N.B This is the unedited transcript of the interview.
Can you introduce yourself?
I'm Jakob Nielsen, and I'm the principal of the Nielsen Norman Group, which was founded by myself and Don Norman.
You recently published a book called Mobile Usability.
Exactly. Together with my co-author, Raluca Budiu, who is one of my colleagues in this company, Nielsen Norman Group. We've been working on Mobile Usability for many, many years. The first study we did was back in 2000 when mobile was so terrible that there was no reason to really write a book about it other than to say, "Don't design for WAP phones." It was so bad. I used to say that WAP means, "Wrong approach to portability." But recent years have become good enough that's it worth really giving people guidelines for how to design for the kind of current generation of smartphones.
You say that mobile websites will soon beat mobile apps. Why do you think that?
I would say today, apps still tend to beat websites. One of the reasons is that mobile websites are often not truly mobile optimized websites, but rather kind of a version of a generalized webssite that's trying to target desktop computers, tablets, and mobile phones. You can make it more optimal for that platform, but the downside, is that it requires some specialized development resources, and most companies can't afford that, therefore, they end up doing a half-baked job.
Whereas the benefit of doing it as a mobile website is that it becomes a little bit more capable of automatically adapting to a variety of different phones. That has a benefit for companies with fewer resources, which are most companies. I would say if you are a big company or if you have something you're doing on the internet that is really highly valuable, which can be the case for a small company. Company with twenty people can make an application worth billions as we've seen here recently. If you have something that is very highly valuable, definitely it's worthwhile doing an optimized app for each of the platforms.
But for most companies, not so much. As we move into the future, we will see that web coding is becoming more capable, also there are more tricks to not just have one website to fit all, but rather have the website respond. I would say adjust, and thus it can self-optimize. If you code it correctly, it can self-optimize for each individual phone or each tablet.
I really want to point about tablets. There are two main classes of tablets, by which I mean size wise, not Android versus iPad, but size wise. What you might call full size tablet and sort of the mid-size tablet. They really are quite different. We've done a lot of user testing of those. The sort of the smaller tablets, the mini tablets, if you just take a regular tablet, a big sized tablet and squeezed that application down, or that website down, you're not getting good usability. Conversely, if you take a phone design, and you just enlarge it a little bit, then it will work, but it's not going be that great either. If you want to have great usability, you have to design for each of these different screen sizes. That of course, again becomes a bigger job as the different platforms multiply.
With all these different screen sizes, testing, it then becomes overwhelming, no?
It does, and that's one of the benefits of going the web app approach. We all know that the standards are not true standards, and they all have their bugs.
I remember I worked at Sun Microsystems back when we invented the Java programming language. The slogan for Java was, "Write once and run everywhere." That was the official company slogan, but my version of the slogan, which I never told was, "Write once, maybe run everywhere, but debug everywhere for sure." There are all the little tweaks and differences in each individual version.
We still have that same problem today. There are of course, different types of testing I want to point out. There's the software testing. Does it even run? Does it crash? Then, there's a usability testing, which is what does the user experience? Obviously, there is no user experience if it crashes, so you want to eliminate those. I would point out, one of the simplest things in designing for any touch screen is the size of a touch target, which needs to be kind of finger sized for what we call, "Fat fingers."
Typically, we say that one by one centimeter is the minimum size. Almost all applications have much smaller touch targets. Even like Netflix, which is a nice video application that I otherwise like and use a lot, but some their buttons are just too small. Particularly if you're trying to use it on treadmill, you're bumping up and down a little bit as you're deciding what to watch. It doesn't work.
What app has immediate utility?
Well, something that sometimes fades to the background, does not intrude on you. I think a lot of designers want to wave their flag and say, "Hey! Look at my design," but the truth is that users don't want that. They don't go to a website to admire it. They go to a website to get things done. The same is true for apps. Now, the controls have to, when you are using them, be big enough that you can touch them. They have to be explained clearly enough that you can understand them, so there has to be some verbiage.
