Synonyms: contextual interview, field study, in-context research, contextual observation
Do not index
Definition
Contextual inquiry is a field research method where you watch users do their real tasks in their own environment and interview them about it as they go. Developed by Hugh Beyer and Karen Holtzblatt, it runs on a master-apprentice model: the user is the expert, and you're the apprentice learning how the work actually happens.
Use cases
People are bad at describing their own workflows from memory. In a conference room you get the tidy version. Watch them work and the messy reality shows up.
Hidden workarounds: A nurse says she uses the charting software as designed. Watch a shift and you find the paper cheat sheet she keeps because the login times out every 10 minutes.
Environment surprises: A warehouse app tested clean in the lab. On the floor, screen glare, gloves that block the touchscreen, and workers holding the phone in their teeth tell a different story.
Tool-switching: A team assumes users live inside their product. Shadowing shows it's one of nine open tabs, and the real work happens in the copy-paste between them.
How it's used in practice
Go to the user. Run the session at their desk, home, or floor, not in your lab or on a scheduled call.
Watch first, ask in the moment. Interrupt when something surprises you, not in a debrief an hour later.
Follow the four principles. Context: be where the work is. Partnership: let the user lead. Interpretation: say your read out loud so they can correct it. Focus: know what you came to learn.
Debrief within a day. Run an interpretation session with the team, dump observations onto notes, and build an affinity diagram while it's fresh.
🪄
Pro-tip: Bring a second person. One watches and probes, one captures quotes and timestamps. Solo, you'll miss half of what happens because your head is down taking notes.
Challenges & limitations
Time and access: Sessions run about 2 hours each, plus travel and scheduling. Some environments, like hospitals or secure sites, are hard to get into at all.
Observer effect: People act differently when watched, so the first stretch of a session is often performance, not routine.
Small samples: You get depth from a handful of users. It answers how and why, not how many.
