samarv/pmf-threshold-gating

PMF Threshold Gating

To drive massive organic word-of-mouth growth, you must move past "good enough" to "great enough." This framework uses a high-threshold Sean Ellis Score as a hard gate to prevent "scaling a small problem into a big mess." By demanding that 50% of users feel "very disappointed" if a product disappeared, you ensure the product is fundamentally different rather than incrementally better.

The Pre-Build Gate: Mock Press Release

Before assigning engineering resources, validate the value proposition:

• Write a two-paragraph mock press release or customer announcement.
• If you cannot articulate why the target customer should care in two paragraphs, the strategy is not clear enough to build.
• Rule: "We may not be right, but at least we are clear."

The PMF Measurement Workflow

1. The High-Bar Sean Ellis Survey

Ask a cohort of early users: "How disappointed would you be if this product went away?"

• The Scale: 1. Not disappointed, 2. Somewhat disappointed, 3. Very disappointed.
• The Nubank Threshold: 50% must answer "Very Disappointed."
• Why 50%? The standard 40% often includes "politeness bias." A 50% bar ensures the product is a "must-have" that drives fanaticism.

2. Identify the "Bullseye Cohort"

If your overall score is below 50%, find the segment where it is above 50%.

• Filter the data by usage patterns (e.g., "Users who used the feature 4+ times").
• Look for the common denominator among the "Very Disappointed" group.

3. The "Direct-Line" Investigation

Once the bullseye cohort is identified, the PM/Designer must skip the research department and call them directly.

• The Rule of 10: Call 10 customers from the bullseye cohort personally.
• The Observation Method: Ask indirect questions to find their pain points rather than asking if they like your solution.
• Goal: By the 5th or 6th call, you should be able to predict what the 7th person will say. This indicates you have found the "truth" behind the PMF.

4. Iterate to Scale

Do not scale the product to the mass market until the Sean Ellis score hits the threshold in the target segment.

• If the score is low, use the direct call insights to adjust features until the 50% mark is cleared.
• Mantra: "We are not going to take a small problem and scale it."

Examples

Example 1: Payments Assistant (Assistente de Pagamentos)

• Context: A bill-pay automation tool launched in Brazil.
• Initial Result: Sean Ellis score was borderline (40%).
• Bullseye Cohort Discovery: Analysis found a segment with 4+ commitments on 2+ different payment rails had a 70% score.
• Application: The team pivoted the roadmap to focus on multi-rail onboarding and automated registration for high-volume bill payers.
• Output: Grew from hundreds of thousands to over 10 million monthly actives.

Example 2: High-Income Rewards Card (Ultravioleta)

• Context: A premium card with a monthly fee and rewards.
• Initial Result: Mixed feedback.
• Bullseye Cohort Discovery: Only users who spent enough to trigger the fee waiver loved the product fanatically.
• Application: The team spent 2 years in "the lab" iterating on the specific benefits for that segment (fee waiver transparency, specific rewards) before scaling.
• Output: Achieved decisive PMF in the high-income segment before mass marketing.

Common Pitfalls

• Scaling a "Leaky Bucket": Scaling a product with a low Sean Ellis score usually results in high churn. You can "bounce a dead cat" (make metrics look good for months) using a large existing user base, but the product will eventually fail.
• The "Lawyer" Mindset: Becoming a lawyer for your hypothesis (defending it) instead of a judge (evaluating it objectively). Use customer anecdotes to trump your data.
• Hedging the Strategy: Trying to serve all customers at once. Startups must concentrate bets on a specific segment to build wealth; diversification is only for preserving it.
• The Research Distance: Relying on a third-party researcher to summarize findings. PMs must "feel" the customer's tone of voice via direct phone calls to understand the nuance of their pain.