NPS tells you how users feel about you. PMF tells you whether they would miss you if you disappeared. They are different questions, they predict different things, and most NPS tools cannot measure PMF even though they look like they could. This is the difference, the overlap, and why running both at once is usually the right call.

Most founders I talk to started with NPS because it was the metric their investors asked about. Then they hit a stage, usually somewhere between 50 and 500 users, where the NPS number stopped giving them useful signal. That is the moment to add PMF measurement. Here is why.

Quick Answer: What Is the Difference Between PMF and NPS?

PMF (product-market fit) measures whether your product is a need or a nice-to-have, using the Sean Ellis question "How would you feel if you could no longer use [product]?" with a 40% "very disappointed" threshold as the signal. NPS (Net Promoter Score) measures customer sentiment on a 0–10 scale using "How likely are you to recommend [product] to a friend?" and tracks loyalty as a time series. PMF is a gate ("do I have a real product?"), NPS is a thermometer ("how do users feel about the product I have?").

What NPS Actually Measures

NPS was created by Fred Reichheld at Bain & Company in 2003. One question, one number:

"How likely are you to recommend [product] to a friend or colleague?" (0–10)

Scores 9–10 are promoters. Scores 7–8 are passives. Scores 0–6 are detractors. Your NPS is the percentage of promoters minus the percentage of detractors, yielding a number between -100 and +100.

What NPS is good at:

Tracking customer sentiment over time
Flagging drops that correlate with product or support issues
Comparing against industry benchmarks
Driving customer success workflows (follow up with detractors, get referrals from promoters)

What NPS is not good at:

Telling you whether you have product-market fit
Distinguishing "loyal but replaceable" from "would miss deeply"
Predicting retention at the pre-PMF stage
Giving you a clear go/no-go signal on whether the product is worth building more of

What PMF Actually Measures

The product-market fit score comes from Sean Ellis, who developed it while working at Dropbox, LogMeIn, and Eventbrite. One question, one percentage:

"How would you feel if you could no longer use [product]?"

Very disappointed

Somewhat disappointed

Not disappointed (it isn't really that useful)

N/A – I no longer use it

Your PMF score is the percentage of respondents who say "very disappointed." The 40% threshold is the line Sean Ellis identified as the reliable signal of product-market fit across hundreds of startups he worked with.

What PMF is good at:

Telling you whether your product is a need versus a nice-to-have
Gating decisions about whether to scale, raise, or pivot
Identifying which user segments have fit even when overall does not
Focusing your roadmap on the users who would actually miss you

What PMF is not good at:

Continuous sentiment tracking
Industry benchmarking (PMF data is sparsely published)
Predicting short-term churn from passive unhappiness
Measuring support or UX quality directly

The Core Difference in One Sentence

NPS asks "How do you feel about the product you already have?" PMF asks "Is this product a need or a nice-to-have?"

Those sound similar. They are not. A user can rate you 10 on NPS (they love recommending you) and still say they would be "not disappointed" if you disappeared, because a competitor is one click away. Conversely, a user who rates you 6 on NPS (too buggy, too expensive, frustrating UX) can absolutely say they would be "very disappointed" without you, because the core thing you do is irreplaceable for them.

The first user is loyal but replaceable. The second is grumpy but hooked. PMF catches the second. NPS does not.

Two Examples That Make This Concrete

Example 1: SaaS With NPS 60 But PMF 22% (Loyal But Replaceable)

Imagine a project management tool. NPS is 60, genuinely high. Users like the interface, the onboarding is smooth, support responds in under an hour. They would recommend it to a friend.

But when you ask "how would you feel if this product disappeared," only 22% say "very disappointed." The rest say "somewhat disappointed" or "not disappointed," because Notion, Asana, Linear, Trello, and five more tools do roughly the same thing. The switching cost is low. The loyalty is real but shallow.

This is where NPS alone misleads you. You look at 60 and think you have a great product. You do. But you do not have product-market fit at the "need" level, and your long-term retention is going to reflect that when a cheaper or shinier competitor shows up.

Example 2: SaaS With NPS 35 But PMF 52% (Polarizing But Essential)

Now flip it. Imagine a niche compliance tool for indie SaaS founders dealing with EU VAT. NPS is 35 (mediocre). Users complain about the UI, the dashboard is ugly, some workflows are clunky.

But 52% say they would be "very disappointed" if the product disappeared. Because no one else solves this specific problem at this price point. The alternative is spending $400/month on a tax accountant or getting it wrong and paying penalties.

This is where NPS alone also misleads you. You look at 35 and think you are failing. You are not. You have strong product-market fit with a polarizing user experience, which is a fixable problem. If you only measured NPS, you would be tempted to rebuild the UI. If you measured PMF, you would realize the UI is fine enough, and you should invest in widening the wedge.

