The most expensive mistake early-stage founders make is scaling a GTM motion before genuinely validating product-market fit. They have enthusiastic customer interviews (say validation). They have active free trial users (do validation). They interpret both as evidence that they are ready to hire sales reps and scale spend. Then the new customers churn, the sales team cannot replicate the founder’s close rate, and the company realizes it built a scaling motion on top of a foundation that was not yet solid.
Knowing how to validate PMF correctly requires understanding that validation has three distinct stages, each producing a different type of evidence with a different predictive value for scaling decisions. This guide walks through all three stages with specific thresholds, methods, and the Double Experimentation Loop that structures the full validation process.
Why PMF Validation Is a Process, Not a Moment
The conventional description of PMF as a moment — “you’ll know when you have it” — is unhelpful for operational decision-making. It creates a binary mental model where founders either have PMF or do not, which leads to both premature scaling (“we have customers, so we must have PMF”) and excessive caution (“we are not sure we have it, so we will not invest”).
The more useful model is PMF as a progressive accumulation of evidence across three stages. At each stage, the evidence is stronger and the confidence threshold for proceeding with investment is higher. No single stage is sufficient for scaling decisions; the combination of strong evidence across all three stages is what justifies moving from validation to execution.
Stage 1: Say Validation (Weakest Evidence)
Say validation is the evidence produced by customer conversations: interviews, surveys, feedback sessions, and demo discussions where customers articulate the problem, their urgency, and their interest in the solution.
What Say Validation Looks Like
- Customers describe the problem in specific, urgent terms without prompting
- Customers have tried other solutions and found them inadequate
- Customers use language that indicates urgency: “this is costing us,” “we need to fix this,” “this is a priority”
- Customers describe the problem consistently, with similar language and similar pain points, across interviews
What Counts as Valid Say Validation
Minimum threshold for Stage 1 say validation: 50+ customer conversations with the target segment, with consistent problem confirmation from at least 70% of those conversations. Below 50 conversations, the pattern is not statistically reliable. Below 70% consistent problem confirmation, the pain is not universal enough in the target segment to support a scalable motion.
What Say Validation Does NOT Prove
Say validation does not prove willingness to pay. It does not prove that customers will change their behavior. It does not prove that the problem urgency is high enough to drive a purchase decision. Every customer discovery interview produces say validation. Almost none of it reliably predicts what customers will actually do when asked to pay.
Founders who build business cases on say validation alone are working from the weakest form of evidence available. The say-do gap — the difference between what people say they will do and what they actually do — is one of the most well-documented phenomena in behavioral economics and is consistently underestimated by early-stage founders.
Stage 2: Do Validation (Medium Strength Evidence)
Do validation is the evidence produced by behavioral actions that do not involve money. It is substantially stronger than say validation because it requires customers to actually change their behavior — signing up, activating, returning, engaging — rather than just expressing interest.
What Do Validation Looks Like
- Email signups and waitlist conversions from a landing page describing the product concept
- Free trial activations that reach the core feature of the product
- Demo requests from prospects who found the product independently
- Return usage of a free tier or trial after the initial activation
- Completion of key activation milestones that indicate genuine engagement with the product’s value
What Counts as Valid Do Validation
Threshold for Stage 2 do validation:
- Landing page conversion rate above 10% from targeted traffic (not broad ads)
- Free trial activation rate — reaching the core value feature — above 30% of signups
- Return usage at Day 7 above 40% of active trial users
- Demo-to-trial conversion rate above 50% for prospects who match the ICP
These are directional thresholds, not absolutes. The key question is whether behavioral engagement is strong enough to distinguish genuine demand from curiosity or novelty effects.
What Do Validation Does NOT Prove
Do validation does not prove willingness to pay at the target price. Free tier engagement and paid tier conversion are systematically different. Customers who love a product at $0 or $10/month may not be willing to pay $200/month for the same product. The do-pay gap is smaller than the say-do gap but still substantial, particularly in markets where free or very low-cost alternatives exist.
Stage 3: Pay Validation (Strongest Evidence)
Pay validation is the evidence produced by actual purchase transactions at or near the target price point. It is the only stage that simultaneously confirms value delivery (the product solves a real enough problem to justify payment), value capture (customers pay what the business needs them to pay), and market reality (the price-value equation works for the target segment).
What Pay Validation Looks Like
- Unaffiliated customers paying standard pricing without significant discounting
- Pre-sales: customers paying for access to a product that does not yet exist in final form
- Renewals: customers choosing to continue paying after the initial contract period
- Expansion: customers paying for additional seats, usage, or features within their account
- Referrals: paying customers referring similar companies who then also pay
What Counts as Valid Pay Validation
The threshold that the investment community now treats as minimum pay validation: 5+ unaffiliated paying customers in the target ICP who purchased at or near target pricing without founder-relationship shortcuts and without heavy discounting, who have retained for at least one full contract period.
