when Sun was trying to improve the user interface on its operating system. By 1994, he figured out that the Web itself represented the future of interfaces and started to focus his ideas more broadly. In 1998, Nielsen joined forces with Donald Norman, who thinks about usability as it applies to physical objects. The cover of Norman’s book, The Design of Everyday Things, shows an imaginary teacup-from-hell with the spout on the same side as the handle.

Nielsen thinks that Web designers can commit comparable kinds of sins. His own best-selling book, Designing Web Usability: The Practice of Simplicity, has sold more than 250,000 copies in 21 languages since its publication in 2000 (He updated it this year, re-titling it Prioritizing Web Usability. Nielsen's other books include Homepage Usability: 50 Websites Deconstructed (2001); International User Interfaces (1996), Usability Inspection Methods (1994), Usability Engineering (1993). Nielsen's Alertbox column (www.useit.com/alertbox) on has more than 250,000 readers.

Nielsen argues that good web design has never mattered more because, for many people, a company’s website is the very equivalent of the company itself. Customers may never talk to a service representative, never read a paper brochure, never visit a store. Their entire interaction may be via the Web. So if the website is difficult to use, that means that the company is difficult to do business with. You are your website and your website is you. So design it with care.

Perhaps your biggest contribution to Web design is in citing empirical evidence as proof of usability?not just relying on instinct. How did that come about?
It probably comes from my parents, who both were psychologists. While I was growing up, they did many scientific studies?on everything from monkeys to actual human beings. So I understood early on that you can learn things from experiments. Over the past 25 years, it’s been my firm belief that you can’t just imagine how people behave with computers: you have to find out. It’s amazing how few others actually do that.
People have a sense about the aesthetics of good design, but not necessarily how well a design will function.
Exactly. That distinction is often not made. You can know what looks good?a lot of people have reached prominent jobs by having impeccable taste. But that does not at all mean that you know what’s easy to use. That’s a completely separate question. In fact, while a good Web design has both of these qualities?ease of use and good looks?different people sometimes provide each of them.
And there’s a further problem: the people who are in charge of these projects are not themselves the average user. They are typically smart and highly educated. And, for sure, they know enormously more about their own project than the average outside person would. So there’s a big difference between what they think is easy and what an outside person would think. What is needed runs contrary to instinctive human behavior that dates all the back to the pre-Copernicus model of the universe. Which is to say we all think, on some level, that we are the center of the universe?because we observe the universe through our own senses. It’s just ingrained in human behavior that we ourselves are the measure of all things.
How does that self-centricity play out in Web design?
For example, for many people can’t believe that the average person doesn’t admire beautiful pictures on a home page. And yet eye-tracking experiments often show that nobody looks at them. But when we point that out at seminars, somebody in the back of the room will say: ⁠but our branding requires that we must include this pretty picture.”
So eye tracking lets you know, literally, where people are looking on the screen?
That’s the only thing it does. It’s very simple. Eye tracking measures what people are looking at. We then add something called the ⁠mind-eye hypothesis,⁠⁠ which says that people look at what they are interested in. This is not 100 percent true, but it is overwhelmingly true. If people don’t look at something, it’s because they are not interested in it ? that’s basically the point. So you can measure whether or not people look at these big pictures: and they basically don’t. I’m referring to the sort of glamorous, frivolous, feel-good pictures that people think are on a website for a purpose.
That’s another of the big differences between the insiders working on the website and the outside customers. When you work on a project, you admire it for it’s own sake. But if you are an outside customer, you only go to a website for a purpose. That purpose could be very broad: finding the day’s news on a newspaper website, for example. But if it’s a company website, people often have just a single purpose in mind: a product purchase, for example, or support. They are trying to get something done.
So what we call the ⁠user experience⁠⁠ is, in fact, the user’s experience. It’s not what you would wish them to do, or what you think you would do, or how you would interpret the page because you already know all the information. If I give you five things to click on, then of course you can do it quickly, because you are familiar with the design. But if you are an outside person, two things happen. First, you don’t know what these five things mean and therefore you often pick the wrong one. And secondly, you may want to do a sixth thing that the designers never even thought of putting in.
Do your tests ever show someone glancing all over the screen as if they were completely lost?
Oh yes, that definitely happens. That’s a characteristic of a website that is horribly designed, because it doesn’t have any structure for guiding people to where they want to go. Sometimes, people can make no sense of the page at all. And other times, they may be looking at the right area, but don’t know how to pick from the choices. That’s actually just as bad.
Do you think that Web designers do bad designs, or do good designs get derailed by others?
I think both. There are definitely some good web designers out there, but there are a lot of people who don’t really know anything about how users actually behave. So they are only focused on good looking designs, because that’s their skill?they came out of graphic arts or advertising. But advertising design has different goal, which is to grab people and make them look. Whereas, with a website, people are already looking.
And you can’t navigate an advertisement.
No. With an advertisement, you don’t have to do anything but admire it. Then hopefully you will get that warm, fuzzy feeling about the company. But of course for a website, this notion of an interactive environment is relatively new. It’s odd that I’m still saying this. If you and I had been talking 10 years ago, the idea of designing for an interactive experience would have been literally new. But even today, people still don’t understand that core challenge.
