Interview Plan

Use this when you need to plan a user interview or series of interviews and want a structured guide, a capture template for consistent note-taking, and a process for extracting insights afterward. Covers the full arc from research objectives through synthesis.

After interviews: Use /research-synthesize to turn your interview notes into themes, backlog impact, and planning inbox entries. For concept validation (testing an idea rather than exploring a problem), use /concept-test-plan instead. Pair with /screener-design for participant recruitment.

Reference materials

Detailed guidance on specific aspects of interviewing:

• Question quality — question categories (behavior, opinion, feeling, knowledge, sensory, demographic), leading vs. non-leading rewrites, probing techniques, and interviewing dos and don'ts.
• Rainbow chart — visual cross-interview analysis technique using color-coded participant mapping. Use after 3+ interviews to spot convergence and divergence patterns.
• Finding participants — where to recruit, screening questionnaires, scheduling logistics, compensation guidelines, and a recruitment message template.

Process

Step 1: Separate the three question levels

Before writing a single interview question, force clarity on three distinct levels. These are not the same thing and must be articulated separately:

1. Business question -- What does the organization need to decide? (e.g., "Should we invest in a self-serve onboarding flow or keep the white-glove model?") This is the decision the research informs, not a question you ask participants.
2. Research objectives -- What do we need to learn to inform that decision? (e.g., "Understand where new users get stuck in the first 10 minutes" or "Identify which onboarding tasks users can do independently vs. need help with.") These are learning goals, not interview questions.
3. Interview questions -- What do we actually ask participants? (e.g., "Walk me through the last time you set up a new tool at work. What happened?") These are the words that come out of your mouth during the session.

If the business question and the research objective are identical, push back -- the team is asking participants to solve their business problem. If the research objective and interview questions are identical, the questions are too abstract to yield useful data.

Step 2: Gather research context

Ask the user:

1. Business question -- What organizational decision will this research inform? (The decision, not the learning goal.)
2. Research objectives -- What do you need to learn to make that decision? (These should be distinct from the business question.)
3. Target user segment -- Who are you interviewing? (Role, experience level, relationship to the product.)
4. Hypotheses or assumptions -- What does the team currently believe? What are you testing? (e.g., "We assume users find the export flow confusing.")
5. Number of interviews planned -- How many participants? (Affects synthesis approach. See finding participants reference for recruitment guidance and sample size recommendations.)
6. Time per interview -- How long is each session? (Typical: 30-60 minutes.)
7. Existing context -- Any prior research, personas, or known pain points to build on?
8. Observation opportunity -- Is there a chance to observe participants in their natural environment? If the research is about behavior (what people do) rather than attitudes (what people think), an observational study or contextual inquiry may yield stronger data than interviews alone. Consider whether observation should supplement or replace some interview sessions.

Step 3: Identify which personas are being researched

Before generating the guide, prompt the user:

Persona check: Which user persona(s) does this research target? If you have defined personas (e.g., from /persona-create or /persona-draft), name them. If not, describe the archetype briefly — their role, goals, and what makes them distinct from other user types.

This grounds your interview guide in real user context and ensures your questions probe persona-specific behaviors, not generic ones.

Step 4: Generate the interview guide

Produce a structured interview guide:

## Interview Guide — (Topic)

**Research objectives:** (From Step 1)
**Target persona(s):** (From Step 2)
**Session length:** (N minutes)
**Date range:** (When interviews are planned)

---

### Warm-Up (3-5 minutes)

(2-3 rapport-building questions. Purpose: make the participant comfortable and establish context about their role/background.)

- (Question 1 — about their role and daily work)
- (Question 2 — about their general experience with the problem space)

### Core Questions (15-30 minutes)

(5-8 open-ended questions directly tied to research objectives. Mix question types across categories — behavior, opinion, feeling, knowledge, sensory — to build a complete picture. See [question quality reference](reference/question-quality.md). Each question includes:)

**Q1: (Open-ended question tied to Objective 1)**
- Why this matters: (What this question tests or reveals)
- Probes if they give a short answer:
  - "(Follow-up probe 1)"
  - "(Follow-up probe 2)"
  - "Can you walk me through a specific example?"

**Q2: (Open-ended question tied to Objective 1 or 2)**
- Why this matters: (Explanation)
- Probes:
  - "(Probe)"
  - "(Probe)"

(Continue for each core question)

### Hypothesis-Testing Questions (5-10 minutes)

(2-3 questions designed to validate or challenge the team's assumptions. Frame these neutrally — don't lead.)

**H1: (Question testing Hypothesis 1)**
- Hypothesis being tested: (What you believe)
- Neutral framing: (Why this question avoids confirmation bias)
- Probes:
  - "(Probe)"

### Wrap-Up (3-5 minutes)

- "Is there anything I didn't ask about that you think is important?"
- "If you could change one thing about (the product/process/workflow), what would it be?"
- "Is there anyone else you'd recommend I talk to?"

