Find out if you have
product-market fit
in 60 seconds.
Product category
Benchmarks adjust per category — infrastructure has longer time-to-value tolerances than dev tooling.
Go-to-market motion
PLG measures retention as D7; enterprise uses annual logo retention.
Annual revenue
This sets your PMF level on the revenue axis (0–7).
Time to first value
From signup to the user experiencing the core benefit.
Retention
Day-7 active retention of new signups.
Net revenue retention
Expansion − contraction − churn across your existing cohort.
Free-to-paid conversion
Share of free users who become paying.
Organic signup share
Portion of signups arriving without paid acquisition.
Two complementary methods, one tool.
Survey mode uses the Sean Ellis test: if ≥40% of your users would be "very disappointed" if they could no longer use your product, you likely have PMF. The best-known teams have used this test to lift their score from the twenties into the fifties in a single year.
Compass mode scores 5 behavioral metrics (TTFV, retention, NRR, conversion, organic) against devtools-industry benchmarks plus an ARR-based revenue stage. Signal ahead of revenue = distribution problem; revenue ahead of signal = foundation risk.
Questions we get.
How many survey responses do I need?
You get directionally useful signal around 40 responses, reliable data above 100. Below 40 our tool flags "low confidence" and shows a wider Wilson 95% CI so you can see the noise.
Why 40% and not 50%?
Sean Ellis derived 40% empirically from ~100 startups: companies that beat 40% almost always found growth; those below almost always struggled. It's calibrated from data, not a round number.
Why "disappointed" instead of "satisfied"?
Loss framing surfaces dependency. Positive framing invites politeness bias — users say they'd miss a product they rarely use. Disappointment reveals who actually relies on you.
Does this work for B2C?
Survey mode works for any product. Compass mode benchmarks are devtools-specific (based on 37 companies) — so the buckets will skew for consumer apps. Use Compass as a framework, not a scorecard.
Is NPS enough?
No. NPS measures referral intent; PMF measures dependency. Promoters who are also "very disappointed" are your most valuable cohort — use both and cross-reference.
Does my data leave my browser?
No. Everything is computed client-side. Share links encode only aggregate numbers — never raw survey answers or CSV rows.
What if I only have 35 responses?
You'll see a "low confidence" badge and a wide confidence interval. Interpret cautiously; re-run when you hit 100. Do NOT nag users for responses — non-response is baked into the 40% benchmark.
Ready to turn signal into action?
Hitting 40% "very disappointed" is the start of the climb, not the summit. Let's talk about what compounds next.
Schedule session