One I would like to kind of point out is the Amazon Kindle app. The Kindle app and Kindle device actually, because they have the hardware version and they have a software version. Which is another interesting aspect of today's modern world, that we sometimes call ‘cross-channel’, something that can work in a lot of different environments. One of the nice things about the Kindle is that, first of all, as you're reading, everything kinds of fades away and you're just looking at the book really and just turning the pages, which is a very simple user interface.
On the other hand, let's say you’re waiting at the dentist, you can read on your phone, and the if you're sitting at home in your nice comfy chair, you can read on your big tablet, or if you're on a business trip, you can pull it up on you laptop. It remembers your place in the book for all of those. We call it, "seamlessness" in the cross-channel, so as you move from one channel to the channel, it's seamless. It just transfers. It remembers you from one to the other, which is a very simple usability. We also have an article about it on our website, but it's an easy guideline. It's actually harder to make that happen.
Why do designers get in the way of the user? We are meant to be facilitating whatever problem they're trying to solve, but we just tend to sort of create that friction.
One of it is just featuritis, "We can do one more thing." You could easily come up with a scenario under which users would want or would need that particular thing. There's always somebody who would like to do something. If you go and look at Microsoft Office, which for decades has been criticized for having too many features. Those features are not there because Microsoft wants to make it harder to use a product. Each feature is there because some customer, some paying customer wanted it. Like law firms want to have line numbers, so they can refer to line seventeen on page eighteen.
You can see how adding numbers to the lines in a document can be useful for a law firm. That said, it's completely useless to me. Yet, you have this complexity of the user interface. The challenge here becomes how to first of all, say no to as many feature requests as you can. Then secondly, have a way of prioritizing the features.
It's actually sort of wrong thinking that all design needs to be equally approachable or equally apparent. In testing websites, we have found that the right hand column of websites tend to be looked at less. Some people think, "Wow. We can't put anything in the right hand side of a web page." That's not true. You can put secondary things there. You just have to recognize and know that it's not going to be looked at as much. We'll know it will not be looked at as much, but some things are primary and some things are secondary, and we have to have the courage to say, "Yes. The secondary things go in a place where it's not looked at as much."
It's hard to keep on this, but one of the plagues that's really on the web right now is overuse of carousels and this notion that home page can't have one main feature. It has to have ten of them, and you rotate through them at either a rapid pace that nobody can read them, or so slow a pace that nobody knows that they're there because they always scroll past.
It's not a great way of exposing richness of content, the richness of offerings, and yet I believe carousels are often there because of political considerations. For example, the vice president has an important release this month. It goes on the home page, and five other people have something important to say. There's no design director who is strong enough to be able to say, "No. We got to pick one as the main feature, and then we can five others as secondary features, and that's it."
Back to the Microsoft case of trying to satisfy the lawyers, that was a business decision, not a design decision.
No, of course. Definitely you would probably sell more to law firms if you include this feature. If you do not include it, maybe some other company will come up with, ‘writing for lawyers’ and sell into that customer base. In some sense, all design is business decisions ultimately because you have to have a business and if it's a website, it has to sell. If it's a software application, you have the have the customers for that, and so forth. Designers always trade off. That's for sure. I'm just saying that we need to put more emphasis on the criterion of being easy to use.
Again, that varies depending on what type of project we're talking about. Because if we're talking about an enterprise application that’s, "Mission critical," that people use every day maybe for several hours every day, well then, we can afford to send them to a training course. It's more important that they have a lot of power at their fingertips, and so it's a very full-featured application. These complex applications still have a lot of usability issues, but that said, they should be full featured.
On the other hand, the vast majority of apps and website are really almost used once or intermittently. There are some websites that people go to all the time like Facebook, but most websites are not used that frequently, and that means that users will not really learn all of the features, so you have to reduce the feature set there.
I've recently released an app, and one of my concerns is, in trying to keep it simple, it's much easier for a competitor to copy. What do you then have to sort of say, "We're better because we do X," when the competitor can do X also quite easily.
