The Evolution of Usability

Expert Joseph Dumas has seen how changes in software interfaces have changed the way we work

Joseph S. Dumas is a nationally recognized thought leader in user experience methods. Since 1982, he has written books, taught at universities, and consulted for major corporations on the subject of usability. Profit Online spoke with Dumas, now a user interface consultant with Oracle, about the evolution of usability, the impact of user interface on user productivity, and what we can expect from the next generation of Oracle products.

Profit: How did you get involved in the subject of usability?

Joseph S. Dumas: I have a PhD in cognitive psychology, and I started out doing basic research on short-term memory and teaching at a university. Then I started working in applied research areas. This was during the ‘70s when the usability profession didn’t exist, but there was a precursor profession called human factors. I took a job with the U.S. Department of Transportation and after five years of working with the government and doing applied research, I joined a consulting firm that was doing work in human factors. [Note: human factors, also known as ergonomics, is the study and design of human machine systems of all kinds.]

In 1982, for the first time, people working in human factors and psychology got together and began to realize that there was this blossoming new area—the usability of software. It’s hard to believe now, but the CRT [cathode ray tube] screen was just spreading—computing was moving out of computing centers and onto people’s desks.

One of the things about software is that you only see its surface. It’s the interface that’s viewed on the screen, while the rest is hidden. In order to understand how to use software, the user needs to develop a mental model of how it is organized. A usability specialist tries to figure out what that mental model looks like.

I wrote a book in 1998 called Designing User Interfaces for Software, and started working on designing software with companies. I also started helping to develop methods, such as usability testing, that were just at their beginnings. Soon, usability laboratories started, and things took off. People interested in usability, which had been part of human factors, started evolving a new profession. We refer to these people as usability specialists, and now we have the broader term “user experience” with several thousand people in the profession.

Early on, there wasn’t much research being done—but there were a lot of consulting opportunities with major high-tech companies. I worked with Lotus, Microsoft, Hewlett-Packard, and Digital, helping refine their methods to measure and improve the usability of their products.

Profit: What were the usability issues software companies were struggling with in the early days of the industry?

Dumas: The biggest issue at the beginning was whether usability was important or not. In the early days, functionality often overwhelmed the usability issues. When Lotus jumped in with Lotus 1-2-3, they didn’t have to be particularly concerned about its ease-of-use because it had no competition, and it had so much more functionality than anything else. But as the ‘80s went on into the ‘90s, things changed. First of all, the Macintosh came out. Right from the beginning, they pushed the ease-of-use of their product. Then, more competitors came into the national software market—Microsoft came up with Excel, and there were three or four companies putting out suites of software. All of a sudden, the competition for ease-of-use became important, because the basic functionality was there for all of them.

Then in the early ‘90s, there was what I would call the “suite wars,” when Microsoft started pushing Office as a suite. They had three or four products that all had similar user-interface designs. There are similar menus, they all have “file” and “edit,” and so forth. So they claimed that if you learned Microsoft Word, it would be easier for you to go to Excel than Lotus 1-2-3. Once ease-of-use became important, people started putting it in their ads and touting it without much data behind them—they simply said, ‘our product is easier.’ Both these companies and their customers no longer could dismiss usability.

So the issue became, ‘How much of our resources do we need to put towards usability in our organization?’ rather than, ‘Is usability an important thing for us at all?’ By early 1993, all of the major manufacturers— Oracle, Microsoft, IBM, Lotus—had built usability labs and were doing testing. Then, the internet came along and accelerated everything that was going on.

Profit: How has the field of usability changed in terms of user expectations?

Dumas: The internet changed everything. The fact is that almost everybody now is on computers and the internet. It was a hard evolution to create software for the non-technical users, and in many cases it still is. Also, if you look at enterprise software, one of the trends is that organizations don’t want to make their own applications any more. They want smaller or no IT departments, rather than bigger ones. They want the products to work for everybody in the company without them having to hire consultants to tailor the products and so forth. This is a big challenge for a company like Oracle.

Profit: What usability work have you been doing for Oracle’s next-generation applications? What are the challenges, and what kind of efforts have you been making to address those challenges?

Dumas: First, I’m helping them standardize and document their design and evaluation methods— developing guidelines and templates for the usability folks to use. Second, I’ve been working on ways to measure end-user productivity of software. Productivity, in general, has to do with the amount of output you get for the labor you put in. If an enterprise software product allows you to be more productive, that means you get more work done in the same amount of time. Or, you can do the same work in less time.

Profit: How do you go about finding the places where you’re going to get the most productivity benefit from re-engineering the task?

Dumas: Saving time with repetitive tasks is one way to think about productivity. You certainly get the most benefit from the tasks that are done frequently. Also, one of the big issues with a new product is the way it’s organized and structured. When the structure matches the way people do their work, that’s when you get the most usable products. That means you have to understand people’s work and how they do it. And that means watching customers, talking to them, and measuring what they’re doing. In some cases, it means bringing them into the lab and watching what they do with applications, and thinking about how you can reorganize what they’re doing to make it work.

Profit: Online, people are getting used to functions like drag and drop, partial refresh of pages, and really streamlined interfaces. Does that create that a tension between what their expectations are when they’re on an internet application, and when they work with enterprise software?

Dumas: People do expect that they’ll be able to drag and drop, right-click, get pop-ups, and do all the things they can do with desktop software. They don’t understand—nor should they have to—why they can’t do those things in enterprise software. The new products that are coming from Oracle will incorporate all of those things. People do expect it, and it’s coming.