The UX Research Library
Contextual Inquiry Quick Guide
A practical companion — not the full method. Pick a section.
Tap a section to jump straight to it
What It Is
Watching and asking, at the same time, in the real place it happens.
The core idea
- You go to where the work or behavior actually happens — their desk, their kitchen, their actual workflow — instead of bringing them to you.
- You watch first, then ask about what you just saw, rather than asking someone to describe it from memory.
It's not just observation
- A pure fly-on-the-wall observation misses the "why" behind what you're seeingContextual inquiry blends watching with brief, in-the-moment questions — you're a coach or apprentice, not a silent camera.
Common misconception
- It's not the same as a usability test in someone's officeThere's usually no task you designed — you're watching their real, ongoing work, on their terms, not a scenario you scripted.
CI vs. Interviews
Same goal — understanding people — different tool for different situations.
Pick contextual inquiry when
- The environment shapes the behavior — physical layout, interruptions, other people, tools lying around.
- People genuinely can't describe their own workflow accurately — it's too automatic, or too fragmented, to narrate from memory.
- You need to catch workarounds nobody would think to mention.
Pick interviews when
- You need a fast answer and don't have days for site visitsCI is slower and more resource-intensive per person than an interview.
- The behavior happens entirely on a screen, and remote is genuinely fineIf nothing about the physical environment matters, you may not need to be there in person.
The common mixup
- Treating CI as "an interview, but in their office"If you're not actually watching them work and asking about what you see, it's just a relocated interview — you've lost the whole point of going.
Before You Go
What to lock down, and what to leave loose on purpose.
Lock this down
- A specific window of real activity to observe — not "come see how I work," but a task or period likely to actually happen while you're there.
- Consent and recording logistics, confirmed in advance — including whether other people who might appear (coworkers, family) are aware.
- A loose list of things you're curious about, so you're not starting from nothing.
Leave this loose
- Your exact questionsScripting too tightly fights the whole method — the best questions come from what you actually see, not what you guessed you'd see.
Don't do this
- Ask them to "act normal" for a task that wouldn't naturally happen that dayManufactured normalcy isn't real behavior — if the task genuinely isn't happening, that's useful information too, not a scheduling failure.
In the Field
What to actually do once you're standing there.
Do this
- Ask about what just happened, right after it happens — not saved up for later, when the detail and the reason for it have both faded.
- Notice workarounds, sticky notes, and improvised tools — those are often the most honest evidence of where the real design gap is.
- Let silence sit. People narrate more when you don't rush to fill the gap.
Use with care
- Taking notes while they workHeavy note-taking can make people perform for you instead of just working. A few key words in the moment, fuller notes right after, usually works better.
Don't do this
- Correct or teach them a "better way" while you're observingThe moment you help, you've changed the behavior you came to see — save any teaching for after the visit is fully over.
The one thing to remember
You're a guest in someone's real workflow — not a director staging a scene.