“Unaffiliated” is important: customers who bought because of a personal relationship with the founder, who received heavy discounts as a favor, or who are in industries completely outside the target ICP are not pay validation for the target motion. They may be revenue, but they are not PMF evidence for the segment you are trying to scale.
Pre-Sales as the Most Direct Pay Validation
The most direct and capital-efficient form of pay validation is pre-sales: charging customers for access before the product is complete. This removes the say-do gap entirely — customers either pay for the concept or they do not.
A real-world example of pre-sale validation strength: a GTM strategy course that attracted 3,000 signups on a waitlist page converted 600 pre-sales at the target price point — a 20% conversion rate from say-validation (signups) to pay-validation (purchases). This conversion rate provided strong PMF evidence before a single piece of content was produced.
Pre-sale conversion benchmarks: above 15% from a qualified waitlist is strong PMF evidence; 8–15% is moderate and warrants further testing; below 8% suggests the price, segment, or value proposition requires revision.
The Double Experimentation Loop
The Double Experimentation Loop is the structured framework for running all three stages of PMF validation efficiently. It has two phases that correspond to different evidence requirements and different resource investments.
Phase 1: Qualitative Validation (10–30 Data Points)
Phase 1 uses small-scale qualitative methods to test the core hypothesis: is the problem real, is the urgency high, and does the proposed solution direction resonate? Methods: customer discovery interviews, small beta access groups, manual outreach to gauge interest in a landing page concept, and early-stage prototype testing.
Phase 1 is not statistically significant — it is directional. It filters hypotheses before Phase 2 investment. If Phase 1 produces consistent problem confirmation with urgency language from 7 of 10 conversations, the hypothesis passes Phase 1 and moves to quantitative validation. If Phase 1 produces weak signals, the hypothesis is revised before Phase 2 investment.
Phase 2: Quantitative Validation (100+ Data Points)
Phase 2 uses larger-scale structured tests to confirm or reject the hypothesis with statistical significance. Methods: landing page tests to 500+ targeted visitors, outbound sequences to 200+ ICP-matched contacts, pre-sales campaigns to a qualified waitlist, and free trial experiments with controlled activation flows.
Phase 2 produces the pay validation evidence required for scaling decisions. The thresholds: pre-sales above 15% from a qualified list, or 5+ unaffiliated paying customers within the ICP at target pricing, or a trial-to-paid conversion rate above 20% with 90-day retention above 80%.
The GTM hypothesis validation framework details the full experimental design protocol for both phases. The pre-sale validation guide covers the specific mechanics of running an effective pre-sales campaign.
When to Stop Validating and Start Scaling
The transition from validation to scaling is justified when Phase 2 evidence meets the pay validation threshold AND the do validation metrics indicate the product delivers on its promise. Specifically:
- 5+ unaffiliated paying customers at target pricing in the target ICP
- Retention above 80% at 90 days for those customers
- At least one customer who has referred a similar company
- A documented sales process that produced the first 5 customers and can be handed to a hire
When these conditions are met, you are not trying to find PMF anymore. You are trying to prove you can acquire customers with a repeatable process — the GTM Fit stage. The early customer profile at this stage should be evolving toward the ideal customer profile as validation extends from early adopters into the leading edge of the mainstream segment.
Frequently Asked Questions
What is the difference between say, do, and pay validation?
Say validation is customer interviews confirming the problem exists and is urgent. Do validation is behavioral actions without payment — signups, trial activations, return usage. Pay validation is actual purchase transactions at target pricing. Each stage is stronger than the previous. Only pay validation from unaffiliated customers at target pricing is sufficient evidence to justify scaling investment decisions.
How many paying customers do you need to confirm PMF?
A minimum of 5 unaffiliated paying customers who match the target ICP, paid at or near target pricing, without heavy founder involvement in the sales process, and who have retained for at least one full contract period. This is a floor, not a target — more paying customers from a consistent ICP with consistent metrics provides stronger evidence and higher confidence for scaling decisions.
What is the Double Experimentation Loop?
A two-phase PMF validation framework. Phase 1 uses qualitative methods (interviews, small tests) with 10-30 data points to filter hypotheses and identify direction. Phase 2 uses quantitative methods (structured outbound, paid tests, landing pages, pre-sales) with 100+ data points to confirm or reject the hypothesis with statistical significance. Phase 1 filters; Phase 2 confirms.
What is a good pre-sale conversion rate for PMF validation?
Above 15% from a qualified waitlist (people who have seen the product concept and explicitly expressed interest) is strong PMF evidence at the price point tested. 8-15% is moderate and warrants testing whether price adjustment or message refinement improves conversion before concluding on PMF. Below 8% indicates the price, value proposition, or target segment requires revision before the pre-sale test is a reliable PMF signal.
When does validation end and scaling begin?
Scaling is justified when all three conditions are met: pay validation threshold reached (5+ unaffiliated paying customers at target pricing), retention is healthy (above 80% at 90 days), and at least one customer has provided a reference or referral. At this point, the validation work has produced enough evidence to justify hiring and investment in systematizing the acquisition process into a repeatable motion.
