User Interview

Purpose

Conduct structured one-on-one conversations with users to explore motivations, experiences, pain points, and unmet needs in depth that quantitative methods and think-aloud cannot reach. User interviews answer why questions: why users value certain features, why they churned, why they work around problems instead of reporting them, and why their stated preferences differ from observed behaviour.

In A3, user interviews serve phase-specific purposes:

• A3.2: Post-testing interviews explore unexpected behaviours observed during task-based testing—“You completed the task but hesitated for 20 seconds. What was happening?”
• A3.3: Pilot interviews explore sustained usage patterns, churn reasons, and value perception after weeks of real-world use—“You stopped using the product in Week~3. Walk me through what happened.”

Interview Types in A3

When to Use

Use user interviews when:

• Quantitative data reveals what but not why—e.g. cohort analysis (the referenced method) shows Week~3 churn spike but not why users leave
• Think-aloud (the referenced method) surfaced unexpected behaviours that need deeper exploration
• NPS (the referenced method) Detractors need investigation—what would change their score?
• A3.3 pilot users develop workarounds, feature requests, or usage patterns the team did not anticipate
• A3.4 evaluation needs qualitative evidence to complement quantitative metrics (user quotes, stories, context)

Do NOT use when:

• The question is answerable with analytics—“How many users completed onboarding?” does not need an interview
• Seeking statistical evidence—interviews provide depth, not representativeness (sample too small for generalisation)
• As validation theater—interviewing only enthusiastic users to “prove” the product works
• Users have insufficient experience—interviewing someone who used the product for 5 minutes yields surface impressions, not deep insight

Sample Size and Duration

Per interview type:

• Post-test debrief (A3.2): 8–15 interviews, 15–20 minutes each
• Pilot check-in (A3.3): 10–20 interviews, 30–45 minutes each
• Churn interview (A3.3): 5–10 interviews, 20–30 minutes each
• Power-user deep-dive (A3.3): 5–8 interviews, 45–60 minutes each

Saturation: New themes typically stop emerging after 8–12 interviews for a given user type. If interview #13 yields no new themes, saturation is reached.

Analysis time: 1.5–2× interview duration per recording (a 45-minute interview requires 60–90 minutes to code)

Prerequisites

• Interview guide: Semi-structured guide with 8–12 open-ended questions, organised by theme, with follow-up probes
• Participant selection: Purposive sampling —select for diversity of experience (satisfied and dissatisfied, active and churned, novice and power user), not random sampling
• Prior data review: Interviewer has reviewed the participant's usage data, test results, or NPS score before the session (enables targeted probing)
• Recording setup: Audio (minimum) or video recording with participant consent
• Trained interviewer: Skilled in open-ended questioning, active listening, and avoiding leading questions
• Time: 30–45 minutes per interview; 60 minutes maximum (fatigue degrades quality)

Complete Procedure

Step~1: Design Interview Guide (2–4 hours)

Structure the guide in four phases:

1. Context (5 min): Understand the user's world. “Tell me about your role. Walk me through a typical day.” Establishes rapport and context before product discussion.
2. Experience (15–20 min): Explore product interaction. “Walk me through the last time you used [product]. What were you trying to accomplish? What happened?” Focus on specific past events, not general opinions.
3. Pain points and value (10–15 min): Probe friction and delight. “What's the most frustrating part? What would you miss most if we took it away? How does this compare to what you used before?”
4. Future (5 min): Explore unmet needs. “If you could change one thing, what would it be? What's missing?” Treat responses as hypotheses, not requirements—users describe symptoms, not solutions.

Key question design principles:

Step~2: Recruit Participants (2–5 days)

A3.2 post-test: Select from testing participants —choose those who exhibited interesting behaviours (unexpected failures, creative workarounds, strong positive/negative reactions).

A3.3 pilot: Purposive sampling across the engagement spectrum:

• 3–5 high-engagement users (power users)
• 3–5 moderate-engagement users (typical)
• 3–5 low-engagement or churned users (critical insight source)

Step~3: Conduct Interviews (30–45 min each)

• Start recording (with consent)
• Follow the guide but adapt based on responses —the best insights come from unexpected paths
• Use the 5 Whys technique when a response is surface-level: “You said checkout was confusing. What specifically confused you? Why was that confusing? What did you expect instead?”
• Take light notes for immediate recall; rely on recording for detailed analysis
• End with: “Is there anything I haven't asked that you think I should know?”

