An interview can explain how work is supposed to happen. Watching someone do it may reveal the extra spreadsheet, the message to a colleague and the scrap of paper that keeps the process moving. Contextual inquiry combines observation with questions during the activity, so the researcher can investigate those details while they are present.

The participant acts as the expert in their own work. The researcher observes, asks for clarification and checks interpretations, with a defined area of interest to keep the session useful. The method is associated with Contextual Design, developed by Hugh Beyer and Karen Holtzblatt, and its emphasis on learning from activity in context.

Let the work set the conversation

Suppose a fictional repair company wants to improve how technicians record a completed visit. In a meeting, a technician may describe entering notes into an app. During a visit, the researcher might see that the notes are postponed until there is mobile reception, while a photograph and a voice message temporarily hold the relevant details.

That observation changes the questions worth asking. The difficulty may concern working offline, remembering what happened or coordinating with a colleague, rather than the design of the notes field. The researcher can ask about the actual sequence and check whether the same arrangement applies on other visits.

InContext Design’s guidance on contextual inquiry stresses the importance of seeing people perform their real activities. A session that becomes only a conventional interview in a different room loses much of what makes the method useful.

Prepare access as carefully as questions

Choose an activity and recruit people whose work can illuminate it. Agree what can be observed, photographed or recorded, and address the presence of customers, colleagues or sensitive information before the session. Permission from one participant does not automatically cover everyone and everything in their environment.

Begin with enough background to understand the role, then make the transition to observation explicit. Ask the participant to continue as they normally would where that is safe and practical. Avoid turning the session into a demonstration of the official process merely because the researcher is there.

A remote session can capture some computer-based activity through screen sharing, but it may omit physical materials, nearby conversations and other context. Describe those limits in the research plan. A recorded walkthrough can be useful too, although a reconstruction is different from observing the activity as it happens.

Ask without taking over

Questions should follow consequential moments and respect the demands of the work. Interrupting someone during a delicate procedure may be inappropriate; noting the moment and asking later may be the better choice. The researcher needs to learn from the activity without becoming an avoidable obstacle to it.

Offer interpretations tentatively. “What is that list helping you keep track of?” leaves more space than “You use that because the app is unreliable, right?” The participant may correct the researcher’s account or reveal that the workaround serves several purposes.

Compared with shadowing, contextual inquiry typically gives questioning and shared interpretation a more explicit role. Neither label dictates a universal session length or guarantees that every part of the work will be visible.

Preserve the context in the finding

Record the sequence, materials, constraints and explanations together. In the repair example, a recommendation to add offline notes is more persuasive when it is connected to the loss of reception and the temporary methods technicians use, rather than presented as an isolated feature request.

Use task analysis or affinity mapping to examine the material across sessions. Check differences as well as common patterns. One workplace’s workaround can suggest a valuable question without establishing that every customer has the same need.