The Product/Market Fit Engine
Turn PMF into a number you can raise: focus only on the 'somewhat disappointed' users whose reason for loving you matches the mass market's
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 95%
A systematic method to measure and increase product/market fit. You survey users with the Sean Ellis question, treat the '% very disappointed' as your PMF score, then grow it by deliberately ignoring both the very-disappointed and not-disappointed cohorts and working only the 'somewhat disappointed' users whose stated main benefit matches what your core fans love.
Origin
Built by Rahul Vohra at Superhuman, extending the 40% 'very disappointed' benchmark from Sean Ellis (who coined the term 'growth hacker'). Documented in Vohra's First Round Review post 'How Superhuman Built an Engine to Find Product/Market Fit'.
Core principles
- 01PMF is measurable, optimizable, and can be numerically increased
- 02The metric is '% who would be very disappointed to lose the product'; >40% predicts fast growth, <40% predicts a growth struggle
- 03Very-disappointed users already love you — don't over-index on their asks
- 04Not-disappointed users are a lost cause — ignore their feedback entirely
- 05The growth lever lives in the somewhat-disappointed segment, filtered to those whose main benefit matches your core value
How to run it
- 1
Measure the score
Ask users: 'How would you feel if you could no longer use this product?' with three options — very disappointed, somewhat disappointed, not disappointed. Your PMF score is the % answering 'very disappointed'.
Pro tip This metric is more predictive of success than NPS. 40% is the threshold benchmarked by Sean Ellis.
Watch out If you ever change survey method, all prior numbers are invalidated — treat it as a fresh baseline.
- 2
Learn what fans love
Ask the very-disappointed users what they love most about the product. For Superhuman it was speed, keyboard shortcuts, design aesthetic and time saved.
Watch out Don't build too heavily on this group's feature requests — they already love you, so their asks move the score least.
- 3
Split the somewhat-disappointed by main benefit
Ask the somewhat-disappointed users whether they'd be disappointed because of your core benefit (e.g. speed) or something else. Keep the group whose reason matches your fans; politely disregard the rest.
Pro tip Even if you built everything the mismatched group asked for, they'd keep pulling you in a different direction than your mass-market fans.
Watch out This is emotionally hard — you are deliberately ignoring paying users who are giving you sincere feedback.
- 4
Build the two-column roadmap
From the matched segment, list what they love (double down) and what holds them back (objections to overcome). Each planning cycle, spend half your time reinforcing what fans love and half systematically removing the objections of the matched somewhat-disappointed users.
Pro tip Run the engine on individual sub-products (e.g. Superhuman for Sales) once the whole product is too broad to survey as one thing.
In the wild
Superhuman's very-disappointed users loved speed, keyboard shortcuts and design. Vohra asked somewhat-disappointed users whether they liked Superhuman for its speed or something else — and disregarded the 'something else' group so the roadmap stayed aligned with the mass-market reason for love.
→ A roadmap 'guaranteed to increase PMF', concentrating engineering on the segment whose value driver matched the core fan base.
Common mistakes
Acting on all customer feedback equally
Building for not-disappointed and mismatched somewhat-disappointed users dilutes the product and lowers the PMF score.
Relying on NPS instead
NPS is far less predictive of growth than the '% very disappointed' question.
Is it for you?
Best for
Early-stage founders hunting for or trying to deepen product/market fit with a measurable, survey-driven loop
Not ideal for
Products with too few users to survey meaningfully, or mature suites so broad that one survey blends unrelated jobs-to-be-done
From the transcript
“you have to deliberately not act on the feedback of many of your early users”
“the companies that struggled to grow almost always had less than 40% very disappointed”
“but to focus on the segment of the somewhat disappointed people”
“spending half your time doubling down on what people really love and half your time systematically overcoming the objections of the somewhat disappointed users”
From the episode
Superhuman's secret to success: Ignoring most customer feedback, manually onboarding every new user, obsessing over every detail, and positioning around a single attribute: speed
Rahul Vohra (CEO)