---
# === IDENTITY ===
id: business/customer-research/customer-interview-guide-template/2026
canonical_question: "What questions should I ask in customer discovery interviews — by stage, with anti-patterns for leading questions?"
aliases:
  - "customer discovery interview questions by stage"
  - "the mom test interview guide template"
  - "customer interview script for startups"
entity_type: execution_recipe
domain: business > customer-research > customer interview guide template
region: global
jurisdiction: global
temporal_scope: 2024-2026

# === VERIFICATION ===
last_verified: 2026-03-11
confidence: 0.91
version: 1.0
first_published: 2026-03-11

# === TEMPORAL VALIDITY ===
temporal_validity:
  status: evolving
  last_breaking_change: null
  next_review: 2026-09-07
  change_sensitivity: medium

# === CONSTRAINTS ===
constraints:
  - "Never ask hypothetical questions ('Would you use X?') — they produce false validation 80%+ of the time"
  - "Interview guide must be adapted per discovery stage — problem interviews differ from solution interviews differ from usability interviews"
  - "Minimum 5 interviews per persona segment before drawing conclusions — small samples produce misleading patterns"
  - "Interviewer must not mention their solution until the final 5 minutes — mentioning early contaminates all subsequent answers"

# === SKIP CONDITIONS ===
skip_this_unit_if:
  - condition: "User specifically wants to uncover purchase motivations using JTBD methodology"
    use_instead: "business/customer-research/jobs-to-be-done-interview-methodology/2026"
  - condition: "User wants to validate a specific feature, not discover problems"
    use_instead: "Search knowledgelib.io for usability testing methodology — no dedicated unit yet"

# === AGENT HINTS ===
inputs_needed:
  - key: discovery_stage
    question: "What stage of discovery are you in?"
    type: choice
    options: ["Problem discovery (pre-product)", "Solution validation (have MVP)", "Post-launch (have customers)", "Win/loss analysis"]
  - key: interview_type
    question: "What type of interview?"
    type: choice
    options: ["B2B (decision-maker)", "B2B (end-user)", "B2C consumer", "Expert/industry"]
  - key: interview_experience
    question: "How experienced is the interviewer?"
    type: choice
    options: ["First time", "Some experience (5-20 interviews)", "Experienced (20+ interviews)"]

# === EXECUTION METADATA ===
execution:
  required_inputs:
    - name: "Target persona or segment description"
      source: "business/customer-research/buyer-persona-development-methodology/2026"
      format: "Persona profile or segment hypothesis"
    - name: "Key assumptions to test"
      source: "Founder / product team"
      format: "List of 3-5 testable hypotheses"
  outputs:
    - name: "Stage-Specific Interview Guide"
      format: "Markdown document"
      description: "Complete interview script with opening, questions by section, probing follow-ups, anti-pattern warnings, and closing"
    - name: "Interview Analysis Template"
      format: "Spreadsheet"
      description: "Structured template for capturing insights across multiple interviews with pattern detection"
  tools_required:
    - name: "Zoom / Google Meet"
      purpose: "Video call interviews with recording"
      tier: free
      cost: "$0 (40-min limit on Zoom free) or $13.33/mo"
      alternatives: ["Google Meet (free, unlimited)", "Phone calls"]
    - name: "Otter.ai"
      purpose: "Real-time transcription during interviews"
      tier: free
      cost: "$0 (300 min/mo) or $8.33/mo"
      alternatives: ["Zoom transcription (free)", "manual notes"]
    - name: "Google Sheets"
      purpose: "Interview tracking and pattern analysis"
      tier: free
      cost: "$0"
      alternatives: ["Airtable", "Notion"]
  credentials_needed:
    - service: "Zoom"
      type: "Account login"
      where_to_get: "https://zoom.us/signup"
      free_tier_limits: "40-minute meetings, 100 participants"
  estimated_duration: "2-3 hours to build guide; 30-60 minutes per interview; 5-15 interviews per persona"
  estimated_cost: "$0 (DIY) to $2,000 (recruited participants at $50-200 each)"