---

### Interviewer Reminders

- **Listen more than you talk.** Aim for 80/20 — participant talks 80% of the time.
- **Follow the energy.** If the participant gets animated about something, follow that thread even if it's off-script.
- **Don't fix or explain.** If they describe frustration with your product, resist the urge to explain why it works that way.
- **Silence is a tool.** After they answer, wait 3-5 seconds. They often add the most valuable insight in the pause.
- **Note exact words.** Their language matters more than your paraphrase.

Step 5: Generate the capture template

Produce a note-taking template for use during or immediately after each interview:

## Interview Capture — (Participant ID or Name)

**Date:** (Date)
**Interviewer:** (Name)
**Participant:** (ID/name, role, segment)
**Persona match:** (Which defined persona does this participant map to?)
**Session length:** (Actual duration)

---

### Key Quotes
(Capture verbatim quotes — these are gold for synthesis and stakeholder communication.)

1. "(Exact quote)" — Context: (what prompted this)
2. "(Exact quote)" — Context: (what prompted this)
3. "(Exact quote)" — Context: (what prompted this)

### Observations per Research Objective

**Objective 1: (Restate objective)**
- What they said: (Summary of relevant responses)
- What they did: (Any behavioral observations — hesitation, confusion, excitement)
- Strength of signal: Strong / Moderate / Weak

**Objective 2: (Restate objective)**
- What they said: (Summary)
- What they did: (Observations)
- Strength of signal: Strong / Moderate / Weak

### Hypothesis Validation

| Hypothesis | Supported? | Evidence |
|---|---|---|
| (Hypothesis 1) | Yes / No / Mixed | (Brief evidence) |
| (Hypothesis 2) | Yes / No / Mixed | (Brief evidence) |

### Surprises
(Things the participant said or did that were unexpected — not directly related to your questions.)

- (Surprise 1)
- (Surprise 2)

### Emotional Moments
(Points where the participant showed frustration, excitement, confusion, or strong feeling. These often reveal the highest-value insights.)

- (Moment — what triggered it, what they said)

### Follow-Up Needed
- (Any topics that need deeper exploration in future interviews)
- (Any artifacts the participant mentioned that you should request)

### Interviewer Reflection (fill within 30 minutes of session)
- What was the single most important thing I learned?
- What surprised me most?
- What should I ask differently next time?

Step 6: Generate the synthesis plan

Produce a plan for how to aggregate findings across multiple interviews:

## Cross-Interview Synthesis Plan

**Total interviews planned:** (N)
**Synthesis approach:** (Recommended based on interview count)

### After Each Interview (within 30 minutes)
1. Complete the capture template while memory is fresh.
2. Highlight the top 3 moments (strongest quotes, biggest surprises, clearest signals).
3. Rate signal strength per research objective.

### After All Interviews Are Complete
1. **Build a rainbow chart** — Assign each participant a color and map their responses onto a shared grid by question/theme. See [rainbow chart reference](reference/rainbow-chart.md) for the technique and template.
2. **Aggregate by objective** — For each research objective, collect all evidence across participants. Look for patterns, convergence, and divergence.
3. **Surface themes** — Identify 3-5 recurring themes across participants (use `/research-synthesize` for this step).
4. **Quote bank** — Compile the strongest quotes per theme for stakeholder communication.
5. **Hypothesis scorecard** — Tally hypothesis support across all participants.
6. **Confidence assessment** — Be honest about sample size:
   - 3-5 interviews: Exploratory — themes worth investigating, not building on.
   - 6-10 interviews: Directional — patterns are likely real if consistent.
   - 10+ interviews: Strong — patterns are reliable for decision-making.
7. **Recommended actions** — What should change in the backlog or plan based on findings?

### Handoff to /research-synthesize
When you're ready to synthesize, gather:
- All completed capture templates
- Your interviewer reflections
- The original research objectives and hypotheses

Then invoke `/research-synthesize` with this data. The synthesis skill will produce themes, backlog impact, and a planning inbox entry.

Step 7: Review and validate

Ask the user:

• Do the core questions actually test your research objectives? (Map each question back to an objective.)
• Are the hypothesis-testing questions neutrally framed? (Would a participant feel led toward an answer?)
• Is the capture template practical for your interview format? (In-person vs. remote, solo vs. paired interviewing.)
• Does the synthesis plan match your timeline and team size?
• Are you testing the right hypotheses, or should some be reframed based on the guide review?

Adjust as needed.

Related skills

• /research-synthesize — use after interviews to turn notes into themes, backlog impact, and planning entries. This is the natural next step.
• /persona-draft — if you don't have personas yet, draft them from existing data before planning interviews.
• /persona-create — if you need shareable persona artifacts (PDF, slides, Miro) for the team.
• /experiment-design — if the interview findings suggest a hypothesis worth testing with an experiment.

Output location

Present the interview guide, capture template, and synthesis plan as formatted text in the conversation. The user copies the guide into their project documentation and uses the capture template for each session.