Why Most NPS Tools Cannot Measure PMF (Even Though They Look Like They Could)

Both surveys are single-question tools sent to users. Both produce a number. Both claim to measure "customer love." So can any survey tool run either?

Technically yes. Practically no.

An NPS tool can build the Sean Ellis question as a custom survey. What it cannot do:

Calculate the PMF score automatically. NPS tools calculate promoters-minus-detractors. They do not understand "very disappointed = positive signal." You end up calculating PMF in a spreadsheet.
Segment by High Expectation Customer. The whole Vohra PMF methodology hinges on filtering responses by HXC segment. NPS tools have no HXC primitive.
Track the 40% threshold as a success state. NPS tools chart trend lines but do not understand that PMF is a threshold metric. 39% and 41% are categorically different outcomes, not just 2 points apart.
Power the Vohra implementation loop. The 50/50 "double down on love / remove top blocker" workflow requires specific reporting on what very-disappointed users say they love and what somewhat-disappointed users say is missing. NPS tools do not ship that.

Delighted (sunsetting June 30, 2026) is the most famous example. You can run the Sean Ellis question inside Delighted. Most people never did, because the rest of Delighted was built around NPS workflows and fighting the tool to make it behave like a PMF tool was not worth it.

When to Use Each

Use NPS when

You already have confirmed PMF and are optimizing customer experience
Your business model relies on referrals (consumer apps, freemium SaaS)
You need continuous sentiment tracking, not a gate
Investors or boards require NPS specifically
You have a customer success team that acts on per-response feedback

Use PMF when

You are pre-PMF and need to know if you should keep building
You are early-PMF and trying to strengthen it
You are deciding between product iterations and need user-value signal
You are picking which segment to double down on (PMF by segment is often revealing)
You want a retention leading indicator rather than a sentiment lagging one

Use both when

You are between 50 and 5,000 users and serious about the product
You want both the current-state thermometer (NPS) and the depth-of-need gate (PMF)
You want to catch the "loyal but replaceable" and "grumpy but hooked" cases above
You have investors asking about NPS but know PMF is the number that actually predicts your business

Most indie developers benefit from running both from the start. They are not expensive, they are not redundant, and together they tell a complete story that neither one tells alone.

How to Run Both Without Doubling Your Survey Load

The mistake is sending two separate surveys to the same users every month. Survey fatigue kills response rates.

Better approach:

Send the Sean Ellis PMF survey once per quarter to your active user base. That is the PMF baseline. Re-measure each quarter to catch drift.
Send the NPS survey on a trigger: after a completed onboarding, after 30 days of activity, after hitting a key feature usage milestone. That gives you continuous sentiment data without hammering users.
Ask the same "what is the main benefit" open-ended follow-up on both. Over time, the qualitative responses converge and tell you which value props are loadbearing for retention versus which just drive recommendations.

In practice, a user sees the PMF survey once per quarter (2 seconds to answer) and the NPS survey maybe 2–3 times per year on triggers. Total survey load: under 10 interactions per user per year. Manageable.

PMF and NPS in the Same Tool

Historically you needed two tools. NPS specialists (SatisMeter, Delighted) were strong on NPS but weak on PMF. PMF measurement was usually a custom setup in Typeform or Google Forms, which meant manual segmentation and no integration.

That is why FitSignal was built. The Sean Ellis survey is a first-class object with automatic scoring, HXC segmentation, and the Vohra engine roadmap view baked in. NPS surveys are a secondary survey type in the same tool, so your users, properties, and historical data live together in one place.

If you already run NPS elsewhere and just want to add PMF measurement, the free tier handles 250 survey sends per month, enough to run a quarterly PMF baseline alongside your existing NPS setup without changing anything.

Best For / Not Best For

This article is best for: Indie developers, early-stage SaaS founders, and growth-stage PMs who run NPS (or are about to) and want to understand what they are missing. Also useful for anyone confused by the "NPS is dead" vs "NPS is gospel" debate. Both camps are wrong because they are pointing at the wrong metric.

This article is not best for: Large enterprise ops teams with established NPS programs running through CX platforms (the tool switch cost outweighs the tooling gains), or consumer apps where referral loops are the whole business (NPS alone is probably the right call for you).

Bottom Line

PMF and NPS are not competing metrics. They are complementary ones. NPS gives you the month-over-month pulse. PMF gives you the gate. Without the pulse, you miss emerging quality issues. Without the gate, you scale a product that users do not actually need. That is how most startups die politely, at NPS 50, wondering why their growth keeps flattening.

Run both. Send the Sean Ellis PMF survey quarterly, send NPS on triggers, and pay attention when they disagree. The disagreements are where the real product insights live.

If you want to measure both in one place without wrestling two separate tools, FitSignal handles both on the free tier. Or stick with your existing NPS setup and add PMF measurement alongside. The important thing is having both numbers, not where they live.