How to Measure Product-Market Fit (Not What You Think)
PMF is not a feeling. It's when your product satisfies a need in a way your market is willing to pay for. Measure it with retention, NPS, and churn trajectory—not vanity metrics.
The Core Answer
Product-market fit is not a moment, a feeling, or a milestone you pass. It’s a state where your product reliably satisfies a market need in a way the market is willing to pay for. Measure it with three observable facts: cohort retention is stable (new cohorts behave like old cohorts after the same elapsed time), NPS from power users is consistently >50, and churn is declining or flat as you grow. If retention is flat or negative, cohorts are unpredictable, or NPS is below 40, you don’t have PMF—you have early-stage product with maybe product-market hypothesis. The distinction matters because it changes what you optimize for: PMF means optimize for profitability and expansion; pre-PMF means optimize for learning and retention.
Retention: The True Signal
Retention is the North Star because it’s the hardest to fake. A user can download your product for vanity, click a link because they’re curious, or try a feature because a competitor told them to. But a user who comes back week after week is signaling that the product solves a problem they have.
For B2B SaaS, measure cohort retention: group users by signup week (or month), then track what percentage are active in week 4, 8, 12, 16, 24, 52. Plot each cohort. Do they look the same? If cohort 1 (January signups) has 60% retention at week 4 and cohort 10 (October signups) has 55%, you have repeatable behavior. That’s signal.
For B2C, the same logic applies, though the retention curve is steeper. A fitness app might have 40% week-1 retention, 20% week-4 retention, and 10% month-3 retention. The shape matters. If new cohorts follow the same curve, that’s signal.
If retention is declining (cohort 10 has 50% week-4 retention when cohort 1 had 60%), you’re getting weaker signals—either product is degrading, product-market fit is deteriorating, or you’re signing up the wrong users.
NPS and Power User Satisfaction
NPS (Net Promoter Score) is a coarse instrument, but it’s useful as a directional signal when applied to power users—people who have deeply engaged with your product and would notice if you took it away. Don’t ask all users “would you recommend this?” Ask power users (top 10% by engagement) and segment the answer: How many are 9–10 (promoters)? How many are 6–8 (passives)? How many are 0–5 (detractors)?
If 60% of power users are promoters (NPS = 60), you have strong signal. If 30% are promoters, you have weaker signal. If detractors outnumber promoters, you don’t have fit—you have product that some users have learned to work around but wouldn’t choose again.
NPS matters because it’s a leading indicator of retention and expansion. A power user with NPS 9 will keep paying, refer colleagues, and upgrade. A power user with NPS 4 will churn when a competitor shows up.
Churn: The Trajectory Matters More Than the Number
A 5% monthly churn sounds healthy until you realize it’s climbing: month 1 was 2%, month 2 was 3%, month 3 was 5%. That trajectory says something is breaking. Flat or declining churn (even at 8–10%) says something is holding.
Segment churn by cohort age. Customers in their first month churn at one rate; customers in their 12th month churn at another. If first-month churn is rising, your onboarding is degrading. If year-1 churn is rising, product is deteriorating or you’ve saturated the use case.
PMF looks like: early churn (month 1) moderates over time, annual churn stabilizes below 10%, and the curve is flat quarter-over-quarter.
The PMF Checklist
You likely have PMF if:
- Cohort retention curves are overlapping (new users behave like old users)
- Week-4 or month-1 retention is >30% (B2B) or >25% (B2C)
- NPS from power users is consistently >50
- Churn is declining or flat
- Inbound demand is growing without increased marketing spend
You likely do NOT have PMF if:
- Retention curves are diverging (cohorts are unpredictable)
- Month-1 retention is below 20%
- NPS from power users is below 40
- Churn is rising month-over-month
- Growth requires sustained high marketing spend with declining CAC payback
Common Mistakes in Measuring PMF
Vanity metrics. DAU, signup velocity, or funding announcements feel like PMF but they’re not. They measure attention, not attachment. A user who downloads your app but doesn’t open it again next week is not evidence of fit.
Averaging cohorts. If you average all users’ retention, you’ll miss the signal. One cohort might be 60% at week 4, another 40%. The average is 50%, which looks healthy but masks two different products. Always segment by cohort.
Ignoring churn composition. If 50% of churn is “customer deleted account” and 50% is “license expired because company downsized,” those have different meanings. One says product failed; one says customer survived but contract ended. Know your churn.
Treating PMF as binary. PMF is not a light switch. You have increasing levels of fit: product solves a problem for some customers (weak fit), solves it better than alternatives (good fit), and customers are willing to expand (strong fit). Most products are somewhere on that spectrum.
How to Measure This Month
1. Pull cohort retention data. Segment users by signup week. Track active users in weeks 1, 4, 8, 12. Plot each cohort. Do they overlap?
2. Segment power users and survey them. Define power users (top 10% by engagement, revenue, or logins). Ask 5 of them: “On a scale of 0–10, how likely are you to recommend?” Look for the pattern, not the individual response.
3. Measure churn by segment. Calculate monthly churn. Track it month-over-month. Is it flat, rising, or falling? Is early churn stable?
4. Articulate the gap. Where’s the weakest signal? Is retention too low? Is churn rising? Is NPS weak among power users?
5. Hypothesize why. If retention is weak, why? Product isn’t solving the problem, or onboarding isn’t clear? If churn is rising, why? Competitor advantage, feature degradation, or wrong customer segment?
The Bottom Line
PMF is measurable: stable cohort retention, NPS >50 among power users, and declining or flat churn. These three signals together say your product is reliably solving a market need in a way the market will sustain. Vanity metrics (DAU, signup velocity) don’t measure fit. Averaging cohorts hides the signal. PMF is not binary—it’s a spectrum—and your job is to know where you are and what’s pulling you forward or backward.