mfwarren Product-Market Fit
PMF surveys, Sean Ellis test, retention analysis, and pivot frameworks.
Purpose
Determine whether you've found product-market fit and, if not, what to change. PMF is the single most important milestone for any startup.
Workflow
Step 1: Gather Context
- Product/service description
- Current customer count and engagement
- How customers find you today
- Retention/repeat usage data (if available)
- What customers say they love (and hate)
Step 2: Sean Ellis Test Design
Survey question: "How would you feel if you could no longer use [product]?"
- Very disappointed
- Somewhat disappointed
- Not disappointed
- N/A — I no longer use it
Benchmark: 40%+ "Very disappointed" = PMF signal
Design the full survey (5-8 questions) to understand:
- Who are the most passionate users?
- What's the primary benefit they get?
- What would they use as an alternative?
- How did they discover you?
Step 3: PMF Assessment
Based on data provided:
- Strong PMF signals: High retention, word-of-mouth growth, pull from customers
- Weak PMF signals: High churn, feature requests that change the core, price sensitivity
- No PMF signals: Growth only from paid acquisition, low engagement, high support volume
Step 4: Pivot Framework (if needed)
If PMF isn't there, evaluate:
- Zoom in: Double down on the feature users love most
- Zoom out: Your feature should be a platform
- Customer pivot: Same product, different audience
- Need pivot: Same audience, different problem
- Channel pivot: Same product, different distribution
Step 5: Action Plan
Concrete next steps based on assessment.
Output Format
## PMF Assessment: [Product]
### Current Signals
| Signal | Status | Evidence |
|--------|--------|----------|
| Retention | [Strong/Weak] | [Data] |
| Word of mouth | [Strong/Weak] | [Data] |
### Sean Ellis Survey
[Survey questions]
### Assessment
[PMF status and reasoning]
### Recommended Actions
1. [Action]
2. [Action]
Constraints
- PMF is not binary — present it as a spectrum
- Don't declare PMF based on vanity metrics (signups, downloads)
- Be honest if the data suggests a pivot is needed
- Note that PMF can be lost — it's not permanent