That's true, but the question is can they do it as well. There's a lot to be said for polish, of "fit and finish" as we sometimes call it, and having everything be truly well integrated and no rough edges. There's a lot of things in the background that will help make things just run more smoothly. If we think about search engines as an example. The most popular engine is just one blank box you can type in. Particularly if you think back to fifteen years ago, there were a lot of other search engines that had a different design approach, which had very busy home pages.
Then, Google came out with, "I'm just going to give you one box to type in." Again, user experience is the entirety of all that meets the user, not just the dots on the screen.
Behind that, there was a lot of smarts. For example, if you type something, if you make a typo, and most people can't really spell that well, it will still tend to find it correctly, as opposed to having a very literal interpretation of exactly what you typed. There's no end to that. You can employ as Google does, thousands of programmers to constantly improve that interpreting, what people really mean to say rather than what they did say. That can continuously get better and better and better. There is a lot of things one can do there in the background to make it smoother, just work better.
That's a high value to users as well, but I have to say that it's often difficult to communicate. I do believe that in the technology industry, the marketing side of our industry has not been great communicating value propositions of simplicity and integration. It's almost as if people assume that if it's technology, it will crash, but not everything crashes. Not everything only does exactly what you say. Some things actually understand better what you mean to say, and so we need to be able to communicate that better. The easiest thing is the feature checklist. That is the easy way of promoting a product. The harder way is to really communicate it works well.
You sort of touched on that during your talk at Google about the projector, and how you had initial difficulty of getting started. Do you remember that?
Oh, yes! I was just trying to connect a regular laptop to a regular project, and a small twist, which was play video with some sound. That's really another big twist these days. That's a normal thing, and yet it didn't work. Luckily, they have a very good AV guy who came up and tweaked all the cables five times, and he knew the things to try. It finally worked, but it really should be plug and play, not "plug and pray" as they say. It's a little bit sad that to this day, we have this problem.
I have to say though, I remember how it was to give presentations twenty years ago and even ten years ago, where it was anybody's guess if you could get projector to work. Nowadays it is actually better.
Are things getting better?
Yes. They are getting better. It's just that they are getting better from such a low point that the glass is not full yet. There's still so much to be done. Particularly if we think about the true mission, the human factors field, which is to make technology adapt to humans, put humans in control, empower humans, and empower not just a little minority of elite people, but rather a vast base of almost everybody, and we are not there yet.
If you think about the majority of the population, how they have no technical skills whatsoever, they have relatively low levels of education. I'm not even talking about people who have disabilities. We are not even catering to those at all. Accessibility is very poorly done. Let's not talk about those guys, let’s just talk about people who have a high school education. Using technology today is very hard for vast groups of people.
What needs to be done?
From my entire career, which is now more than thirty years, we've had the one, kind of broad answer to this question, which is called, "User centered design." That means not, "Ego Centered Design" or centered on what I feel should be good, but rather centered around users and user needs. It starts with discovering user needs because if you're designing something that doesn't meet user needs, it doesn't matter how good it is, if it's meeting the wrong need, it's not going to be good. You have to figure out user needs.
We have to focus early on users, involving users throughout the design process, iterative design, knowing that there's no such thing as the perfect user interface. We have never yet seen the perfect user interface. Don't assume that your design is going to be the first one in the history of the universe to be the perfect user interface.
Iterative design is another thing that's really important for quality of user experience. A good design does not mean it could not be improved more. Sometimes we face this problem, in particular for like design agencies, like if they say to their clients, "We should really build in some time to schedule for doing like a few rounds of user testing." The client says, "But I thought you guys were good designers. Why do we need to test your design?" Yes. They are good designers, but they can be even better designers if they improve the design through iterations.
Another analogy would be to hire a programmer saying, "Well, you're a good programmer. There's no need to debug your code." That's just not how the world works. People can't build a perfect program the first time they sit down to code, and similarly the best designer can't design a perfect user interface with their first attempt. These are the basics, so focus early on users, involve users throughout the design, an iterative design. Not really that complicated, but you just have to do it.