# === DISTRIBUTION ===
canonical_source: "https://knowledgelib.io/business/customer-research/customer-interview-guide-template/2026"
suggested_citation: "Source: knowledgelib.io — AI Knowledge Library (verified 2026-03-11)"

# === RELATED UNITS ===
related_kos:
  depends_on:
    - id: business/customer-research/buyer-persona-development-methodology/2026
      label: "Persona profiles to target for interviews"
  feeds_into:
    - id: business/customer-research/ideal-customer-profile-framework/2026
      label: "Interview insights inform ICP criteria"
    - id: business/customer-research/buyer-journey-mapping/2026
      label: "Interview data maps buyer touchpoints"
  related_to:
    - id: business/customer-research/jobs-to-be-done-interview-methodology/2026
      label: "JTBD interviews for deeper motivation analysis"
  alternative_to: []

# === SOURCES ===
sources:
  - id: src1
    title: "The Mom Test: How to Talk to Customers"
    author: Rob Fitzpatrick
    url: https://www.momtestbook.com/
    type: industry_report
    published: 2013-10-01
    reliability: authoritative
  - id: src2
    title: "Customer Discovery Interviews — The Real Startup Book"
    author: Kromatic
    url: https://kromatic.com/real-startup-book/3-generative-market-research/customer-discovery-interviews
    type: community_resource
    published: 2024-06-15
    reliability: high
  - id: src3
    title: "The Mom Test for Better Customer Interviews"
    author: Looppanel
    url: https://www.looppanel.com/blog/customer-interviews
    type: technical_blog
    published: 2025-01-10
    reliability: high
  - id: src4
    title: "8 Essential Customer Discovery Interview Questions for 2025"
    author: ProblemSifter
    url: https://www.problemsifter.com/blog/customer-discovery-interview-questions
    type: technical_blog
    published: 2025-02-20
    reliability: high
  - id: src5
    title: "The Ultimate Guide to Customer Discovery"
    author: CardBoard
    url: https://cardboardit.com/2023/08/the-ultimate-guide-to-customer-discovery-lessons-from-airbnb-steve-blank-and-the-mom-test/
    type: technical_blog
    published: 2023-08-15
    reliability: high
  - id: src6
    title: "Customer Discovery Done Well"
    author: Richard von Kaufmann
    url: https://richievk.medium.com/customer-discovery-done-well-179a05c4726f
    type: technical_blog
    published: 2024-04-10
    reliability: moderate
---

# Customer Interview Guide Template

## Purpose

This recipe produces stage-specific interview guides (problem discovery, solution validation, post-launch, win/loss) with exact question scripts, probing follow-ups, anti-pattern warnings for each question, and a structured analysis template for synthesizing insights across 5-15 interviews. The output follows The Mom Test methodology — asking about real past behavior rather than hypothetical futures — to produce actionable insights that survive contact with reality.

## Prerequisites

- [ ] **Target persona or segment description** — from [Buyer Persona Development](/business/customer-research/buyer-persona-development-methodology/2026) or a documented hypothesis
- [ ] **3-5 testable hypotheses** — specific assumptions about customer problems, behaviors, or preferences
- [ ] **Interview scheduling** — ability to book 30-60 minute calls with target segment
- [ ] **Recording + transcription tool** — [Zoom](https://zoom.us) + [Otter.ai](https://otter.ai) or equivalent
- [ ] **Analysis spreadsheet** — Google Sheets template for cross-interview pattern detection

## Constraints

- Never ask hypothetical questions ("Would you use X?") — they produce false validation 80%+ of the time [src1]
- Interview guide must be adapted per discovery stage — problem interviews differ fundamentally from solution interviews [src2]
- Minimum 5 interviews per persona segment before drawing conclusions [src5]
- Interviewer must not mention their solution until the final 5 minutes [src1]

