Jakob Nielsen is best known as the guru of Webusability. Through his consultancy, the Nielsen Norman Group, he has lectured, hectored and vectored firms towards the goal of Web sites that actually work for the user.
But Nielsen's views are controversial. To some, his insistence on simplicity, straight lines and limited colours make him the Internet's Mondrian - a minimalist genius and scourge of design overload. But to many in the Web design fraternity he is just a killjoy. What good is the Web, they ask, if you can't push the boundaries set by standards bodies that necessarily work to the lowest common denominator.
Nielsen flies in to London this month on a "world tour" of lectures and seminars dedicated to spreading the word on usability (see box p58). But just what is usability, and is it getting better or worse as the Web becomes commercialised?
"Commercialisation made things worse, but e-commerce is actually making things better," says Nielsen. "Many big companies produce very bad Web sites. But many commerce Web sites - sites that actually sell something - are reasonably good. Not all are up to my complete standards, but pretty good.
"The difference between a big company Web site and a commerce site that actually sells things is simple: on the e-commerce Web site, they are accountable for the quality of the design. If it is easy to use, it will sell; if it is difficult to use, people will go away. So they have this very easy metric that they can see every day.
"Contrast that to a big corporate Web site. In a big company there is no accountability for the actual user experience because the way they rate employees or give them promotions is not based on whether the users can use the Web site.
The way they judge things in a big company is that you give a presentation to the executive committee and, if they like it, you are doing well. That's why there are so many big company Web sites that actually look very good but are impossible to use. The managers never had to use them."
Has Nielsen noticed any difference as Web front-ends become integrated with back-end systems? Does this inevitably lead to functionality being defined in IT terms rather than end-user terms?
"I don't think it's getting better as fast as I would like, but it is definitely getting better. There are fewer of the really 'loaded' Web sites, with huge splash pages, those Flash screens, big photos everywhere. There is more focus on the things customers want.
"The new emphasis on back-end connectivity may help here because back-end engineers are not so obsessed with graphics. They are more obsessed with making a solution and I think that is a better stage to move from when you're trying to make something good.
"It just changes the question in terms of usability. In a traditional Web site the question was how to provide things people want to know. In the new scenario it's more: 'Let's make the features work the way people think, make them more natural, make it an easy task flow.' Those are the classic problems that we've been dealing with for 20 years. They are not going to go away."
What are the horror stories on today's Web sites? Nielsen has a famous vendetta against Flash, the graphics plug-in that allows Web sites to run animated films.
"Flash is just as bad as the old 'splash page'. Before you get in to use something - which was why you came to the Web site - you have to sit and suffer through something. At least most of the Flash intros have a little button that says 'skip the intro'.
"It is annoying that, after all the investment sunk into the Internet in the past five years, we've got a new technology and a new way of annoying users."
Apart from Flash his biggest gripe at present is site designers' inability to provide logical search facilities. "A search engine is a completely different thing than all other issues in Web design - an add-on. It typically gets neglected in every process. And yet if you look at what the user does when they arrive at the Web site, they very often go directly to the search engine."
Nielsen believes there are two types of users: those who know vaguely what they are looking for and those who know exactly what they want to find.
"If you try that on almost all Web sites today, you will fail miserably because the search engines are of random quality: they just do full word, keyword matching."
Does Nielsen's hostility to Flash extend to all the proprietary extensions to HTML that many Web purists fear will destroy it as an open standard - for example .php, .asp?
"You have to distinguish between two different types of those proprietary languages. There are ones that only live on the back-end server, then they spit out standard HTML, like .php and .asp. With this we can say Microsoft did something right. With these products the only place where you have an incompatible technology is on the server and it's purely up to a user company to make sure that their own server works and that they hire staff who know the programming language.
"The problem is when the stack of things transmitted across the Internet arrive at your machine in a non-standard format. Flash is right now the worst offender for that. I'm not saying never use it but I'm just saying only use it when it really adds value. When you use non-standard technologies, there's a lot of problems you run into. Macromedia would say: 'But 90% of all Web users do have our programme installed.' To which there are two answers: first, which version? Second, it means you can lose 10% of your business just by that decision."