People don't do it.
No. They don't. There are all these different reasons. One is this sort of fallacy of the genius designer thinking, "Well, I'm good enough so therefore ..." That's a natural thing, right? To think that. Or the client is saying, "We're paying you to be good, so therefore do your best work and that should ought to be good enough.” For either reason, that's certainly a big problem.
There's also the problem that people do it wrong as well. I only touched on some of the broad rules here, but when we say focus on users, that doesn't mean things like focus groups because you can't ask users, “What do you want”, and they will tell you because they’re not designers. They don't know what they want, so that's not the methodology. If you show a demo of something and say, "Does this look cool to you?" Yeah, maybe it looks cool, but it would be terrible to use. If you ask people to speculate, "If we built this feature, would you want use it?" "Well, it sounds good. Yeah. I would like to use it." Those are all the wrong methods.
This approach is common because this is the way people tend to do testing marketing campaigns and branding propositions. There's a big tradition and very well-funded tradition for probably a hundred years or so of different types of marketing research, and yet that's the wrong approach for design research. That's a subtlety that takes a while to get through to old school management, who is used to doing things one way.
The right methodology is to say, "Perform this task," and watch them?
You have to watch people, watch what they do, not what they say, and you have to have to do actual tasks. Now, luckily we can have them do actual tasks with very simplistic, sort of simulations of a user interface. You don't have to actually go through the trouble of building it. You can do paper prototyping, very simple mockups. Of course you know it's just a piece of paper; it's not a computer, but you can still kind of make believe that it's a car reservation system, and you go in and book your car from the airport for your next trip or something like that, even if of course nothing really happens, but people can make believe.
You can see a lot of the things that trip them up. Like, “The photo of this car doesn't tell if it's going to be easy for my old grandmother to get in the seat," or something like that.
What's the reciprocity principle?
Well, the reciprocity principle says that whatever people do to you, you feel likely to do back to them. It's a part of social psychology, which comes from people having evolved in small groups back in the really old days. If somebody's nice to you, you'll tend to be nice to him or her because that's the only way you can get cohesion in the tribe. On the other hand if somebody's nasty to you, you'll tend to want to punish them. If you think about in terms of interaction design, it means that if the design of the site or the app does something good for you, you will be more likely to feel positive towards it and want to do something.
Conversely, if they’ve done nothing good, why should I do something good for them, right? One of them goes for things like registration, which is a classic usability problem, which we actually discovered in testing back in 2000. Websites have premature requests for registration, "Come and join our community. Have a relationship ..." You don't want to have a relationship with people that early.
To this day it's classic problem on a lot of the mobile apps. You install the app, and the first thing that happens when you start it up is it comes up and says, "This app wants to send you notifications," and basically disrupts your entire day. Most people say, "No. Let me first get to know and like you." Then, the reciprocity principle will kick in. If you do something for me, I'll do something good for you, or I'll say, "I trust you," and I will want your notifications, and I'll register with your site, and I'll allow you to store my credit card number, and to sort of build gradually like that. You first have to do something good to the user, and then they will reciprocate.
But you understand why one does that right? We're concerned that we might not capture that user.
Oh, of course, but the problem is that quite often you don't catch them anyway, and it's often too late. You start off on this kind of nasty footing, you start by annoying them. Again, if you think back to the reciprocity principle, the dark side of reciprocity principle is that if somebody does something bad to me, I feel like I should do something bad to them in return.
We want to punish cheaters, and there's this notion of fairness, and that comes from, "That is the only way you can run a tribe". That if people can get away with just being nasty and nobody does anything against them, everybody will just look out for themselves, and you will never be able to put a team of hundreds together and bring down the holy mammoth.
What epiphanies have you come across recently that you think we designers need to hear more about?
We need to look more at going from designing for me to designing for you. Honestly, it is so easy to believe that if you design something that you like, then it must be good.
I think that.
