In its unscientific form, contextual inquiry is simply observing people do what they do. As professionals, however, we can be more effective by applying some science to the approach.
In a nutshell, the following items are recommended for a good contextual inquiry.
1. Understand your audience – do some research on what they do, what their tasks and goals are, and the unique characteristics of the environment in which you will observe them.
2. Make a list of what you want to observe – time with users is usually a precious commodity, so make a wish list of what you would like to get out of the session.
3. Plan a technique for getting rapport – Beyer and Holtzblatt (see reference below) identify several approaches such as the Master/Apprentice. In this approach, you are the apprentice observing the master at work.
4. Define all of your measuring and recording techniques in advance – you may wish to use audio, photography, video, etc., but even for simple note taking you may wish to prepare note sheets in advance to make the best records possible. For example, a "swim lanes" note sheet can be used to record the interactions between what the user is doing vs. what the system or tool is doing.
For more in depth information, you can look to two very good books. The first is Beyer and Holtzblatt’s Contextual Design: Defining Customer-Centered Systems. The second is Mike Kuniavsky’s Observing the User Experience.