Nielsen thinks the Web's "killer app" is its ability to provide instant gratification. "I get what I want, that's what the Internet gives to me and it really is a very satisfying feeling that I personally control my destiny. I am the emperor, the ruler of the universe - a very small universe, admittedly, constructed on my screen. I rule that little domain. And that's the feeling that the Internet really emphasises. Anybody who messes with that, it's like a rebellion against the emperor And that's what the people do when they kind of freeze your screen with a Flash that takes forever to download. Or they do other things that interfere with my ability to do what I want."
What does Nielsen say to colleges that are training people in Flash and various other multimedia plug-ins to the Web? "My main worry is that they're telling people 'this is what you can do tomorrow'. It should be: 'these are things you should consider for five years from now'. I would say that it is more important to teach interaction theory - which is to say the principles of human behaviour; what is easy for people to use and what is difficult for people to use, because those are constant principles."
Nielsen refutes the charge that his approach to Internet usability has not changed much since HTML's early iterations. "That's because I'm really focused on the human side of technology and that's based on the human brain and other aspects of the human perception: it's basically the brain and that's going to stay the same.
"In contrast there's technology which is changing all the time. So when you talk about the educational system, I would rather have them not focus on the technology. Whatever people are going to do to build good Web sites in five years it's almost certainly not going to be Flash - but what is going to remain constant is how many things the human brain can keep and sort in its memory at any one time."
For Nielsen usability and the defence of global Internet standards go hand in hand. "The first reason is that one of the first criteria for usability is that it works. And that's where the standards are really important: because it's just impossible to develop or test their designs on all possible platforms."
"It is very dangerous to predict into the future from the current scenario. Right now we have a temporary situation - call it an aberration - where Microsoft Internet Explorer rules the Web and there's a few people left that use Netscape, but not very many. Internet Explorer has 85% of the market and Windows has an even bigger market here. So I can understand people saying that's what we're aiming at. But, actually this is exactly the same scenario we had five years ago, it was just Netscape instead. You also had one dominant browser that everyone used except for a few percent, who used something else. But I think those two scenarios were temporary. For example we had a big transition period from Netscape to Internet Explorer - a four-year period where the two were almost equally big.
"In the future devices and browser programs will be very different. So for your Web site to work across all those different environments, it's got to follow the standards. The second reason that standards are important for usability comes back to human behaviour. Even if we are left with Microsoft ruling the world, we are still left with the fact that eyeballs, people, are human. When someone moves around from one Web site to another, if they know what to expect it is much easier for them to use it.
"Take a simple example: the 'back' button on a browser. Does the back button work? Well if people follow the standard, then users can navigate and they can move forward by clicking an underlined link, or they can move backwards by clicking the back button. And if they come across something they like, they can copy the URL and stick it in an e-mail and send it to one of their friends, saying: 'You should look at this Web page.'
"Every time we do studies of users we see the back button is, like, magnetic. It's like the back button is the lifeline of the user. Having the authority for me to say: 'This is the wrong path, I want to go back.' That's one of the things that makes the very strong feeling of empowerment and satisfaction on the side of the user.
"Following the standards encourages that feeling of empowerment of people knowing what to do. And the more they know what to do the more they will dig into it and enjoy it."
Nielsen moves on to cascading style sheets - another add-on to basic HTML that, in this case, has been accepted into the World-Wide Web Consortium standard for Internet authoring.
"The only issue of concern that I have about style sheets is that if they have been very poorly implemented in the browsers, particularly the Version 4 browsers. Internet Explorer 5 actually did a decent job, and there again, one can always criticise Microsoft for making a few mistakes but they actually did a fairly good job, so the critique is that they didn't do a perfect job. I think we've just been through a temporary period where we have had a good idea and a good specification but a bad implementation. Style sheets have not taken off as well as they should have but I must say I see them used more and more when I go to well designed Web sites.
"Style sheets give designers more freedom, which is good, but it also gives designers more opportunities to do things that actually hurt you. For example, you can specify the text size, the font size completely accurately which graphic designers love - but here's the danger - how do you know that you're not showing your material to a user who needs different text because they have poor eyesight. And so if you freeze the font size and say 'this is going to be 11 point for ever after', that is going to be very unpleasant to a user who needs 14 point to read."