You’ve noted that some things have improved, like the disappearance of splash screens.
Yes, there definitely have been improvements. One of the things we have measured is called the success rate, which means the rate that people succeed at what they are trying to accomplish on some system. The success rate for websites is 66% percent.
That’s all?
Back in the 1990s, it used to be less than 50 percent. Now, only one-third of the time do people actually fail, which is still horrible, but it’s at least better.
What bad assumptions do people believe about their own websites?
The biggest is the question about branding, because marketing-oriented websites tend to be run by people who are firm believers in the concept of brand representation: a company’s image, logo and style. In the print and non-interactive world, that’s all you have?a representation. However, on the Web, the experience itself becomes the brand message. If the experience is pleasant, you think good things about the company. If the experience is unpleasant, you think nasty things: that the company doesn’t give good service or is difficult to deal with. You’ll notice that the sites that do a lot of business, like eBay and Amazon, are very attuned to the user experience. As companies, they ask themselves: do I treat customers well? Do I give them quick access to what they want? Do I have the features they need at their fingertips without overwhelming them with others?
They provide a good experience, and that experience becomes their brand representation.
That reminds me of some of the things you wrote about B2B [business-to-business] websites. For example, that some sites require contact information before you can get information.
That’s exactly right, and I think it’s because in the B2B world, the sales cycle is much longer, involving many more sales calls and steps before they actually get the deal. In B2C [business-to-commerce], on the other hand, sites like Amazon close the loop much more quickly.
You would think that designers would look at sites like Google, eBay, Amazon and Yahoo! for inspiration.
They do, but then they make the mistake of thinking they are different?that they just sell, say forklift trucks. But the difference between a major site and a small one is in the details, not the user experience. In fact, when we talk to people making B2B purchases, they see the problem. They ask: ⁠why can’t I get as good an online service when I’m about to spend $10,000, as when I’m buying a paperback book from Amazon?”
It’s tempting to say that sure, the major websites can be simple, because they are so widely used. But I think it’s the other way around: Because they are simple, they are widely used. Because they put so much power at the user’s fingertips, they get so much traffic. We forget that these companies were not always famous brands. They all started out small, but they have always emphasized simplicity.
Are you’re suggesting there is now a de facto web style applied across many websites?
Exactly. People get used to a certain way of things working. And one of the myths is that in order to be special, a site should engage people by giving them a creative new interface. This is completely wrong. The way to be engaging is for your products, services, and concepts to be engaging. A good analogy is the novel. They are written on pages of a book that you turn one at a time. You don’t jump ahead ten pages and then five pages back. And for the most part, writers use normal language and syntax. (It’s true that some poets try to break all those rules?but they only sell five copies.)
In other words, the just because a book’s ⁠interface⁠⁠ is standard doesn’t mean that writers can’t still tell completely different stories. That idea is accepted in so many other media. Yet for Web design, people still have trouble distinguishing between the interface and the content.
Does this hold true internationally? Does it hold true in Asia or Japan?
Yes it does. We have done some studies in Asia, including a few in Japan, and we find the same thing. You wouldn’t want to use the same website in Japan and, say, Germany. You would want to optimize it for the particular country. But aside from language, the biggest difference is what topics you emphasize. Right now, the World Cup is big news in Germany, but maybe less so in Japan?so how the scores are displayed may differ. But that doesn’t change the general principle that you should put the biggest score on top. The general guidelines come out of basic human nature, which tends to be the same the world over.
Japanese sites do look stylistically different than those in the U.S.
I agree. But it’s more of an aesthetic difference, not so much a matter of navigation. There’s a tendency to have slightly more cluttered pages in much of Asia. That may partly be a matter of subjective preference, but it also tends to be what you see in the earlier stages of building websites, before testing. And while there hasn’t been much testing done in U.S., there’s even less in Asia?though more in Japan. I think it’s an emerging discipline over there.
People also think there is a special Japanese or Chinese way of doing things. But that’s not necessarily true. There was a university study done in Hong Kong that tested Chinese e-commerce sites with Chinese users, and then compared those scores with our own guidelines. It turned out that there was a very strong correlation between the two: high scoring sites also scored well on our guidelines. The actual content may differ by country, but the principles of design are universal.
Many of the newer websites use AJAX. Financial pages, for example, now have more of a drag-and-drop interactive feel. Do you see that as a plus, a minus, or a temptation?
It depends. We have done studies with the investor relations people for their section of the corporate website. It turns out that the average user doesn’t actually want a lot of advanced tools for looking at the financial numbers in five different ways. Sure, a few investment professionals may want them, but they can use a site like Bloomberg.com for that.
That’s the general danger of AJAX?that is that people will add too many features that too few people will use. As for drag and drop?it works very well if you’re using an application on a regular basis. But on the Web, use is usually more intermittent.
You’ve also criticized the overuse of Flash.
At worst, Flash has encouraged those completely bogus ⁠welcome to my website⁠⁠ screens?forcing everyone to look for the ⁠skip intro⁠⁠ button. And Flash has too often been used as if it were a television, a passive media, whereas you want to make a web page as interactive as possible. But we’ve found some nice uses for Flash, as well. For example, Sony’s website has image of MP3 players that you can spin 360 degrees. The web design is very simple, but gives you a good idea of what the actual product looks like.