Step~4: Analyse and Synthesise (3–6 hours per 10 interviews)

1. Review recordings: Highlight key quotes, stories, and observations
2. Code themes: Tag insights by category (value drivers, pain points, feature requests, workflow context, churn triggers)
3. Affinity map: Cluster related insights across participants using affinity mapping (the referenced method)
4. Identify patterns: Which themes recur across 3+ participants? Recurring themes are signals; one-off comments are noise (unless from a critical user segment)
5. Connect to quantitative: Link interview themes to metric patterns—e.g. “Churn interviews reveal that Week~3 drop-off correlates with onboarding complexity, consistent with cohort analysis showing 70% activation but 45% Week~4 retention”

Step~5: Report (2–3 hours)

Produce interview synthesis:

• Key themes with supporting quotes (3–5 themes)
• Participant journey maps (if appropriate)
• Actionable recommendations linked to themes
• Connection to quantitative data
• Verbatim quotes for A3.4 evaluation report (qualitative evidence)

Quality Criteria

1. Semi-structured guide: Prepared but flexible—not a rigid script, not an unstructured conversation
2. Open-ended questions: No yes/no or leading questions in the guide
3. Diverse participants: Satisfied and dissatisfied, active and churned, different segments
4. Recorded and coded: Interviews recorded; themes systematically coded (not cherry-picked quotes)
5. Connected to data: Interview themes linked to quantitative patterns (cohort retention, NPS, task completion)
6. Actionable output: Themes translated into specific design, product, or strategy recommendations

Theoretical Foundation

Seminal references:

• : The definitive practitioner guide to user interviewing. Established key principles: interviews reveal context (the user's world), not just opinions (what they think about your product). The interviewer's job is to understand the user's mental model, workflow, and environment—not to validate the team's assumptions.
• : Provided the methodological foundation for qualitative research interviews, including question design, rapport building, and analysis techniques. Established the distinction between structured (fixed questions), semi-structured (guide with flexibility), and unstructured (conversation-led) formats.

Contemporary references:

• : Positioned user interviews within the continuous discovery framework: weekly interviews as a product team habit, not a phase-gated research activity. While A3 uses interviews at specific points (post-testing, during pilot), the continuous discovery mindset informs the approach—interviews are conversations, not interrogations.
• : Introduced the “Mom Test” principle: questions that even your mother cannot lie to you about. Ask about past behaviour (“When did you last…”), not future intent (“Would you use…”). Past behaviour predicts future behaviour; stated intent does not.

Challenges and Solutions

Challenge 1: Leading Questions

Symptoms: Interviewer asks “Don't you think the new checkout is much faster?” User agrees (social desirability). Team concludes checkout is validated.

Solutions: Use the Mom Test principle: ask about past behaviour, not opinions about your product. Pilot the interview guide with a colleague who challenges every question for leading language. Record and review early interviews to catch interviewer bias.

Challenge 2: Only Interviewing Happy Users

Symptoms: 10 interviews, all with active Promoters. Glowing feedback. Team misses the 40% who churned.

Solutions: Mandate that ≥30% of interview participants are churned, inactive, or Detractors. These users hold the most actionable insight. Churn interviews are uncomfortable but essential—they reveal whether problems are fixable (UX friction) or fundamental (wrong value proposition).

Challenge 3: Interview Data as Anecdote

Symptoms: One user says “I'd pay $50/month for this.” Team uses this as pricing evidence.

Solutions: Interview data is qualitative —it generates hypotheses, not conclusions. A single user's willingness to pay is an anecdote; 50 users' conversion rates are evidence. Report interview findings as themes with frequency (“7 of 12 participants mentioned…”), not as individual claims.

Relationship to Other Methods

User Interviews receive input from:

• Think-Aloud (the referenced method): Observed behaviours and verbalisations generate interview topics
• NPS (the referenced method Detractor/Promoter classification identifies who to interview and what to explore
• Cohort Analysis (the referenced method): Retention patterns identify churn triggers to investigate
• Usage Analytics: Feature usage data reveals behaviours to probe (“I noticed you use feature X daily but never feature Y…”)