You get:
- “Would you buy this?” questions (people say yes, but won’t)
- no discovery of actual pain points
- no understanding of current alternatives
- no price validation (willingness to pay unknown)
- wasted time building features nobody wants
But customer discovery is not selling.
It is listening without bias.
- Problem questions: how do they currently solve it?
- Severity questions: how painful is it (1-10)?
- Solution questions: would they try your solution?
- Price questions: what would they pay?
- Never ask “would you buy this?” (biased question)
Without good interviews, you build the wrong thing.
This framework forces AI to create unbiased customer discovery scripts.
Assume the role of a customer discovery expert who creates unbiased interview scripts.
Your task is to create a customer interview script.
Generate:
1. SCREENING QUESTIONS (2-3)
- To qualify participants
- Ensure they have the problem
2. PROBLEM DISCOVERY QUESTIONS (5-7)
- Open-ended, non-leading
- Focus on past behavior, not future intentions
3. CURRENT SOLUTION QUESTIONS (3-5)
- How they solve the problem now
- What they like/dislike about current solutions
4. SEVERITY ASSESSMENT (1-2 questions)
- How painful is this problem (1-10)
- How often do they experience it?
5. SOLUTION REACTION QUESTIONS (2-3)
- How they react to your proposed solution
- Still non-leading ("What would you expect from a solution?")
6. WILLINGNESS TO PAY QUESTIONS (2-3)
- Price sensitivity
- Budget range
7. RECRUITING SCRIPT
- How to invite potential customers to interviews
INPUTS:
Product Idea:
[DESCRIBE]
Target Customer Profile:
[WHO ARE YOU TALKING TO?]
Problem Hypothesis:
[WHAT PROBLEM DO YOU THINK THEY HAVE?]
Proposed Solution:
[WHAT ARE YOU BUILDING?]
Interview Format:
[IN-PERSON / VIDEO CALL / PHONE]
Estimated Interview Length:
[15 MIN / 30 MIN / 45 MIN]
RULES:
- Never ask "would you buy this?" (people lie)
- Ask about past behavior (more reliable than future intentions)
- Problem questions first (before solution)
- Severity question: 8+ means urgent problem
- Listen more than you talk (80/20 rule)
- Record interviews (with permission) for analysis
- Interview 10-15 customers before building
- Interview 10-15 potential customers before writing code.
- Ask about past behavior, not future intentions.
- Never ask “would you buy this?” (people lie).
- Record interviews (with permission) for analysis.
- Look for problem severity of 8+ (on 1-10 scale).
Product Idea: Mobile app for freelancers to track time and invoice clients
Target Customer Profile: Freelance designers, writers, developers with 1-5 years experience
Problem Hypothesis: Freelancers waste 5+ hours/week manually tracking time and creating invoices
Proposed Solution: All-in-one mobile app with timer, invoice generator, and payment tracking
Interview Format: VIDEO CALL (Zoom)
Estimated Interview Length: 30 MINUTES
This framework improves outcomes by forcing:
- screening questions (qualify participants)
- open-ended problem discovery (uncover real pain)
- current solution analysis (understand alternatives)
- severity assessment (prioritize urgent problems)
- unbiased solution reaction (not “would you buy?”)
Great customer discovery doesn’t ask for opinions — it uncovers behaviors and pain points.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI business strategy tools, startup frameworks, and practical strategies for founders and entrepreneurs.