## Tool Selection Decision

```
Which path?
├── Problem Discovery (no product yet)
│   └── PATH A: Problem Interview Guide — understand if the problem exists and matters
├── Solution Validation (have MVP or prototype)
│   └── PATH B: Solution Interview Guide — test if your approach solves the validated problem
├── Post-Launch (have paying customers)
│   └── PATH C: Customer Feedback Guide — understand satisfaction, expansion, and churn risk
└── Win/Loss Analysis (have both wins and losses)
    └── PATH D: Win/Loss Interview Guide — understand why deals close or die
```

| Path | Focus | Duration | Key Output |
|------|-------|----------|-----------|
| A: Problem Discovery | Pain validation, current solutions, willingness to pay | 45-60 min | Problem severity ranking + behavior patterns |
| B: Solution Validation | Solution fit, usability, pricing reactions | 30-45 min | Feature priority list + pricing sensitivity |
| C: Customer Feedback | Satisfaction, expansion, referral potential | 30 min | NPS drivers + churn risk factors |
| D: Win/Loss | Decision process, competitor comparison, deal dynamics | 45 min | Win/loss patterns + competitive intelligence |

## Execution Flow

### Step 1: Build the Interview Guide

**Duration**: 1-2 hours
**Tool**: Document editor

Select the appropriate guide template based on your discovery stage. Each template includes question-level anti-pattern warnings.

**PATH A: Problem Discovery Interview Guide**

```markdown
## Problem Discovery Interview Script

### Opening (3 minutes)
SCRIPT: "Thank you for making time. I'm researching how [target persona]
handles [domain]. There are no right or wrong answers — I'm just trying
to learn from your experience. Is it OK if I record for my notes?"

ANTI-PATTERN: Do NOT say "I'm building a product that..." — this primes
the interviewee to give polite feedback instead of honest answers.

### Section 1: Context & Current Behavior (10 minutes)
Q1: "Tell me about your role and what a typical week looks like."
   → FOLLOW-UP: "What takes up most of your time?"
   → ANTI-PATTERN: Don't ask "Are you busy?" (everyone says yes)

Q2: "Walk me through the last time you [dealt with problem area]."
   → FOLLOW-UP: "What happened next? How did you handle it?"
   → ANTI-PATTERN: Don't ask "Do you have a problem with X?" (leading)

Q3: "What tools or processes do you currently use for [domain]?"
   → FOLLOW-UP: "How did you end up choosing those? What do you like/dislike?"
   → ANTI-PATTERN: Don't ask "Have you tried [your competitor]?" (fishing)

### Section 2: Pain Point Deep-Dive (15 minutes)
Q4: "What's the most frustrating part of [current process]?"
   → FOLLOW-UP: "Can you give me a specific recent example?"
   → FOLLOW-UP: "How often does this happen? Daily? Weekly? Monthly?"
   → ANTI-PATTERN: Don't ask "Wouldn't it be great if...?" (leading)

Q5: "What happens when [problem] occurs? What's the consequence?"
   → FOLLOW-UP: "How does that affect your team/revenue/time?"
   → ANTI-PATTERN: Don't suggest the consequence — let them name it

Q6: "Have you tried to solve this? What did you try?"
   → FOLLOW-UP: "What worked? What didn't? Why did you stop?"
   → ANTI-PATTERN: Don't ask "Why didn't you try X?" (judgmental)

Q7: "If this problem went away tomorrow, what would change for you?"
   → FOLLOW-UP: "How would you measure the improvement?"
   → ANTI-PATTERN: Don't describe your solution as the answer

### Section 3: Willingness to Pay / Priority (10 minutes)
Q8: "How much time/money do you currently spend dealing with this?"
   → FOLLOW-UP: "Is that your estimate or a tracked number?"
   → ANTI-PATTERN: Don't ask "Would you pay $X?" (hypothetical)

Q9: "If you ranked your top 3 work challenges, where would this fall?"
   → FOLLOW-UP: "What's above it? Why?"
   → ANTI-PATTERN: Don't ask "Is this your biggest problem?" (binary, leading)

Q10: "Have you actively looked for a better solution in the past year?"
   → FOLLOW-UP: "What did you search for? What did you find? Why didn't you buy?"
   → ANTI-PATTERN: Don't ask "Would you look for a solution?" (hypothetical)

### Section 4: Buying Context (5 minutes)
Q11: "If you found a great solution, what would the buying process look like?"
   → FOLLOW-UP: "Who else would need to approve? What's the budget process?"
   → ANTI-PATTERN: Don't ask "Would you buy this?" (hypothetical)

Q12: "What would make you NOT buy a solution in this space?"
   → FOLLOW-UP: "What are the dealbreakers?"

### Closing (2 minutes)
SCRIPT: "This has been incredibly helpful. [Optional: brief mention of what
you're working on, max 30 seconds.] Would you be open to seeing something
we're building in the future?"

RECORD: Contact info for follow-up, permission level granted.
```

