How to Improve Your PMF Score (From Below 40% to Funded)

Let's start with the feeling, because it's the thing standing between you and your next move. A low PMF score feels like a verdict. Like the survey just told you the truth you were afraid of: it's not working.

It didn't. Here's what a 28% actually means: more than a quarter of your users would be genuinely upset to lose your product. That's not a failing startup. That's a foothold — a real group of people who already get real value — and your entire job now is to widen it.

The founders who win aren't the ones who score 40% on the first try. They're the ones who treat the first score as a baseline and run a disciplined loop to raise it. Superhuman did it from 22%. So can you. Here's the loop.

Don't know your current number yet? Calculate your PMF score here first — then come back and start climbing.

First, the mindset shift that makes the rest work

Before any tactics, internalize this, because it governs every decision below:

Most founders, staring at a low score, do the natural thing: try to fix everything for everyone. More features, more use cases, more flexibility, fewer "no"s. And the score stays flat or drops — because a product that's a little better for everyone is a must-have for no one. The whole loop below is built on the opposite instinct: get narrower, not broader.

Step 1: Segment — split your users into three

Your survey gave you more than a number. It sorted your users into three groups, and the entire strategy comes from treating them differently:

If you do nothing else, do this split. It converts a vague "the score is low" into three specific, different to-do lists.

Step 2: Define your ICP from the people who love you

Zoom all the way in on the "very disappointed." Forget everyone else for a moment. Ask: who are these people, specifically?

Profile them across whatever you can see — role, company size, stage, industry, what they used before you, how they found you. You're hunting for the pattern that makes them different from the users who shrugged. Almost always, one emerges: a particular kind of person, with a particular problem, for whom your product is a genuine relief.

That pattern is your ideal customer profile, and you didn't invent it — your happiest users handed it to you. This matters enormously, because the fastest way to raise a PMF score is often not to change the product at all. It's to acquire more of the people who were always going to love it and stop diluting your numbers with people who never would.

Step 3: Find your core value (and protect it)

Now read what the "very disappointed" wrote when you asked about the main benefit they get. The answers will cluster. For Superhuman it was speed. For your product it's something specific — and probably narrower than your marketing currently claims.

That cluster is your core value, and it's now sacred. Every roadmap decision gets measured against it: does this make the core value stronger for the people who already love it, or does it pull attention away? The number-one way startups lose a hard-won score is by neglecting the thing fans love while chasing features for people who don't care.

Step 4: Build the 50/50 roadmap

Here's the engine itself, straight from Rahul Vohra's playbook. Split your roadmap roughly in half:

Half 1 — Double down on what fans love

Take your core value and make it more pronounced. If fans love speed, make it faster. If they love simplicity, cut more. This deepens the moat with your "very disappointed" users and pushes the merely-very-happy toward evangelism. It also defends the score you've already earned.

Half 2 — Convert the "somewhat disappointed"

Go to the fence-sitters' answers to "how can we improve this for you?" Those are concrete blockers — a missing integration, a workflow gap, a feature they need before they can rely on you. Fixing them moves people from "somewhat" to "very," which is exactly what raises your score.

The critical filter: only build the blockers that fit your ICP. A fence-sitter's request that would pull you toward the wrong audience gets declined, even if several people ask. You're not trying to satisfy everyone who responded — you're trying to deepen fit for your kind of user.

Step 5: Re-survey and read the trend

Ship a cycle of improvements, then run the survey again. Watch the score. Three things to look at:

• The headline number — is the percentage of "very disappointed" climbing?
• The migration — are "somewhat" users moving into "very"? That's the conversion working.
• The ICP concentration — is your fan base getting more sharply defined? Good. That's focus.

Then do it again. And again. This is the part that separates the 22%-to-58% stories from the startups stuck at 25% forever: they keep running the loop. One cycle rarely cracks 40%. Four or five focused cycles often do.

Real trajectories: what movement looks like

To make this concrete, here's the shape of the journey — the famous one, and a more typical one.

Notice these aren't overnight jumps. They're the result of sustained iteration — measure, narrow, build, re-measure — over quarters. The score is a slow-moving number precisely because real fit is hard to fake, which is also what makes it credible to investors when it does move.

The fundraising payoff

Here's why the trend matters as much as the destination. A founder who walks into a raise saying "we have product-market fit" is asking the VC to take their word for it. A founder who says "we went from 28% to 44% over two quarters, here's the ICP driving it, and here's what those users say they'd miss" has handed over evidence — a story of demand they understand and can steer.

Even a sub-40% score becomes a strength when it's rising with a clear ICP narrative attached. Investors aren't only buying fit; they're buying your grip on it. See exactly what they want in an investor-ready PMF report, and how to pair the score with retention in our complete measurement guide.

What NOT to do (the score-killers)

• Don't build for the "not disappointed." Every hour spent winning over people who'll never love you is an hour stolen from your fans.
• Don't spread the roadmap across every request. Breadth is the enemy of fit. Pick the blockers that serve your ICP.
• Don't acquire faster than you fit. Pouring the wrong users into a sub-40% product just dilutes the score and burns cash. Earn the fit, then scale it.
• Don't measure once. Without the trend, you can't tell whether your changes are working — you're back to guessing.

Your next cycle, starting this week

1. Run the survey on your engaged users to set (or refresh) your baseline.
2. Segment into very / somewhat / not disappointed.
3. Define your ICP from the very-disappointed group and their open-ends.
4. Set a 50/50 roadmap — half doubling down, half converting fence-sitters who fit the ICP.
5. Ship, then re-survey. Track the trend, not the snapshot.

A low score was never the bad news you thought it was. It's the clearest instruction set your users will ever give you. Run the loop, and watch your number climb toward the line — and past it.