If your score is below 40%, two moves matter most: find the segment that already values you, and remove the specific blockers holding back the users who almost do. The common mistake is trying to lift the blended number by building for everyone at once.

Here's the practical roadmap: quick wins that can move your score fast, plus deeper product changes that create lasting improvement.

How Do You Improve Product-Market Fit?

You improve product-market fit by finding the user segment that scores highest on the Sean Ellis survey, doubling down on what they love, and systematically fixing blockers for "somewhat disappointed" users. Most teams see 5-10 point PMF score improvements within 1-2 quarters using this approach.

Understanding Your PMF Score

Before improving anything, make sure you're measuring correctly:

Run the disappointment survey (original methodology) with activated users (not new signups, not churned users)
Calculate (Very Disappointed ÷ Total) × 100
Segment. Your overall score hides the real story

Your overall score might be 35%. But what if your power users are at 55% and your casual users are at 18%? That is the finding. The overall number is an average, and it hides two completely different user populations.

As you work to improve, always track by segment. An overall improvement from 35% to 38% means nothing if your power user score dropped from 55% to 50%.

Quick Wins to Boost Your Score (Low-Hanging Fruit)

These don't require product changes. They can move your score in days or weeks:

1. Survey the Right Users

Are you surveying everyone who signed up? Stop.

Survey only users who have:

Activated (completed your "aha moment")
Used the product 3+ times
Been active in the last 30 days
Had two or three weeks of real usage

If you're including users who signed up and never came back, you're diluting your score with people who never gave your product a chance. That is not a fit problem, it is an activation problem.

2. Fix Your Survey Timing

When you send the survey matters. Too early and users haven't experienced enough value to have a strong opinion. Too late and you've already lost the "somewhat disappointed" users to churn.

Sweet spot: 2-4 weeks after activation, when users have had meaningful experience but are still engaged enough to respond.

3. Shift Your Acquisition Mix

Where are your most engaged users coming from? If, hypothetically, organic signups score 50% and paid ads score 20%, then some of your acquisition is bringing in people the product does not suit, and your blended number hides it.

That does not make paid acquisition bad. It does suggest the targeting is bringing in people the product does not suit. Refine your ad targeting to match the profile of your high-PMF users.

4. Simplify Onboarding

If users aren't reaching the "aha moment" fast enough, your onboarding is the bottleneck. Look at:

Time to value. How long does it actually take a new user to reach the point where the benefit is obvious, and can you shorten it?
Drop-off points. Where exactly do users abandon the setup flow?
Unnecessary steps. Are you asking for information you don't need upfront?

Every extra step between signup and value delivery costs you potential "very disappointed" users.

Deep Work: Product Changes That Move the Needle

Quick wins only get you so far. For lasting PMF improvement, you need deliberate product changes:

1. Double Down on What Works

Your "very disappointed" users love something specific about your product. Find out what.

Ask them:

"What is the main benefit you receive from [product]?"
"What would you miss most if you couldn't use [product]?"

Their answers will usually cluster around one or two themes. Build more of that. Whatever it is, the speed, the simplicity, the specific insight it provides, invest there.

2. Fix the Blockers for "Somewhat Disappointed" Users

Your "somewhat disappointed" users are the highest-leverage group. They see value but something's preventing them from becoming advocates.

Ask them:

"What is missing from [product] that would make it better?"
"What almost prevented you from signing up?"

Their top 2-3 answers are your product roadmap. These are not feature requests from random users. They are specific blockers named by people who already see the value but need something more to cross the line.

3. Narrow Your Focus

Trying to serve everyone? That's the single most common PMF killer.

If your product does five things, work out which one or two your "very disappointed" users actually named, and let the rest wait. A product that does one thing unmistakably well is easier to depend on, and easier to describe to the next person, than one that does five things adequately.

Cutting scope feels like cutting value, which is why this is hard to do. Users rarely want more features. They want their problem solved clearly.

4. Find and Clone Your Power Users

Who are the people who love you most? Study them:

What's their role or company stage?
How did they find you?
What features do they use most?
What problem were they solving when they signed up?

That profile overlaps with what Julie Supan called the high expectation customer: the most discerning person in your target market, who recognizes and enjoys what your product is best at. Rahul Vohra applied her framework to survey data at Superhuman. Once you can describe that person, you can go looking for more of them, and a score measured on an audience that fits the product tends to read higher than one measured on everybody.

5. Improve Time-to-Value

The longer it takes someone to reach the point where your product is obviously useful, the more of them you lose before they could ever miss it. Compress the distance between signup and that moment.

Tactics:

Pre-populate with sample data so users see value immediately
Guide users to one specific outcome in their first session
Remove every step that doesn't directly contribute to the first "aha"

Building a Feedback Loop with Early Users

PMF improvement isn't a one-time project. It's a continuous loop:

The Loop:

Survey. Run the disappointment survey on users with enough experience to have an opinion
Segment. Look at scores by user type, plan, tenure, source
Analyze. What separates "very disappointed" from "somewhat disappointed"?
Prioritize. Build what "somewhat disappointed" users are missing
Ship. Release changes quickly; small bets, fast iteration
Measure. Survey the next set of eligible users. Did the score move, and in which segments?
Repeat

Best Practices:

Keep the survey consistent. Same wording, same population criteria, same timing relative to activation
Track trends, not snapshots. A single measurement tells you where you are; a trend tells you where you are going
Don't react to individual comments. Look for patterns across 10+ responses before acting
Involve the whole team. Fit is not only a product concern. Marketing, support, and sales all shape how users experience your product

When to Re-Measure (And When Not To)

Measure After Major Changes

Shipped a significant feature? Changed your positioning? Adjusted pricing? Measure 2-3 weeks after to see the impact on PMF.

Don't Measure Too Often

Measuring every week buys you noise and survey fatigue. There is no cadence that is correct for every product: match it to how often people actually use yours, and to how long a change realistically takes to show up in how they feel.

Ignore Small Fluctuations

A move from 35% to 37% usually is not meaningful, because it sits inside the uncertainty of a typical sample. Work out the interval for your own response count rather than trusting a rule of thumb: 16 "very disappointed" out of 40 is a 40% score with a 95% interval running roughly from 26% to 55%.

Celebrate Sustained Improvement

The win is not touching 40% once. It is holding a strong score across several readings while the user base grows, and seeing it corroborated by retention and repeat use. A survey score on its own measures stated attitude, not fit.

The Bottom Line

Improving your PMF score comes down to three things:

Measure the right users. Survey activated users, not everyone who signed up
Build for your best segment. Double down on what power users love, fix blockers for "somewhat disappointed" users
Run the loop continuously. Measure, analyze, ship, measure again

The founders who improve PMF fastest treat it as a system, not a project. They measure consistently, segment ruthlessly, and ship based on what the data tells them.

FitSignal runs the disappointment survey, breaks the score down by persona and segment, and tracks the trend, so the analysis does not live in a spreadsheet.