by: Christian Sarkar

Jakob Nielsen, author of Designing Web Usability and Homepage Usability: 50 Websites Deconstructed, advises companies on how to make their websites, public or internal, more user-friendly.

NOTE: This interview was conducted way back- April, 1999 - to be precise. Since then, I've interviewed Jakob twice more- see Creating The Loyal Customer (AgileBrain.com) and Use it or Lose it: Improving the Online Experience (qualityofexperience.org). He is always very insightful and this interview (below) remains relevant today.

If there's one thing Jakob Nielsen knows plenty about, it's website usability. He has, after all, made a career out of advising companies on how to make their websites more efficient, more effective and, above all, more user-friendly. According to Nielsen, that last point is the key to having a website that works well – "keep it simple" is his website mantra.

The high-tech engineer-turned-consultant spoke with me about the different levels of website design and how a company can determine whether or not its website will pass a usability test.

How do you define Web usability, and what are your criteria for measuring it?

Usability is always defined by two big questions: Who are the users, and what are they trying to achieve? Relative to those two things, is the website easy or difficult? On an intranet, the users are your employees. For a public website, the users are your customers. Those two different groups of people differ in how much they know about the company. So the structure of the website will differ because of that.

That's one of the huge mistakes that almost all companies make the first time they go on the Web. They don't actually think about the customer's view of the information space or what the customer is trying to achieve. They're looking at it more from the view of, 'How do we understand the problem inside the company?' So the next question is: What are the users trying to achieve?

For an internal website, it's different types of job-oriented or employee tasks. For a public website, it depends on the goal of the site. If it's a shopping site, site products could be very important. If it's an entertainment site, having fun is the real goal. There's no single, optimal design for a website. It really depends on what the users are trying to accomplish.

One of things you make a big point about is that people read differently on the Web than in print. All too many websites are designed by print people with advertising backgrounds.

That often leads to Web pages that may look good but aren't good to interact with. The user interface is the 'look and feel' of a website. Of those two terms, the feel is the most important. How it feels to use something is more important than how it looks. People aren't sitting there just enjoying the website, they're trying to do something with it. My personal website is based on the notion that getting information to people quickly is really important.

The difference in behavior is that online, there's much more moving around. The amount of time people spend on any individual page is much shorter because there are so many things they can do online. Also, when people try to find something online, there are often many different places they can go to read about a certain issue, problem or product. It's almost in self-defense that people want short, punchy answers on the Web.

Is that why you recommend a very objective style of writing?

Yes. Every user we've interviewed has said, 'Just give me the facts.' That's what they want on the Web. The Web is very user-driven. You're sitting there, with a hand on the mouse, ready to click on the places you want to go. This is very different from the traditional marketing problem in which the user is getting catalogs, and your goal is to make sure that they pay attention to your catalog, your advertisement or magazine. With the Web, they're already at your website [and] they've already indicated that they have an interest in you. Now your job is to prove to them that you can deliver something.

How do you go about doing a usability test?

The procedure is very simple – it's one user at a time. That's different from a focus group, which gives you an impression of what people like. If you want to see whether or not they're capable of doing it, which is the usability aspect, you have to have one user at a time sit at the computer and use a website. Typically, I'll take them to a home page and then give them some problem that they have to accomplish. Maybe they have to buy a new monitor for their computer. Then you'll see how fast they can find it, how they decide which monitor to buy and whether the website is providing them with the information they need to make that decision.

You will often find that they click on all the wrong buttons. They don't go directly where they prefer to go because users often have a different understanding about the terms that you put up on the home page. They may not click directly on the button they were supposed to. So the next thing you look at is, 'If they click on the wrong button, how quickly will they discover it was the wrong one?'

Once you know what the problems are, you can redesign with the goal of fixing those difficulties and enhancing those parts of the site that customers liked. The next site won't be perfect, either – it's a continuous process of improvement. I don't think anybody today has what you would call the perfect website [but] people who have the correct process can improve their site every time

Many commerce sites are struggling to sell products. What is the primary thing they need to do to sell more?

Make it easy to buy. The Web really is a buying medium. People will go to a website because they're interested in buying the product, so you have half of your work done for you already. Now you need to make it very easy to actually buy the product. Specifically, how many clicks do you have from the home page before you've finished buying? That's a very simple thing you can count, and it should be as small a number as possible.

Something that's a little more complicated is this: At any given step of the way, how easy is it to make the correct decision as opposed to making the wrong decision? You want users to understand what products and services you're offering. You can't just say, 'This is the product and with so many clicks, you can buy it.' You've got to give people a lot of information about it because they can't touch and feel it.

You can be brief and give people a lot of information. How? Through good use of hypertext and good information architecture. You don't put supplementary information on the front page – you put it on a secondary page with very clear links to it. Some people may care a lot about performance; others may care more about reliability or how well something interoperates with something they already have. You have to provide very detailed information and yet not clutter it [the website] up.

So the trick is to give users choices rather than force them down a certain route?

Exactly. People like the notion that they're in control and can click where they want to go. It just has to be very easy for them to do that. I'm also in favor of having very fast websites with a very fast response time. That means having minimal graphics. At the same time, people also want to be able to see the products before they buy them. So you also need to have photographs of the products. Again, here you have a conflict. You solve it by not putting 12 big pictures all on the first page, but you have a link for people to click and then get a bigger photograph if they want.

What about personalization?

People use personalization as an excuse for not having to fix the basic navigation of their website. If you have an easy-to-navigate website and it's easy with a few clicks to go anywhere you want to go, that takes care of the vast majority of things that you'd otherwise like to use personalization for. So personalization is really the ability of the website to deliver some individualized service for a particular user.

You have said that Web design has three levels: Web management, interaction design and content design. Can you tell us about these and how they go together?

Starting in the middle, the interaction design is the issue of how the user is interacting with the website, how they're moving through the website. That's crucial because any interesting website will have a huge number of pages, products or news articles. Because the Web has a lot of information, good navigation is crucial. Content design is a matter of what's beyond the page, what the user is seeing. That's where you need to be able to give people the information they're actually looking for. It has to be done in a way that matches the characteristics of the Web, not the characteristics of old media. It has to have good hypertext links to background information.

So content design is more of a localized problem, and navigation design is more of a site-wide problem. Those two design problems are a matter of what you're presenting to the user when they're on your website. Above that, you have the Web project management, which is more a matter of how you make sure that the decisions you've made for the other two levels are delivered on the website to the user in a consistent manner. Any given website has so much information on it, that you'll really splinter the website if it's not managed as a whole, in a way that you have a single, unified user experience. A lot of websites have abandoned that part of the project, and they don't have any real coordination. Users really notice that.

A lot of people have websites just because everybody else has a website, not because they know how they're going to make their customers' lives easier or how to benefit from having a website. Unless you have those goals very clearly established, you're going to have a very confused website. The overall experience will be of a site that's not very useful.

Do you believe that every business today is, in the end, a Web business?

Yes, maybe not today, but in a year or two. There are very few businesses that shouldn't have a website. Most companies' websites will become their most important interface with the customer. That's going to increase dramatically because it's still too difficult to do most things on the Web. There are still too many companies that aren't even on the Web yet and too many companies that have a bad website that doesn't serve their customers in a decent way.

But I think that's going to change over the next few years – customers will expect that they can go online and find information about a company. It's similar to having a fax machine a few years ago. If you don't have a fax machine now, you can't be a real company. Similarly, in the future, if you don't have a good website, you can't really offer good customer service.

HOME