Jakob Nielsen is possibly the most well-known personality in the field of usability and user-centered design. In this interview, he revisits history, all the way back to his first experiences with room-sized ‘personal computers’ in the ’70s.
Excerpts From the Interview
Today I am not a programmer – and I don’t think anybody should hire me as a programmer – but I have had experience [as a programmer] and I think that’s valuable. In any project, the discussion inevitably boils down to “can this be done or not,” and engineers have their own attitudes about the answers. Having programmed, even though it was a long time ago, helps me when I talk to engineers about what can and can not be done. It helps me when I ask them about finding ways around technical limitations.
Actually, now I think the field is going to get much bigger, because now I’ve finally come to realize how broad the benefits of usability are. I can see how many projects – even today – don’t do any user testing. Most project teams have no dedicated usability person on board, and they probably don’t even have a dedicated designer or interaction designer on board. And yet, when they do, they double their sales.
An Interview with Jakob Nielsen
Conducted by Tamara Adlin on April 24, 2007 10:25 PM
Today Jakob Nielsen is an inspiration and, through his books and seminars, a teacher to many. But what inspired him to get where he is today?
Until 1998, Jakob was a Sun Microsystems Distinguished Engineer. He also introduced the idea of discount usability engineering.
Jakob has written so many books, I can’t even describe all of them. The first one in our field was published in 1989. Recent books include Prioritizing Web Usability, Homepage Usability, and Designing Web Usability. So, tremendous accomplishments in our field.
Jakob, you are possibly one of the most interviewed people in our field. I want to ask you some slightly what I hope are different questions.
One theme was to make technology easier for humans; to adapt technology for humans as opposed to having humans submit to technology.
His other theme was the universal information system hypertext that we could interlink information; that we could have all this richness of information available. He wrote about this in the 1970s.
Those were very inspiring books to read at the time, when computers were mainly hulking big mainframes that were really very unpleasant to use.
Because it was so old, it had been made available for a few high school students who were into computers, to use more as a hobby. We didn’t really have courses on computers at the high school at the time.
That was really interesting: to use this personal computer and get this feeling of computers being responsive to the human or the person.
Then I started at university. You were supposed to get more advanced stuff when you went from high school to university. In fact, the computer at the university was much bigger but it was a mainframe, time-shared, multi-user system.
The user interface was just horrible. I didn’t really know about usability at the time but you could tell the user experience was much worse than what I had used back at the so-called primitive computer in high school.
That got me interested in the idea that computers could be pleasant or unpleasant. I had had a pleasant one in high school, and now I had an unpleasant one. It was more advanced and powerful – but more unpleasant.
I think people today, who just assume that everyone has their own personal computer, have no concept of how unpleasant the mainframes were in the age of time-shared computers.
That was one of my motivations; having that personal experience with pleasant versus unpleasant user experiences. I knew it could be better.
The other thing that annoyed me at the time, was that everything was based on personal opinion – like this is elegant, this is not elegant, this is how it ought to be, this is how it should not be — as opposed to a more empirical approach that says you can actually find out what works and what doesn’t work. You don’t have to rely on somebody’s opinion; you can actually find out.
I think that I was inspired from an even earlier age when I was 10 years old or so. Both my parents took me into their work from time to time just enough that I kind of saw what was going on with computers.
They were both psychologists at the university. So I saw all these different experiments that were being set up by the psychologists who measure things that are not particularly interesting to outsiders.
But seeing those experiments, even if I didn’t understand the theory, was really exciting for a little kid. Later on when I was a student, having seen that as a little kid, I had a feeling that we can actually study things. We don’t have to argue about them. We can test them.
These different experiences gelled when it came time for me to choose my thesis topic. I decided to go study usability.
But sometimes breaking things down to the basics can be useful. That’s what I did back in 1983. I broke user experience up to different levels of experience.
I worked at the university and then I graduated. I got a job as a university professor to teach the same things to the next generation of students. In academia, you become a student then you become a professor then you teach more students, then they become professors. It’s a circle.
But after a while as university professor, I did think well, it’s just not productive enough. The students loved my course. I always had the highest rankings when the students ranked their professors. We give them grades but they rank us.
The students were always very positive about the course. It was always very over-subscribed, but ultimately it was not completely satisfying to keep on doing that.
The one good thing about being a university professor – this was back when I lived in Denmark – is we did not have a big budget. That’s actually probably true today as well for many universities.
Because we didn’t have a big budget, I could not do the kinds of projects that were done at the big research labs like IBM and MIT Media Lab.
I had to do simpler projects. This led me to develop a discount usability methodology. I had to figure out what we could do with a smaller budget.
