Four AI prompts that cover the full cycle of a contextual inquiry study: preparing the observation protocol, synthesizing field notes, building work models, and reporting findings. Copy the prompt, fill in the bracketed fields with your project details, and paste into any LLM.
Observation protocol preparation
Use this prompt to prepare a structured yet flexible protocol before going into the field.
You are a senior UX researcher preparing for a contextual inquiry study.
Context:
- Product/domain: [describe the product or work domain]
- Target users: [who you will observe, their role]
- Research objective: [what you want to understand about their work]
- Session length: 2 hours
Create a contextual inquiry observation protocol:
1. Introduction script (how to explain the session to the participant, apprentice-master framing)
2. 3-5 focus areas to guide observation (specific aspects of work to watch for)
3. A list of 8-10 "trigger questions" to use during natural pauses (retrospective probing questions tied to observable behaviors)
4. An artifact checklist: physical and digital objects to photograph or document
5. A wrap-up checklist for collaborative interpretation (3-5 interpretation statements to validate with the participant)
Rules:
- Questions should be grounded in observable behavior, not hypothetical
- Include prompts for unexpected moments ("I noticed you did X — tell me about that")
- Note when to stay silent and when to probeField notes synthesis
Use this prompt after completing all sessions to extract themes and insights from your observations.
I conducted [N] contextual inquiry sessions observing [role] performing [task] in their [environment].
Below are my field notes from all sessions:
[Paste field notes]
Please synthesize this data:
1. Create an affinity diagram: group individual observations into 6-10 themes
2. For each theme, provide: a one-sentence insight, 2-3 supporting observations with participant identifiers, and a design implication
3. Identify 3-5 workarounds or shadow processes users have created
4. Map the typical task flow: list the steps users actually follow (not the official process) with variations noted
5. Flag environmental factors that influence behavior (physical space, tools, social dynamics)
6. List contradictions between what users said they do and what you observed them doing
Format: theme name → insight → evidence → design implication.Work model creation
Use this prompt to generate structured work models from your observations.
Based on the following contextual inquiry observations, create work models for [task/domain].
[Paste synthesized observations]
Generate:
1. **Flow model**: How does work move between people? Who talks to whom, what information passes between them, what are the breakdowns?
2. **Sequence model**: What are the actual steps users take to complete [task]? Include decision points, branching paths, and common error-recovery steps.
3. **Artifact model**: What physical and digital artifacts do users create, reference, or modify? What information does each artifact carry?
4. **Physical model**: How does the layout of the workspace affect the work? What is within reach, what requires movement?
For each model, note the design opportunities: where could a product reduce friction, eliminate workarounds, or better support the actual work flow?Research report for stakeholders
Use this prompt to turn analyzed findings into a stakeholder-ready summary.
Based on the following contextual inquiry findings, write a research report for stakeholders.
[Paste your insights and work models]
Structure:
1. Executive summary (3-4 sentences: what we studied, where, key finding, main recommendation)
2. Key insights (bulleted, each with: insight statement + observed evidence + design implication)
3. Work practice highlights: 2-3 most surprising discoveries about how users actually work
4. Recommendations (prioritized: what to build, what to change, what to investigate further)
5. Suggested next steps
Tone: direct, evidence-based, grounded in what was observed. No speculation beyond the data.