We all think that. We just have to have the discipline to step out of that role because it's natural to believe that I am the center of the universe. If you want to design for people, well, you can design for yourself, and it will work. The problem is that you're not representative of the target audience in so enormously many ways, because most people are not designers. If you know technology, most people don't. In any case, you for sure know your own project because you're working on it very intensely. Whereas for the average person, they may spend half minute on a web page or few seconds on an application screen before they click on something. There's a huge, huge gap there in skill and knowledge and motivation as well.
Knowing whom you're designing for is sort of what you're talking about, correct?
That's definitely correct, but it goes beyond knowing it because you can never truly know. This is why we had this notion of having to keep the focus on users and keep this kind of reality check of doing user testing. It is something though that you can drill yourself into becoming more aware of these differences between yourself and others. I do think that this one of the characteristics of really good interaction designers as opposed to people who are just good at designing static things because interactive design is meant to be used, whereas an advertisement is just meant to be looked at. There is a big distinction there in kind of understanding that context of use and what happens during use.
The only way you know is by actually looking at people using your design, so if you do that over the years, you do kind of become more skilled at almost like stepping out of your own body, having this kind of almost out of body experience, of how would another person use it? We have to in some ways become the professional outsiders, not just the loyal employee in the company, the good team member, "I'm building this thing for my team and my company," but you also have to step out and think, “Well, how does it look from the customer angle?” That's something you can become more skilled in doing. You can never become perfect. You can never really be another person, but you can become more skilled at not only just being yourself.
How can I become more skilled? I understand that usability testing is part of that. Is there more we can do?
Well, you can certainly also read a lot of existing research because people like myself and many, many others have studied this for decades now already. There's a lot of material that is known. What I do see though is that when people read, let's say an article I wrote ten years ago, often they will say, "But surely, people don't do that anymore." No. Usually people do that. It's not always true, so generally speaking it's maybe sort of between eighty and ninety percent of the old findings continue to hold because there are some changes that relate to more with habits and people accustomed to something.
For example, let's talk about icon design. Do people recognize that a magnifying glass is an icon representing search? Well, it's a little bit of a stretch, not a big stretch, but it's a little of stretch. Maybe the first time anybody drew a magnifying glass icon, people wouldn't have necessarily gotten that point, but today it's very commonly used, so therefore today, we know from user testing, today it's understood.
Another icon is the hamburger icon, three stripes almost. It's supposed to represent the menu. That's also a little bit of stretch because it doesn't really look like a menu. The first time those came out, we did see people have issues with them. But it's becoming more recognized now.
Even go back to something like let's say traffic signals. Does green mean "Go" and red mean "Stop"? Well, yeah. That's the convention, right? There's no rather inherent reason that should be the case, but because you've seen it a million times already, you completely think that that's how it is. In a traffic light, the reverse though, would just be crashing cars right away.
Sticking with conventions is really important because once people learn something, you are leveraging millions of hours of human time of learning something. It's actually very arrogant to throw away that collective investment of humanity in learning something and say, "Well, I would like to, instead of using a magnifying glass icon, I would like to use an anteater icon to represent search or something like that." Don't do that. There's a lot of that arrogance around, that things people have learned, "Let's just throw it away." No.
Some change because that learning does kind of take place. On the other hand, speaking of resistance to change as well, and things that people have learned, they kind of tend to stick to. Also, a lot of other things, not so much to do with learning, but rather to do with the way the human brain works. That for sure doesn't tend to change very much over time. For example, we know that a lot people have really low literacy levels, and they're not great at reading complicated information and long convoluted sentences and fancy long words. Write simple language.
We keep finding it every new test we do. A lot of things, or let's say people's memory, how many things can you keep in your short-term memory? Well, it's about seven plus minus. People who have great spatial visioning skills and just some people have more powerful brain than others, and they tend to be the ones in our types of jobs because if you have a lot of skills to keep all these things in your head at once, you tend to be able to do this job well. We tend to honestly be a bit smarter than the average person, but that's no reason to celebrate. That's a reason to say, "Remember the average person cannot keep as many things in their mind at one time as you can."