This turned out to be extremely useful in practice, because in development projects, you never have enough time or money. You can’t do these fabulous academic, thorough studies. You have to be pragmatic “what can we learn this week so we can make it better next week?”
That’s actually what I worked on a lot in the 80’s. I wanted to develop cheaper and simpler methods.
I think there is actually some value to having had that experience. Even today I am not a programmer — and I don’t think anybody should hire me as a programmer — but I have had this experience and I think that’s valuable. In any project, the discussion inevitably boils down to “can this be done or not,” and engineers have their own attitudes about the answers. Having programmed, even though it was a long time ago, helps me when I talk to engineers about what can and can not be done. It helps me when I ask them about finding ways around technical limitations.
Today there are still technical limitations. Years from now there will be other technical limitations. We will never be able to do everything that we can dream up.
We want user experience to be perfect, but you can never make it perfect. We can make it better, and being able to talk to the engineers and understand what they have to go through, I think, is valuable.
I think the course was called something like Human Computer Interaction – a very generic broad term.
Usability is really an intersection between psychology and engineering, and even art and visual design. There are a lot of different components and, in traditional university departments, each were studied in ways that didn’t really help our field.
One way of looking at Human Computer Interaction is as a multidisciplinary field; another way is to say this is a new field that draws upon older fields.
I worked with several really brilliant people there. It was just for a year, but I discovered that I liked working in industrial lab as opposed to a university lab.
There were so many people, it’s unfair to even mention any individual ones but John Gould was one of my big idols at the time. I learned a lot from him. He is actually retired now but he was a brilliant person in terms of designing experiments to study new things. For example, he studied how to do three-dimensional user interfaces at a time when we couldn’t really build them.
Because we couldn’t build 3D interfaces, he had people use Mr. Potato Head and other types of physical objects to study the actions and operations people would do when they were given tasks to make certain changes on these toys.
He did these really early studies to come up with very interesting early insights that then later on – maybe 10 years later – might fit into IBM products when the technology caught up with him.
He was really a true pioneer there, and there were many other brilliant people, like Jack Carroll.
That was just for a year. I went back to university but said, “okay, I don’t like being locked up in this ivory tower.” Then I moved to Bellcore (Bell Communications Research), which was the telephone company or, actually, the regional telephone companies’ lab in New Jersey.
It was a spin off of Bell Labs. At the time of the telephone company divestiture, the regional companies and AT&T each got their own research companies. The regional labs had particular emphasis on human factors, and their labs had absorbed most of the old human factors department at AT&T.
Now we’ll go back to the ’50s, when they did things like designing the touchtone keypad and doing experiments to decide how you should lay out the number keys for having the best possible telephone keypad.
Some of the people coming from Bell Labs were real old timers and had done very pioneering early work. That was a really core competency level of expertise in the new group at Bellcore. I was very privileged to join that group; they did brilliant work in different aspects of user interface.
This is also where my interest in hypertext came back a bit. That had been on the back burner during those academic years, because we couldn’t really build hypertext systems. Hypertext was very hypothetical; it was an idea that we could connect all the worlds of information together and have online information sources.
The biggest computer we had was a Mac Plus, which had one megabyte of memory. It did run HyperCard, and we could actually do a little hypertext at the time, but not very much.
On the other hand, at Bellcore, Dennis Egan and another group had built a product called Superbook, which was intended to put technical documentation online.
The intention was to put technical documentation for phone companies online. The documentation for one of the big central offices that held all the big switches took up an entire wall full of binders.
So making that documentation actually searchable and linked, so people could pull up any piece of information they wanted on the screen within a few seconds, was a huge benefit.
Again, this was well before the web, of course. This is a time where we had pre-web hypertexts. It was exciting to get back into that work.
The phone companies have actually always been leading edge. Content usability – things like information foraging, which is now one of the big things – was already being worked on in the late ’80s in the Bell system. I joined them in 1990.
The web in ’91, if you remember that, it was a line-mode interface and you could jump between three different servers or something like that.
It was more like an interesting demo. It wasn’t something that I took particularly seriously, I have to admit.
It was not until two years later that Mosaic and its graphical interface came out. There were several hundred websites by 1993. Back in the early days, we used the web a little bit because our job was to be aware of what was going on, so we could tell the phone companies about it. But we sure didn’t see too much information about the web in ’91.
They were bad because they had no interest in moving on to do new things. For example, some of the other people in the group invented the way of doing simplified email that had been tried out in Florida with a group of senior citizens—who are a very hard audience to get to use new technology—and they just loved it.
What did the phone companies say? They said we don’t want to do email; go back and make the yellow pages easier to use. That was a disappointment for these other guys and all my colleagues.
