The most expensive startup mistake is scaling before product-market fit and mistaking motion for evidence. Paid growth can simulate PMF for quarters. Enthusiastic pilot customers can simulate it in B2B. The only signal that doesn't lie is unprompted repeat behavior: people coming back, and eventually paying, without being pushed.
What a product-market fit validation sprint looks like
Four to six weeks, three phases:
- Evidence audit. I go through your retention cohorts, activation funnel, and churn interviews, working from the data you actually have. Blended averages get unblended; vanity metrics get named.
- Demand probes. Structured user interviews focused on what people did (instead of what they say they'd do), plus concrete tests that measure willingness to act: pricing pages, waitlists with commitment friction, concierge versions of unbuilt features.
- The verdict and the plan. A written read: where you have fit (often a narrower segment than hoped), where you don't, and the experiments most worth running next quarter. Sometimes the verdict is "stop scaling, fix retention", and hearing it from an outsider is what lets the team accept it.
Why my read is worth having
I've validated and invalidated ideas with my own money on the line. Medzin found real fit in healthcare discovery: 18,000+ users, Rs. 60L ARR, and a seed raise built on retention evidence. WisOwl AI reached 10,000+ jobseeker signups with zero paid marketing, which is itself a designed PMF test: organic pull or nothing. And eight years running growth experiments at CaaStle across a ~$50M+ ARR portfolio taught me exactly how convincingly a well-funded funnel can impersonate fit.
If you're debating internally whether you have PMF, you probably know the answer. I provide the evidence, the framing, and the plan to get there.