Usability Empowers Great Products...

Usability is one of the most intangible features of a product, but poor usability will overwhelm even the best feature set. This issue contains ideas on how to measure and improve usability.

Designing for Your Target User...

I recently listened to a senior executive talk about using a new product where operating it was so complex that a rep from the product's company had to be online to operate the product for his group.

Making a product useable is a challenge. Imagine building a feature-rich product that had to be used by: 1) a 50 something CEO who multi-tasks with the phone, in-person meetings, and email and 2) a 20 something who multi-tasks with two instant message sessions, a chat room, and a MUD while reading email. It is a major usability challenge to make a product that they both will like; a product that will not frustrate either of them.

Measuring Usability...

Unfortunately, most people can't accurately tell you how they do routine tasks. Things like brushing your teeth and many tasks at work are done at such a subconscious level that most people can't recall how they do them. Measuring usability needs to be done by observing how they accomplish the tasks.

The first step is to have tasks available to measure. If you have "use scenarios" available from your requirements, they are a good place to start. If not, then you need to develop a few scenarios that are typical of what your users do with your product.

I start with two objective and one subjective measures. Objectively, you can count the number of mouse-clicks and keystrokes needed to accomplish the task or you can time the completion of the task. Subjectively, and more importantly, you listen to the words and observe the body language of the person completing the tasks and note where they are frustrated, confused, make errors or need a shortcut.

Usability Test Setup...

Usability labs come in all flavors. One of the great usability experts I knew sat on a stool in the corner of a control room for weeks and took notes, asking the operators when he didn't understand what they were doing. You can rent high-end usability labs, build a special room in your facility, build or buy a roll-around portable facility, or improvise.

I am in favor of starting pragmatically. You can get started with a conference room, two camcorders, a video mixer to combine the two videos onto one tape (my family room Sony TV can do this) and a VCR. You can find important results with this setup; don't let cost get in the way of getting started.

Running the Tests...

Finding the subjects is critical. You need to find the naive users, the computer intelligentsia, the geeks, the IM gurus, and 50 something executives, but only users representative of your target market.

Timing of the tests is critical. For a software product there is a tendency to delay testing until the end of the project when the user interface is available to test. Unfortunately that is too close to crunch time and if there are major issues it is probably a schedule-buster to fix them. For a new product you need to mock up and test early. I vividly remember forcing a usability test of a training lab on a Friday night and finding that it did not work. That module was fixed over the weekend so 150 customers had a positive experience on Monday.

Listen to your users talk about how hard it is to use your product. Don't dismiss it as "learning curve"... make the interface transparent to the use of your features. It can be done.

There are three sites referenced on the links page that give additional perspective on usability.

Home | Privacy