That sort of general atmosphere permeated everything at the time. The phone companies were not really interested in moving beyond voice and making phone calls. The biggest challenge was how to make call-forwarding easier, which you just can’t do on a touchtone keypad; you need to have a screen for it.
Some of the newer phones are only coming out now with touch screens, which allow you to do things much more easily. But the phone companies were just too conservative to pick up on many of these ideas at the time.
To really go to the next level, I thought, well, I’ve got to go to a place where they have a more aggressive approach to doing new things.
I was quite lucky at the time that I had a little bit of restlessness in me. At the same time, Sun Microsystems had decided that they were going to emphasize usability more, which they had not done in the past.
The original Sun Microsystems was a completely geeky place, but they had finally come to the realization that, to really expand their market, their systems and the software had to be easier to use.
They decided to create a bigger human factors department. They had hired half of Apple’s people away, and they were also looking outside – not just to Apple but in general to see where could they get some experts in the field.
A very attractive thing about Sun is that they had a job category called “Distinguished Engineer,” which is what they were recruiting me to be. The job description is you are the #1 world expert; you figure out what’s important to do.
So when you get a job where that’s the job description, that you decide for yourself what you want to do, that’s attractive of course.
They work on their own project in some department, because you always have to sit in a particular office, but they furthermore have a cross-company mandate. So you can go and talk to any VP you want.
That’s another thing that was attractive about the job. You got to be an expert, and you’re an expert who is actually supposed to make something happen and supposed to make this company move.
It was an attractive job and I was restless, so it was time to move on. That is when I moved into the computer business.
You had also edited Designing User Interfaces for International Use and you had written a textbook, which is no longer in print, on hypertext and hypermedia.
And then in ’94 you published Usability Inspection Methods, Usability Engineering and then Advances in Human Computer Interaction. So you were already thinking about all of these issues by the time you joined Sun.
Even though I said Sun was boosting their human factors department, in truth, of course it was still the smallest department compared to the size of this humongous company. So being efficient was really important to them.
My main interest was on how to have impact on big projects, given the limited budget and people. It was a big issue.
When I joined Sun, the goal of the department was to make Unix easier to use. We were going to get the different development teams to do more usability projects.
Unix is still not really easy to use, because it is inherently very, very complicated. But, this is where the job description of Distinguished Engineer turned out to be a great advantage, because the job description is: you figure out what’s important to work on.
I joined Sun Microsystems in 1994, and that is exactly when the web really boomed. 1993 was when it started to become useful, and ’94 was when, because it was more useful, the web grew by several hundred thousand percent in one year.
Just a few months after joining the company, I decided that, of all the main usability questions one could work on, the most important one was to make the web easy to use. That’s actually what I focused on at the time I was at the Sun: the web and also the intranet, which of course is a big business thing. The web and intranet were important for selling servers, increasing employee productivity, and many, many other reasons.
So I really worked on web sites and the intranet quite intensely for many years at Sun, and did very interesting projects there.
Then, if we fast forward to 1998: as a result of the web growing, and the launch of e-commerce sites, it became clear that usability was no longer a question purely for the computer industry.
The computer and telecommunications industries were where usability started, because they were the first places to do large-scale user interface work. But now making things easier to use had become a broader problem than just a problem for the computer industry.
The web meant that the user interface is your customer interface for basically any company, because your website is your face to the customer. That’s where people turn.
When people go to a trade show, they may pick up some brochures, but they throw them out before they even get on the plane home. Then when they want to look up your company, they look it up on the web.
In ’94, that started to become more prevalent. I was giving a lot of speeches at different conferences and trade shows, and I started to see the people who were coming to talk to me were from a broad range of companies.
Interestingly enough, Don Norman had come from a very different perspective, but he came to similar conclusions: that the usability was becoming a broader problem.
He had worked at Apple; he had been head of Apple’s Advanced Technology Group for a long time. Then he also worked at Hewlett Packard for a while trying to make their products better. He was feeling the same as I was. He had the same general insight that we needed to go beyond just Silicon Valley and beyond just working in that technology business.
I had actually decided that I was going to write a book about web usability. That was my main goal for the year of ’98. Don read about this in the New York Times; that I had started focusing on my book. He called me up and said, “that’s very nice – but shouldn’t we do a company instead to really get some impact?”
I met with him—I’ve known Don forever. Actually the first time I saw Don was back in 1983 when I saw him give a lecture at a conference. He didn’t see me at the time. I heard him speak and he was very inspiring. Then after that, we met many times at different events because we were in those same, relatively small circles of people who specialized in usability. Particularly back in the ’80s, there were not too many people in that area.
I met with Don. I had known him and respected him for a very long time, and we decided, yes, we should start a company.