As people navigate a website, how many things you require them to remember? As a simple example, the guideline to change the color of visited links. Make sure that people can tell what they've seen, what they've not seen. That's a huge memory aid for the average person, and yet for many designers don't like to do that because it sort of makes it a little less clean looking page, but it honestly it's very, very helpful. Again, that doesn't tend to change because it's a matter of the capacity of the human brain, which does not change. It's the same now as it was a hundred years ago, as it will be in a hundred years.
Is good design universal? I'll explain what I mean by that. I have an app, and I have two types of users. One's seventy, and one's twenty-two, and they both use my app. Knowing that, should that aid the way I approach how I design this app? Can't it just be a good design, and it will solve the twenty-two user's problem and the seventy year old user's problem? Do you know what mean?
Yeah. I completely know what you meant. I think generally speaking, yes. There's one thing that's called good design, and it will mainly work for everybody.
But that said, it's very easy to abuse that insight and then say, "Well, I'm only going to design for the twenty-two year old." If you do that, then you may in fact exclude the seventy-year-old users. The simplest example of this is the font size because if you use tiny type, then a young person with keen eyesight might be okay with that, and the older user with reduced vision may not be able to see it, and they will give up using it. On the other hand, if you pick a bigger font size, or you have a way of changing the font size, so there will be two different solutions, then you will serve both audiences because we don't see young users say, "Wow. I don't want to use this app because the text is too big."
In that sense, the same design will be good for both type of users. The same is true for designing for around the world. When we do studies in a variety of counties, different continents, all of the big findings tend to be the same because all of the big findings relate to either just basic human nature, which is the same everywhere or the differences between humans and machines, which is also the same everywhere. Now, of course that said, there is still some cultural differences, and we know that certain colors have certain meanings in different countries and all of that.
There might also be some small information architecture differences based on exactly how people in different cultures kind of think about certain type of things. Most of the findings really are the same. On the one hand, when you localize, yes you should think about the new target country, but on the other hand, the vast majority of the design decisions will remain the same.
What are your thoughts with flat design? When I say flat design, what comes to your head, in terms of do you like it, do you hate it?
No. I really do hate it. It's this hot, strong word, but it's almost true. I do think that it embodies some of the worst aspects of design, which is fads and fashion and disregarding the user’s needs, and saying, "Well, surely people can do it." Again, "Well, if they really work at it, they can do it." Why make them work hard? This notion of signaling, what's clickable, where you can press, where you can type. It doesn't have to be blatant. It doesn't have to be so overly, strongly done as it was in let's say maybe Windows 95 or some of those early, not high quality visual design kind of operating systems that we had. Today, I think it's possible to have somewhat subtle signaling, but still have the signaling. Still tell people what they can do.
My business partner, Don Norman, has an entire book about this, ‘The Design of Everyday Things’, which he talks about "affordances", which means what can you do to things. In some sense, a thing has the same affordance, whether or not it's being signaled. Whether or not you can tell that you can open a door by pushing on it, you can open a door by pushing on it, but if you can tell by looking at it, that's much nicer than the trial and error of pull.
It's the same here. Why make it so flat and so devoid of signaling its functionality that it becomes too complicated to use it. We know that every additional complication you add means more failure. I would say it’s a humiliation of humanity because we try something and we can't use it, and we blame ourselves. I don't. I blame the system. I cannot tell you how many user testing sessions I've been through where the users will say, "Well, I'm not very good with technology. I'm not good with computers." or "I couldn't do this." They blame themselves. It's so sad to see. There's no reason for that.
Coming to a close now. What is it that we need to do more of?
Well, I think I go back to usability testing. That is the one thing that we really need to do more of, but I also think that it has to be that humility of recognizing that, I am not the prime specimen of humanity, or at least I'm not representative of the vast masses of humanity. We have to design for the people and not just for what's the current fashion or "Oh. This is a tired design." Well, it's a tired design to you because you work on it, you look at every day for years. The user looks at it for ten seconds every three weeks. It's not tired to them.