You're guessing who your customer is
Talk to 15 people in your target segment about the problem — not your product. Converge on one sharp ICP and the top three buying triggers.
When to use
You're pre-PMF or entering a new segment. You don't have a sharp, written ICP yet — or your outbound reply rate is under 5%, which usually means the ICP is wrong.
Hypothesis template
If we run 15 problem interviews with [persona], we'll converge on 1 sharp ICP statement, 3 buying triggers, and the top 3 objections — enough to write outbound that lands.
Method
The problem: You're building outbound, decks, and pricing on assumptions. Every conversation surfaces a different "ideal customer." You can't write a cold email that lands because you don't know who hurts most or why.
What works: The Mom Test approach (Rob Fitzpatrick). Don't pitch. Don't ask "would you use this?" Ask about their actual life — what they did last week, what they spent money on, what they're frustrated by. Real behavior, not opinions about your product.
How to do it:
- Pick ONE persona to start (e.g. "Head of Growth at Series A B2B SaaS, 20–100 employees")
- Build a list of 30 — you'll need that to get 15 calls
- Reach out with a no-pitch ask: "I'm researching [problem space], not selling anything, would you do a 25-min call?"
- On the call: ask about the last time they hit this problem, what they tried, what they paid for, what made them give up
- Record + transcribe every call. Tag every quote about pain, current solutions, willingness to pay
- After 15: write the ICP statement, list the 3 most-mentioned triggers, list the 3 most-mentioned objections
Output: A single page with ICP, triggers, objections, and 5 verbatim quotes you can use in cold outreach.
Success metrics
- •Interviews completed (target: 15)
- •Convergent ICP statement
- •Top 3 buying triggers
- •Top 3 objections
- •Verbatim quotes captured
Prerequisites
- A target persona hypothesis to start from
- A list of 30 prospects to reach
- A note-taking or transcription setup
Common pitfalls
- •Pitching the product instead of asking about pain
- •Asking hypotheticals ("would you pay?") instead of past behavior ("what did you pay for?")
- •Talking to friends instead of strangers in the segment
- •Stopping at 5 — patterns don't emerge until ~12
Source: The Mom Test (Rob Fitzpatrick). Standard YC founder advice.
Suggested ICE scores
Ready to run this experiment?
Sign up free to use this playbook with step-by-step guidance and track your results.