So we started a company and we called it Nielsen Norman Group, which has been doing very well every since.
The only problem was that my book, which should have been out in ’98, didn’t come out until ’99. The second we started the company we had clients. Actually we had clients before we even started, to be honest.
People were calling us up hoping we would be available, and when we said we were, of course we got even more clients. So we were very, very busy on client projects and that obviously delayed the book. That’s the only downside.
The truth be told, that was probably for the best, because in ’98, even though we said it’s obvious that usability rules the web and you need to do this stuff and pay attention to users, this was not the prevailing attitude.
The best selling book at the time was called Creating Killer Websites, and the prevailing attitude among the dot com boom people was cool sites and Flash pages and bouncing heads and all that stuff were important.
Actually delaying it until ’99 – it was really a 2000 book, because it came out in December of ’99 — delaying it until 2000 meant that enough people had felt the pain and had projects fail because nobody could use them. So there was more uptake and interest in this book. I think, in terms of book sales, it sold much more because it had been delayed.
Consultants I’ve talked to say that about 50% of consulting is actually being a psychologist to the organization they are consulting for.
Was that a surprising find to you?
I think what was very surprising to me was—this shows how completely naïve I was—was from a business perspective. I thought that when people had agreed to pay a certain amount of money for a project and you did the project and you send the invoice and the check would arrive the next day. In fact, it turns out it takes several months to acquire your money, and the bigger the company, the harder it is to get the money out of them. So the cash flow of the Nielsen Norman Group was very tenuous the first year.
We had plenty of clients. We did plenty of projects. We had invoices out for huge amounts of money, we hadn’t gotten paid that much. It was another six months before we got our money.
That of course in retrospect was probably a little bit stupid.
In the early days, we were just a few people and we just thought we were all an obscure specialized type of people. We all knew each other and we were working to make things better for each other. Now, hundreds of thousands of people have started using all the lessons we learned and all the things we developed.
You can look at the book sales—I’ve been publishing books for a very long time, since the ’80s—and the early books would sell one or two thousand copies and the newer books it was a quarter million or something like that. It’s a very simple measurement unit showing how much broader the interest is now and how many more people are working on these issues now.
I don’t think that at the time I necessarily thought it was going to be as big as it has become. Actually, now I think the field is going to get much bigger, because now I’ve finally come to realize how broad the benefits of usability are. I can see how many projects – even today – don’t do any user testing. Most project teams have no dedicated usability person on board, and they probably don’t even have a dedicated designer or interaction designer on board. And yet, when they do, they double their sales.
I think it’s going to get at least 10x bigger than it is now. The question is how fast.
Not every project needs a world expert and not every company should pay very, very high consulting fees for extremely insightful analysis.
Many of today’s problems are just simple little things that have been well documented for at least 10 years. Even for websites, many of the design issues have been documented for 10 years. You just have to look them up. You just need to actually bother looking them up and have some level of experience in this field so you can understand how to apply and interpret these guidelines.
Then there are the more complicated problems. There’s an entire continuum of things, and there are many things that everybody can do (and actually should do) and shouldn’t have to hire expensive consultants to get the benefits.
Then there are other things, where high level of expertise, deeper level of insight – definitely are needed.
One way of looking at the situation is that for the last 24 years or so, I’ve worked at putting myself out of business, teaching other people how to do it. And now I have more business than I ever had.
There’s nothing wrong with empowering people to do usability themselves, because there’s endless need for usability work in the world. I’ve always been extraordinarily oriented towards working on making the techniques more accessible, more easy to use. I work on making the guidelines more explicit, easier to read, evangelizing them to larger and larger numbers of people, and teaching what we know to others.
There’s no reason that somebody who starts today should make the mistakes I made in 1985.
I don’t like when people say ‘the secrets’ of usability or ‘the secrets’ of web design. There are no secrets. It’s like with astronomy. There are no secrets of astronomy. Just point your telescope up to Jupiter and you can see how that planet looks.
It’s the same here; do the user testing and you can find out how people behave. It’s just that most people don’t bother doing it so there’s some value in having us say well, we have tested 100 people using these different websites and here’s what they do when they get to this type of product page or this type of search results, or whatever it may be.
There’s always more work to be done in this field. I don’t think anybody who is good at usability needs to ever worry, because we’ll always have more things to do.
That of course shouldn’t really annoy me because it’s just human nature. There’s always this desire for the magic bullet. Now it’s things like Ajax or whatever.
But for the classic, corporate website, ecommerce site, government site, American Red Cross, non-profit site, presidential candidate campaign site – whatever type of site, communication is typically one of the main goals. And yet they can’t write.
This has been a wonderful conversation. I want to thank you so much for your time today.