Online Extra: Jakob Nielsen on the Unwieldy Web

The Web usability whiz says some sites are getting better, but far too many remain too complex and frustrating for typical users

Jakob Nielsen has long been the acknowledged leader in making Web sites more usable. As a principal with the business design consultancy Nielsen Norman Group, Nielsen is hired by dozens of companies that want to improve their Web sites. He walks the walk: His Web site,, itself is a paragon of simplicity and speed, with absolutely no graphics on the home page because they slow the experience for millions of people still using dial-up connections.

Nielsen doesn't pull punches. He thinks Web sites generally are still poorly designed and badly written, and that Web logs, the current darling of the Web, are useful in spurring dialogue, but are a step back in usability for the average person. In a recent conversation with Silicon Valley Bureau Chief Robert D. Hof, Nielsen discussed the state of the Web and where it's headed. His conclusion: The Internet still has enormous shortcomings -- especially internal corporate Web sites and e-mail, the hidden majority of Net activity -- that need to be overcome if it's to reach its full potential. Following are edited excerpts of their conversation:

How is the Web of today different from 5 or 10 years ago?

The main utility of the Web is still very information-oriented. People's blogs are fun, but not of any particular value typically. The main problems that we see on the Internet are not being addressed.

Like what?

Bad content and lack of information people need -- either because it's not provided at all or because it's written in a poor, impenetrable style. Companies must stop doing brochureware and start answering customers' questions in plain language. You often can't even tell from the home page what the company does.

Search is the second problem. Even though there has been progress on Webwide search -- Yahoo! (YHOO ), Google (GOOG ), and Microsoft (MSFT ) have done really nice search engines -- when you go to a corporate Web site or an intranet, the search is typically extremely poor. That to me is a particular scandal, because people come to the site on purpose, to find something.

The third problem is e-mail overload. Those three big problems are holding us back much more than any of the new services are pushing us forward.

Tough problems. What would be some solutions?

On the first two problems, you have to recognize that the Web is a communications medium, a two-way medium. It's not a broadcast medium like TV where you want to be glamorous. People go to Web sites because they have questions they want answered. That's just not being done. We don't really need any new technology. We just need to use the technology we have appropriately.

What about e-mail?

E-mail overload, on the other hand, does require new research. There has been some progress in spam filtering lately, such as the collaborative spam filters of Cloudmark and those types of services. They cut down on spam dramatically.

But phishing and other kinds of untrustworthy computing are making computers and the Internet a very scary environment. That is another really big issue.

What do you think of the rise of Web logs?

They support one of the basic concepts of the Web, which is diversity in services and the ability of everybody to participate on equal footing. That has always been one of the benefits of the Web. I've always said that small companies should be the biggest manufacturers of the Web, because you don't have to have a big marble building or send salespeople out to play golf with the clients to have a good Web site.

Web logs are a media form that is optimized for the Web because they live by links. They're able to link to the original material and then provide their spin or twist or commentary. It's much easier to comment on the story than it is to research and report the story in the first place.

When millions of people do that, sooner or later you're going to come up with some interesting comments. Then other people link to the interesting comments more than they link to the uninteresting comments. You run that little loop a few times, and the good stuff will bubble to the top.

Blog, or RSS, feeds, however, still seem quite primitive to me.

The technology is definitely at an extremely geeky level at this time. It's only really for the bleeding-edge early adopters. Some people live by Web logs, but that's not the average user. The average user is too busy. They just want to get work done.

Those little RSS icons? They're meaningless to most people. One of the big lessons from user testing is that life is too short to click on buttons you don't understand. News readers also will have to be dramatically simplified in how you subscribe to feeds and use them.

Apart from the technology, what do you think of the content of Web logs?

Most writers of Web logs simply cannot write something that makes sense out of context. If you read these headlines and you scan down, what is relevant to you, and what you need to click on, is rarely communicated very well.

Is there a prospect for that changing anytime soon?

We're talking about millions of people without any it's hard to imagine getting the content up to a higher quality standard. Even commercial Web sites with professional writers and professional editors...still do not have content that's sufficiently attractive and informative to users. So I'm rather pessimistic on the hope that bloggers will really learn how to do this well. I think they're probably always doomed to being a little bit of a secondary thing that people turn to when they want to check on gossip or whatever, but not really for hard-core purchase decisions.

What new kinds of work on Web design are you doing these days?

There's a broader set of users than there were before. So I've been looking at how teenagers use the Web, or how seniors use the Web, or low-literacy users. Those are large groups of people who have grave difficulty using the Web. Old users, for instance, constitute a very rich market, and they're coming online in big numbers. Yet their needs are ignored. It's very sad.

When it comes to conventional business Web sites, what problems are you working on?

The people who are in charge of these Web sites don't know how difficult they are to use by outside users. People fool themselves into thinking they have the basics right when in fact they have them wrong. They don't know how new customers behave because most companies don't run a study with real customers. This has been a problem for the last 10 years.

Gee, hasn't anything improved?

Every year, it does in fact get better. More companies do user testing. But it takes a lot of time to make changes. It's still too difficult to make these changes happen.

That's where the technology can help. There's less and less required to be hand-built. More and more capabilities are available that you can just sort of plug in. Take the Yahoo! Store, for small companies to sell online. You outsource to Yahoo your shopping cart. It's not a perfect solution, but it's a nice solution.

Who else are Web sites doing right?

Google has done many, many things right. It has such a sparse home page that there's no doubt what you do. It's also very fast. That's something that every study of Web usability has shown people get excited about. This is true for all users and across all applications.

A wealth of tools are available to customize your Web. How much can these change the Web experience, and is that a good idea?

In some ways, it's a good idea, because it allows people to tailor the experience to their own personal needs. Mainly, though, it's just not something you should expect the average user to do. It's just too complicated.

And users don't care about technology. Users don't care about Web sites. They just care about getting their work done and getting the computer out of the way.

What new technology or techniques are needed to move the Web to the next level?

We need to have much more powerful navigation tools in the Web browser. We need more visual navigation on sites themselves. But most of the efforts have been stand-alone attempts to improve particular features. We need a bigger, more integrated solution -- which maybe Microsoft or somebody else can do.

We also need much better collaboration features. The Web is still very much a single-user environment. Wikis -- multi-user Web sites that can be built with many people -- are a great example of what is working. At Wikipedia [the collaborative online encyclopedia], the individual components may not be so great, but they can add up to something valuable. I think there will be many more specialized projects of a similar nature.

Before it's here, it's on the Bloomberg Terminal.