**Verify**: Guide has 10-15 questions, each with follow-up probes and anti-pattern warnings.
**If failed**: If guide feels too long, cut to 8 questions — better to go deep on fewer questions than shallow on many.

### Step 2: Recruit Interviewees

**Duration**: 3-7 days
**Tool**: Email, LinkedIn, personal network

Target 5-15 interviews per persona segment. Quality of interviewee matters more than quantity.

```markdown
## Recruitment Methods (by effectiveness)

1. BEST: Warm introductions from your network (highest show rate, most honest)
2. GOOD: Direct outreach to target persona on LinkedIn (response rate 5-15%)
3. ACCEPTABLE: Paid recruitment via Respondent.io ($100-200/interview)
4. LAST RESORT: Cold email/survey with interview opt-in (response rate 1-3%)

## Recruitment Message Template
"Hi [Name], I'm researching how [role] handles [domain]. I'm not selling
anything — I'm doing customer discovery for [brief context]. Would you
have 30 minutes for a video call this week? I'd love to learn from your
experience. [Optional: I'm happy to share the research findings as a thank-you.]"

## Target Recruitment Mix
- 40% current customers (if you have them)
- 30% target persona who are NOT customers
- 20% people who evaluated but didn't buy
- 10% industry experts or adjacent role holders
```

**Verify**: 5+ interviews scheduled within 2 weeks.
**If failed**: If recruitment stalls, offer $50 Amazon gift cards (standard for B2B) or reduce interview length to 20 minutes.

### Step 3: Conduct Interviews

**Duration**: 30-60 minutes each, spread over 1-3 weeks
**Tool**: Zoom + Otter.ai

Execute interviews using the prepared guide. Follow these rules during the interview:

```markdown
## Interview Execution Rules

1. RECORD everything (with permission). Memory is unreliable.
2. Shut up and listen. Target 80% them, 20% you.
3. Ask "Can you give me a specific example?" at least 3 times per interview.
4. When they say something interesting, ask "Tell me more about that."
5. Never pitch. Never defend. Never explain your product.
6. Take real-time notes on energy shifts — when do they lean in? When bored?
7. Note exact quotes — their words are more powerful than your summaries.
8. At the end, ask "Is there anything I should have asked but didn't?"

## Red Flags During Interviews (The Mom Test signals)
- COMPLIMENTS: "That sounds like a great idea!" → Worthless. They're being polite.
- FLUFF: "I would definitely use that." → Hypothetical. Ask what they DO now.
- WISHLISTS: "It would be cool if it could also..." → Feature requests ≠ buying intent.

## Green Flags During Interviews
- SPECIFICS: "Last Tuesday, I spent 3 hours on..." → Real behavior. Gold.
- EMOTION: "That drives me absolutely crazy." → High-severity pain point.
- ACTION: "I built a spreadsheet to work around it." → They invested effort. Real pain.
- MONEY: "We're paying $X/month for a terrible solution." → Budget exists.
```

**Verify**: Each interview produces 3+ specific behavior examples and 2+ quotable insights.
**If failed**: If interviews are surface-level, add more "Tell me about the last time..." questions and reduce yes/no questions.