Is it still possible to design a good looking Web front-end using plain vanilla HTML?
"I think yes," says Nielsen. "You can spice it up a little bit with some use of style sheets. But one of the interesting points if you do this is what happens if the browser does not support them. If it's done correctly it will have graceful degradation. Users will not see that little extra effect that you designed into to make it look good - but it will still look almost as good. That is a really important point. Non-standard features need have graceful degradation. As opposed to, I guess, miserable degradation."
What does Nielsen think the Internet will look like in five years' time?
"We've added a small amount of multimedia. That's what the past five years have given us. If you look at the next five years, many of the changes will be of the same modest kind.
"Because there's now a really big installed base it becomes more difficult to make radical change. Bandwidth is not going up very fast. It is going up fast in the dreams of telephone companies, and in the demos, and in the trade shows. But if you look at the actual connection of the house of an actual person, the vast majority of people in the UK still don't have broadband.
"In five years' time the majority of people would have reasonably good connections, but there will still be a minority who have just the same bad connections we have today.
"But if we look 10 years into the future, I think we start assuming that everybody is well connected. Other technologies we can look at are what kind of computers people will have.
"Here's a nice prediction, and this is one that I am 100% sure on: what will be a low end computer in five years? It will be the computer you just bought yesterday. Because that same physical computer will most likely be in the world in five years. You won't be using it, hopefully, but somebody will.
"Now if you look 10 years into the future, what will the low end computer be? That will be something very fancy. And what will the high end computer be in 10 years? That will be something fabulous. But in five years, we're still going to have the clunky systems. We're still going to have people using Windows 2000, just as today we still have people using Win95. It's just going to be misery."
Nielsen sees the fastest evolutionary change in the next five years on the mobile Internet.
"Wap phones might as well not exist because Wap is just so unpleasant that it's not going to have any impact on the mobile Internet in the long term. Maybe in the year 2000, 2001 - but in five years from now people aren't even going to remember Wap. It is going to be sort of like discussing some esoteric feature of Netscape 1.2. 'Well, yeah, I remember it, was there'.
"Devices like wireless personal digital assistants will make a big revolution in how people use the Internet. They will make it even more personal, will make it even more the instant gratification medium - where I get what I want and now you can add where and when I want it."
Nielsen believes the next big revolution in Internet usability will be to abandon the "page metaphor".
"We will abandon the Web browser and really have something that I call the Internet desktop which is your information control panel, a new type of software that will integrate your view of all information resources, whether they are e-mail, Web pages, stock trading systems, everything like that."
No more hypertext?
There will be no connection between this interface and the hypertext link, Nielsen believes. Coming from the guru of old school Internet design that sounds like heresy.
"I'm a very big fan of hypertext and I've worked on it since 1983 - but hypertext is often quite jarring and should really serve a smaller percentage of your interaction than it is now.
"Hypertext is like a time warp: it's like being beamed up in Star Trek. You click on something and - boom! - your previous existence vanishes in a puff of smoke and you materialise in some random new place in the universe without any understanding of what's going on."
"It has certainly been successful, but it wasn't the optimal way of managing all your information needs. You would never make a spreadsheet that, every time you changed a plus or minus, the entire whole display goes away and another one appears.
"The question is, can we get rid of the browser in the next five years? In the old days I would have said: 'Yes, because the new ideas are so much better than the old ones.' But I am getting more conservative. The problem is the big installed base."
Small businesses are today confronted with a "shrink-wrapped" Web presence packages - often based on Front Page or some other standard interface design - and they don't have time or money to think about usability as a discipline. What would Nielsen's advice be to a small firm starting out on the Web?
"I would say keep it simple. The fewer resources you have the more you can waste them by trying to do something overly fancy that's not actually going to work. I think it's very important to focus on the content, on the information that customers actually want as opposed to on the glitz which costs a lot of money to do and is not going to be a differentiator."
