Product-market fit is not a feeling and not a funding event. It is a measurable state: a defined segment uses the product repeatedly, pays for it, and would be upset if it disappeared. This playbook sets the thresholds and the sequence for getting there.
The three signals that matter
- Retention flattening - a cohort curve that stops declining and holds flat is the single strongest signal. Falling to zero means no fit, whatever the growth chart says.
- The 40% test - ask active users how they would feel if they could no longer use the product. Above 40% answering 'very disappointed' indicates fit in that segment.
- Pull - inbound signups, unprompted referrals and users asking for invoices before you ask for payment.
Benchmarks to measure against
- B2B SaaS: >85% logo retention over 12 months; net revenue retention above 100%.
- Consumer subscription: >40% of a monthly cohort still active at month 6.
- Marketplace: >50% of buyers transacting again within one purchase cycle.
- Transactional B2B: repeat order rate above 60% within the natural reorder window.
The diagnostic sequence
- Define one segment precisely - role, company size, and the trigger event that creates the need.
- Instrument the core action and measure how many users complete it in week one.
- Chart weekly cohorts for at least eight weeks; look at the shape, not the average.
- Interview the ten most engaged and the ten who churned. Ask what they did, not what they think.
- Rank the gaps between those two groups and fix only the top one before re-measuring.
When the signal is weak
Weak retention with strong signup means a positioning or onboarding problem, not a product problem - the wrong people are arriving, or the right people never reach value. Strong retention in a tiny group means narrow your definition of the market and go deeper before widening.
Do not scale paid acquisition to fix retention. Paid growth on a leaking product converts a solvable problem into an expensive one and destroys the very cohort data you need.
What to tell investors
- Show the cohort chart, not the cumulative user count.
- Name the segment where fit is strongest and state its size honestly.
- State the current retention number and where it was three months ago.
- Describe the one experiment running now to improve it and when it reads out.