### Step 4: Analyze and Synthesize

**Duration**: 3-6 hours (after all interviews complete)
**Tool**: Google Sheets + Miro

Build a cross-interview analysis that identifies patterns, not just individual stories.

```markdown
## Interview Analysis Spreadsheet Structure

### Sheet 1: Raw Data
| Interview # | Date | Name | Role | Company Size | Q4 Response (Pain) | Q5 Response (Impact) | Q8 Response (Cost) | Q10 Response (Active Search) | Key Quotes | Pain Severity (1-5) | Buying Signal (1-5) |

### Sheet 2: Pattern Analysis
| Pattern | Interview Evidence | Frequency (n/total) | Confidence |
|---------|-------------------|---------------------|------------|
| [Pain Point A] | I3, I5, I7, I9 | 4/10 (40%) | Moderate |
| [Pain Point B] | I1-I10 | 9/10 (90%) | High |
| [Current Workaround X] | I2, I4, I6, I8 | 4/10 (40%) | Moderate |
| [Budget Signal] | I1, I3, I5 | 3/10 (30%) | Low |

### Sheet 3: Hypothesis Validation
| Original Hypothesis | Status | Evidence | Confidence |
|-------------------|--------|----------|------------|
| [H1: Users struggle with X] | VALIDATED | 8/10 mentioned X | High |
| [H2: Users would pay for Y] | INVALIDATED | 2/10 showed buying signal | Low |
| [H3: Problem occurs daily] | PARTIALLY VALIDATED | Weekly for most (6/10) | Moderate |
```

**Verify**: At least 3 patterns identified with 40%+ frequency across interviews.
**If failed**: If no patterns emerge, interviews may have been too diverse in segments. Narrow to a tighter persona and re-interview.

### Step 5: Generate Deliverables

**Duration**: 2-3 hours
**Tool**: Document editor

**Output files**:
- `interview-guide-[stage].md` — Complete interview script with anti-patterns
- `interview-analysis.csv` — Cross-interview pattern analysis spreadsheet
- `discovery-findings.md` — Narrative findings report with validated/invalidated hypotheses and key quotes
- `next-steps.md` — Recommended actions based on findings

## Output Schema

```json
{
  "output_type": "interview_findings",
  "format": "JSON",
  "columns": [
    {"name": "total_interviews", "type": "number", "description": "Total interviews conducted", "required": true},
    {"name": "persona_segment", "type": "string", "description": "Which persona was interviewed", "required": true},
    {"name": "validated_hypotheses", "type": "array", "description": "Hypotheses confirmed with evidence and confidence", "required": true},
    {"name": "invalidated_hypotheses", "type": "array", "description": "Hypotheses disproven with evidence", "required": true},
    {"name": "discovered_patterns", "type": "array", "description": "New patterns found not in original hypotheses", "required": true},
    {"name": "pain_severity_ranking", "type": "array", "description": "Ranked pain points with frequency and severity scores", "required": true},
    {"name": "buying_signals", "type": "array", "description": "Indicators of willingness to pay with evidence", "required": true},
    {"name": "key_quotes", "type": "array", "description": "Verbatim quotes that capture critical insights", "required": true},
    {"name": "recommended_actions", "type": "array", "description": "Next steps based on findings", "required": true}
  ],
  "expected_row_count": "1 per persona segment",
  "sort_order": "pain severity descending",
  "deduplication_key": "persona_segment"
}
```

## Quality Benchmarks

| Quality Metric | Minimum Acceptable | Good | Excellent |
|---------------|-------------------|------|-----------|
| Interviews per persona | >= 5 | >= 8 | >= 12 |
| Specific behavior examples captured | >= 3/interview | >= 5/interview | >= 8/interview |
| Patterns with 40%+ frequency | >= 2 | >= 4 | >= 6 |
| Hypotheses tested | >= 3 | >= 5 | >= 8 |
| Verbatim quotes captured | >= 2/interview | >= 4/interview | >= 6/interview |

**If below minimum**: Continue interviewing. If scheduling is blocked, switch to asynchronous interview formats (email Q&A, Loom video responses).

