Jakob Nielsen's Alertbox, June 27, 2005:

Usability: Empiricism or Ideology?

Summary:
Usability's job is to research user behavior and find out what works. Usability should also defend users' rights and fight for simplicity. Both aspects have their place, and it's important to recognize the difference.

There's a duality to usability. On the one hand, it's a quality assurance methodology that tells you what works and what doesn't work in the field of user experience. On the other hand, usability is a belief system that aims to ensure human mastery of the constructed environment.

Both perspectives are valid, but it's important to recognize that these two sides exist and know when to apply them.

Usability as Empiricism

The economist Arnold Kling recently summarized the long-term growth in the economy by a somewhat peculiar metric: flour bags. When measured by how many bags of flour you can buy for a day's wages, the average worker today generates 430 times the value of a worker in the year 1500. (Kling uses bags of flour to compare productivity because it's one of the few things that has been produced continuously and provides the same benefit today as it did in past centuries.)

Ultimately, Kling's point is that the economy is a learning mechanism. This huge increase in wealth resulted from a gradual accumulation of learning: We figured out how to do things faster, cheaper, and more efficiently one improvement at a time. In turn, these improvements built on each other to create ever-increasing productivity.

Learning advances are generated in two ways: the scientific method and Darwinism. In the scientific method, we think up hypotheses and perform experiments to falsify them. If enough studies fail to falsify a hypothesis, then we start believing in it, and engineers can use it to build better products. Business works more Darwinistically, where multiple entrepreneurs bet in parallel on their ability to satisfy a customer need. Most of them go out of business because the economy's "invisible hand" rejects their proposals.

Whether in science or business, the basic point is the same: you propose a solution, then see if it works in the real world. Hypotheses that work become accepted scientific theory; companies that offer the most value to customers become established in business.

Usability is also a reality check. There are two main ways for usability to derive value from reality:

Before design begins, usability methods such as field studies and competitive studies are used to set the design's direction based on knowledge of the real world. These methods are similar to the scientific method's hypothesis-checking elements: you discover principles that explain observed reality and then use them as a guide to build products that are more likely to work.
After a design has been created, other usability methods, such as user testing, determine whether humans can understand the proposed user interface. Just as entrepreneurs compete to see which business ideas create the most value for customers, usability specialists show customers alternative interface designs to see which one works best. The main difference is that it's much cheaper to test a paper prototype of a design than it is to start a company.

Usability explains human behavior in complex systems under strongly context-dependent circumstances, and its predictions are less exact than those of hard sciences such as physics. Usability therefore compounds its findings from past empirical work in guidelines rather than exact formulas.

When something causes problems for many users on many different websites, we issue a guideline warning against it. Similarly, when a design element works well under many different conditions, we issue a guideline recommending it.

Despite these differences, the fundamental approach of usability and harder sciences is the same: conclusions and recommendations are grounded in what is empirically observed in the real world. The job of usability is to be the reality check for a design project and -- given human behavior -- determine what works and what doesn't.

Usability as Ideology

At the same time, usability is also an ideology -- the belief in a certain specialized type of human rights:

The right of people to be superior to technology. If there's a conflict between technology and people, then technology must change.
The right of empowerment. Users should understand what's happening and be capable of controlling the outcome.
The right to simplicity. Users should get their way with computers without excessive hassle.
The right of people to have their time respected. Awkward user interfaces waste valuable time.

These rights have not always been highly valued. In the 1960s, many user interface designs were oppressive, subordinating humans to the needs of technology. Same goes for many websites designed in the "killer site" days.

If designers and project managers don't believe in the usability ideology, why would they implement usability's empirical findings? After all, if you don't want to make things easy, knowing how to make them easy is irrelevant.

Respecting users' rights makes people happier and thus makes the world a better place -- nice, but not reason enough for most hard-nosed decision makers. Luckily, the Web provides a very clear-cut reason to support the usability ideology even if you only care about the bottom line: If your site is too difficult, users will simply leave.

On average, websites that try usability double their sales or other desired business metrics. The reason? The Web is the ultimate competitive environment, and users won't invest their time and mental resources in struggling with websites that violate their right to simplicity. There's always another site they can visit instead.

Balancing the Two Perspectives

As a user advocate, you need both perspectives: usability as empiricism and usability as ideology. Each perspective requires a particular approach.

When taking the empirical approach, you must be unyielding and always report the truth, no matter how unpopular. If something works easily, say so. If something will cause users to leave, say so. The only way to improve quality is to base decisions on the facts, and others on your team should know these facts.

In contrast, when viewing usability as an ideology, you must be willing to compromise. Sometimes decisions must be made that will lower the design's usability quality, either because of limited time and budget or because of trade-offs with other desirable qualities. Of course, project managers can only make good trade-offs when they know the facts about the design elements that help or hurt users. It's the role of the empirical approach to supply those facts.

At most project meetings, everyone has a seat at the table except the poor victims who will have to operate the technology. Often, the usability specialist is the only user advocate in the room. As such, it's crucial that you argue for the importance of usability findings, and supplement bare facts and numbers with ideology and advocacy.

> Other Alertbox columns (complete list)
> Sign up for newsletter that will notify you of new Alertboxes