research methods
Contextual inquiry
What is contextual inquiry?
Contextual inquiry is interviewing people while they work, in their own environment, on their own real tasks - watching what they actually do and asking about it as it happens. You become the apprentice and they the master, so instead of a tidy account recalled at a desk, you get the messy reality: the workarounds, the interruptions, the context that no interview in a meeting room would ever surface.
Also known as: contextual inquiry, contextual interview, master-apprentice
Prefer to watch?
Watch the recap 0:51
The demo
“How do you handle invoicing?” Ask it across a desk, then go and watch it happen where the work really is - and see how much the room you're in changes the answer.
What this demo shows (text version)
The same question - how do you handle invoicing? - is answered two ways. In the "lab interview", recalled at a desk, you get a tidy, idealised account: "I log the job, then I invoice it." In "their world", watching the work happen on site, you see the reality: poor signal, so they photograph receipts and do it later from the van, one-handed, with the phone propped on the dashboard - constraints and workarounds the interview never mentioned.
That's contextual inquiry: interviewing while observing real work in its real setting, with you as the apprentice and the participant as the master. The environment is data, and being there surfaces the tacit, habitual, worked-around reality that recalled answers smooth over. It costs more than a meeting-room chat, but when context matters, it's the only way to see the task as it actually is.
Go to where the work happens and watch it being done. Contextual inquiry combines observation and interview in the real setting, so you see the things people never think to mention - the sticky notes round the monitor, the phone propped against the till, the "oh, I always have to do this bit twice". The environment and the real task carry information that a recalled, out-of-context answer simply loses.
Ask in a meeting room and you get the official version: "I log the job, then invoice it." Watch it happen on site and you see the truth - the signal drops, so they photograph everything and do it later from the van, one-handed, balancing the phone on the dashboard. The context held the real story, and only being there revealed it.
The defining move is studying real work in its real place. People are poor narrators of their own routines - they skip the habitual, forget the workarounds, and smooth over the friction - but when you watch them do the actual task in their actual environment, all of that becomes visible. The setting itself is data: the clutter, the tools, the interruptions, the other people.
The "master-apprentice" stance keeps it honest. You position yourself as the learner and the participant as the expert demonstrating their craft, asking "why did you do that?" in the moment rather than steering. This surfaces tacit knowledge - the things people do without being able to explain - and resists the observer-expectancy trap of nudging them toward your hypothesis.
It's more involved than a lab interview - travel, time, messiness, and you can't fully control the conditions - but that mess is the point. Use it when context matters (specialist tools, physical environments, real constraints), and pair the watching with the asking: observation shows what, the in-situ questions reveal why.