"An SME by definition is a targeted company that does one thing, hopefully, well. So if they can describe that, they can actually stand out. Big companies always provide grand descriptions of what they are doing because no one actually knows what they are doing. A small company can be very specific: this is what we do. Provide a clear and simple explanation of what you do and write the content in a scaled-back manner; not too filled with slogans and bragging but tell you exactly what you do. This is easy advice to give, but it's quite hard to do.
"For example on the Nielsen Norman Group home page we once said: 'The philosophy of Nielsen Norman Group is simple.' There followed eight paragraphs of really heavy-duty text. And when we did our testing, people said: 'If your philosophy is so simple, why did it take you eight paragraphs to explain it.'
"That was a classic mistake because we thought that one paragraph was all really important information, but when users come there they just want one or two simple facts to start with.
"If you sell products, use clean simple product photographs. People actually do want to see what they're getting, so that's a type of graphics that really help.
"Also add prices. For some reason a lot of people don't want to say what things cost on the Web, but nobody's ever going to buy anything without getting the price, so you might as well give it.
"Often users in tests say: 'I don't understand what this is - is this a £20 product, a £2,000, or a £2m thing?'
"So think about it from the customer's perspective: what pieces of information do people really need to know and explain those in no-nonsense, plain and simple language. That's the one thing a small company can do; it doesn't require anything other than a commitment to writing the simple truths."
Nielsen thinks it is a problem for small firms to put themselves completely into the hands of "one-stop" Web presence firms. "You need to maintain the chart of your own destiny on the Internet, because that is really the future of business. So you have to maintain control over your own Web site.
"Of course it's good to have a good graphic designer help you with the layout - for that is the thing the average person does very poorly. But the content is really what people are there for and that has to be written by someone who knows the business."
Jakob Nielsen: CV
>CEO, nielsen norman group
Jakob Nielsen, PhD in user interface design/computer science, founded the "discount usability engineering" movement for fast and cheap improvements of user interfaces. Nielsen has published nine texts focused on usability, the latest being Designing Web Usability: The Practice of Simplicity (2000), which examines Web site design from the perspective of the users' needs. For more information visit: www.useit.com
His User Experience World Tour is in London 28-30 November. The main event is 30 November, Royal Drury Lane Theatre, London, tel: +44(0)20-7494 5399, www.stoll-moss.com. The tutorials are on 28 and 29 November at the Paragon Hotel, 47 Lillie Road, London , tel: +44(0)020-7385 1255 www.paragonhotel.co.uk.
Nielsen on the bottom-line benefits of Web usability
For Nielsen, usability is not just an added extra but crucial to the bottom-line benefits of trading on the Internet. "When Web sites improve their usability, they really specifically get dramatically increased sales or increased usage, whatever way you want to measure success."
Nielsen claims he can improve "measured usability" on any Web site by "50% -100% if I do really well". By measured usability, he means speed, ability of users to remember what they saw etc. However, he claims that a 100% usability hike can deliver four times that benefit in increased sales.
"If you make it twice as easy to find things, people don't just buy twice as much, but they buy four times as much because now they feel this is a Web sites that they can trust, that is welcoming them, and they'll be able to find it in the future.
"Maybe this is a temporary situation that only works because so many other Web sites are so bad so that people feel really thrilled when they come across a good one - but right now it is really true that the impact of usability on the bottom line of an Internet company is dramatically bigger than anything else. I think companies now are wasting endless resources on advertising. This drives people to the home page - but then, if it's unusable, they go away.
"On lots of Web sites 90% of the people who look at the home page, look at no other page. And what usability does is it takes the other half of the equation which is of the people who look at your home page, how many of them become your customers? This is known as the conversion rate, and on most Web sites the conversion rate is about 1%.
"So you have two problems you could attack. You could either make the number of people who look at your home page bigger or you could make the percentage of people who look at you home page who will then convert into customers bigger.
"Usability attacks the second problem and in a sense that is where the vast issues lie. That's where most companies are losing almost all their business."
Nielsen's top 10 mistakes on the Web
Since 1996 Jakob Nielsen has kept a list of the top 10 mistakes in Web design. He says this is mostly still valid, but in 1999 he added a further 10 mistakes resulting from new technology and applications. For the full explanation visit www.useit.com/ alertbox/990530.html.
This was first published in November 2000