## Error Handling

| Error | Likely Cause | Recovery Action |
|-------|-------------|----------------|
| Interviewees only give polite positive responses | Questions are leading or solution-focused | Rewrite questions to focus on past behavior; remove all mention of your product |
| No clear patterns after 10 interviews | Segment too broad or questions too generic | Narrow persona definition; add more specific behavioral questions |
| Interviewees cancel or no-show (>30% rate) | Insufficient motivation or poor scheduling | Add incentives ($50 gift card); offer 3 time slots; send reminder 1hr before |
| All answers are short and surface-level | Interviewee is distracted or questions lack depth | Add "Tell me about the last time..." probes; extend silence after answers (5-second rule) |
| Findings contradict existing team assumptions | Discovery working as intended (this is a feature) | Present raw evidence (quotes, patterns) to team; let data speak |

## Cost Breakdown

| Component | Free Tier | Paid Tier | At Scale |
|-----------|-----------|-----------|----------|
| Video calls | Google Meet ($0) | Zoom Pro ($13.33/mo) | Zoom Business ($18.33/mo) |
| Transcription | Zoom transcript ($0) | Otter.ai ($8.33/mo) | Rev ($1.50/min) |
| Participant incentives | None ($0) | Gift cards ($50/person) | Panel recruitment ($150-200/person) |
| Analysis tools | Google Sheets ($0) | Dovetail ($29/mo) | Dovetail Team ($79/mo) |
| **Total (10 interviews)** | **$0** | **$500-700** | **$2,000-3,000** |

## Anti-Patterns

### Wrong: Asking "Would You Use This?"
The classic Mom Test failure. Hypothetical questions produce hypothetical answers. Your mom would say yes to your terrible idea because she loves you. Strangers say yes because they want to be polite. Neither response has predictive value for actual purchase behavior. [src1]

### Correct: Ask About Past Behavior
Replace "Would you use X?" with "Tell me about the last time you dealt with [problem]." Past behavior is the single best predictor of future behavior. If they haven't tried to solve it, the pain is not severe enough to buy.

### Wrong: Pitching During the Interview
Starting with "We're building a product that..." and then asking for feedback. Once you pitch, the interviewee shifts from honest informant to polite audience. Every answer after the pitch is contaminated by social pressure. [src3]

### Correct: Problem-First, Solution-Last
Spend 90% of the interview understanding their world. Only mention your solution in the final 2 minutes, if at all. The best discovery interviews end without the interviewee knowing exactly what you're building.

### Wrong: Counting "Interested" as Validation
"8 out of 10 people said they'd be interested!" Interest is free. Commitments are expensive. The only signals that matter are: (1) they tried to solve it before, (2) they spent money on workarounds, (3) they asked when they could buy. [src5]

### Correct: Look for Commitment Signals
After sharing your concept (in the final minutes), look for concrete commitment: email introduction to a colleague, agreement to pilot, asking about pricing, requesting a follow-up with their boss. Words are cheap; actions are signal.

## When This Matters

Use this recipe any time you need to learn from customers or prospects through direct conversation — before building, after launching, when pivoting, or when sales patterns are confusing. The interview guide ensures consistent, bias-resistant data collection that agents and founders can act on with confidence.

## Related Units

- [Buyer Persona Development Methodology](/business/customer-research/buyer-persona-development-methodology/2026)
- [Jobs-to-be-Done Interview Methodology](/business/customer-research/jobs-to-be-done-interview-methodology/2026)
- [Ideal Customer Profile Framework](/business/customer-research/ideal-customer-profile-framework/2026)
- [Buyer Journey Mapping](/business/customer-research/buyer-journey-mapping/2026)