Interview - Jakob Nielsen, Ph.D.

SP: You recently signed a deal with Macromedia to help improve Flash. What work have you done there so far? How do you feel Flash needs to be improved to be more usable, do you still dislike it?

The Flash software itself has already fixed most of the big usability problems that I complained about in earlier versions of Flash. Even though any software product can get better, the key priority is no longer to fix Flash, but to fix the Flash designers. Unfortunately, this is going to be harder, because I need to change the thinking of millions of designers and not just a few executives at Macromedia.

The Flash usability project has resulted in three deliverables so far:

1. A thorough report on Flash usability with 115 usability guidelines, based on a research project where we tested the usability of 46 Flash designs in the U.S., Europe, and Asia.


2. A 53-minute film, documenting how users interact with Flash designs and showing video highlights from the usability testing.


3. A preliminary report on how to make Flash designs more accessible and usable for users with disabilities, with 21 guidelines.

The accessibility guidelines are only preliminary because we did not have sufficiently many accessible Flash applications available for the research study to come to firm conclusions. Most likely, we have overlooked some guidelines because we didn't get enough data in the first round of testing. We still decided to publish the accessibility report together with the other two deliverables because it is better than nothing and will be of some help to those Flash designers who want to increase the accessibility of their designs. Hopefully, there will be more accessible Flash designs shipped in 2003 and we are already planning on producing a more comprehensive second edition once we have more data.

I am particularly pleased with our video because I think it will convince many Flash designers to pay more attention to usability once they see how real people struggle with many of the usability problems that are found in many current designs.

SP: Will usability issues put the small design shop (10 employees or less) out of business, or make them more competitive? Usability needn't be expensive - what, in your opinion, are the top 3 measures that all designers should take to ensure that their sites are usable?

Anybody can include usability in their design projects. My current recommendation is to invest 10% of design budgets in usability, so a 10-person design shop might consist of 9 designers and 1 usability specialist. Smaller companies could still do their own user tests by having one of the designers take on a part-time usability role. The basic techniques are easy enough to learn that you don't need to be a full-time usability specialist to run studies. You do get better with experience, and the advanced methods are more difficult to learn, but anybody can run a simple test and there is no excuse for shipping a design that hasn't been tested.

All designers should take the following three measures:

1. Test your design with real users (or, if you are lucky enough to have a full-time usability specialist at your disposal, sit in on the test sessions; don't just wait to get the report because you will learn a lot from watching the users yourself).


2. Check the design against established usability guidelines. This will discover many design weaknesses and will cause you to question some of your design decisions. Sometimes it can be OK to break a rule, but you have to know the rule first, so that you don't break it because of an oversight.


3. Run some user tests of other designs that attempt to solve similar problems as your own design. These tests can be short because it's not your job to fix these other designs. But you will gain deep insights into what works and what doesn't work with design approaches that other companies have spent a lot of money on.

In a real project, you should perform these activities in the opposite order of the one I listed: whenever you start a new project, learn from the usability of other designs, then create your own design and clean up its usability through guidelines, and finally test the revised design with real users. With this sequence of steps, you avoid wasting test users on something that's bound to fail because it hasn't been driven by early usability data or cleaned up by a guidelines inspection.

SP: Which sites, in your opinion, do a good job of combining usability and visual attractiveness?

Williams-Sonoma. Even though they do violate some usability guidelines, such as the ability to make the text bigger, I am very impressed with their product photography which provides content that's very suited for the Web. It's an example of graphics that help rather than hurt.

SP: As input devices, the keyboard and the mouse are both hard-to-use and lead to repetitive-motion ailments. Speech recognition aside, what ideas do you have that could make computers more "usable" devices?

I think the main advances will come in display technology. We need much bigger screens and also higher resolutions. Truly high resolution will help with mobile usability by allowing for more information in a limited space, but PDAs will always be smaller than desktops.

In the future, it will surely be common to have a desktop monitor with 10,000 by 8,000 pixels or more. I personally use an IBM flat panel with 2048x1536 resolution, and even though this is far from what we need it's still a vastly superior user experience to using a small screen as everybody else does.

Regarding input, we will probably get gesture recognition and computers will use video cameras to observe the user. Still, I think that the main research direction should be to minimize the need for users to say everything in as great detail as we do today. The computer should be able to act on some things without the need for explicit commands. At the minimum, we should use simplified commands and be shown better sets of choices to pick from without having to wade through as much crud. Google shows the way.