Contextual Inquiry Terms
Contextual Inquiry Terms refers to the vocabulary used in contextual inquiry - a UX research method where you observe and interview users in their real environment while they perform real tasks, rather than in a lab with synthetic prompts. The terminology overlaps with broader user research but has its own specific working language.
Useful primer if you’ve inherited a research project, are commissioning one for the first time, or need to understand a UX team’s findings without re-learning the field from scratch.
The core terms
Master-apprentice model. The framing for the interview itself: the participant is the master who knows their work, and you’re the apprentice trying to learn it. Reframes the dynamic away from “researcher tests user” toward “user teaches researcher.”
Affinity diagram. The synthesis output. Hundreds of observations from sessions get written on sticky notes and grouped bottom-up by similarity. The patterns surface from the grouping rather than being imposed by a pre-existing framework.
Work models. Visual representations of how work actually flows - sequence model (steps in order), flow model (people, tools, communications), cultural model (norms and pressures), physical model (the environment), artefact model (the documents and tools used). A full contextual inquiry produces several of these per participant.
Triggers and breakdowns. Specific moments in the observed work where something prompts action (trigger) or where the existing tool or process fails the user (breakdown). Most design opportunities sit in the breakdowns.
Workarounds. The improvised systems users build when the official tool doesn’t fit their work. A spreadsheet duct-taped to a CRM. A second browser tab kept open because the first hides important info. Workarounds are gold - they’re a user telling you what their work actually needs.
How contextual inquiry differs from interviewing
A standard interview asks people to describe what they do. A contextual inquiry watches them do it. The gap between the two is usually substantial - people forget steps, omit workarounds they no longer notice, and rationalise behaviour that was actually random.
The other shift: questions are interleaved with observation. You don’t run a 60-minute interview then a separate 60-minute observation. You observe a step, ask about it in the moment, observe the next step, ask about that one. The data is richer because both pieces inform each other.
An example
A B2B SaaS product team running contextual inquiries with five customer ops users observed each one for 3 hours during a typical workday. The official product workflow they had documented took 11 steps. The actual observed workflow averaged 19 steps because users had built workarounds the team had never seen - keeping a separate spreadsheet, using a sister product to do bulk operations the main product made tedious, copying data between tabs.
The product roadmap before the inquiry had 14 features queued for the next quarter. After the inquiry, four of those got cut because the underlying need was being solved by the workarounds users had already built. Three new features got prioritised that nobody had thought to ask for. The roadmap was smaller and more focused - and the eventual NPS lift was bigger than any single feature would have produced.
Related terms
- Audience - the broader concept contextual inquiry targets understanding of
- Buyer Persona - the synthesis output that contextual inquiry can ground in real behaviour
- Audience Segmentation - the analytical step contextual inquiry findings often refine
- Customer Journey - the broader workflow contextual inquiry maps in detail
- Accessibility Terms - an adjacent vocabulary set every UX researcher needs