Blog · Product

7 signals your SaaS is getting closer to product-market fit

PMF does not arrive as a notification. It emerges when several behaviors tell the same story: the right users stay, pay, recommend, and ask for more.

Onny team13 min read

Product-market fit is a body of evidence, not one magic metric

A signup spike may come from a launch. A few customers may stay because they receive custom service. One strong revenue month may depend on a channel that cannot be repeated. None proves that the market is truly pulling the product.

Product-market fit appears when an identifiable group has an important problem and adopts your product as a durable solution. Qualitative evidence explains why; cohorts and revenue show whether the behavior survives over time.

These seven signals are not a checklist where five boxes produce a verdict. Look for convergence within the same segment.

1. Retention stabilizes for a specific group

The hardest signal to fake is a retention curve that stops falling. Some users reach value, return, and keep using the product weeks or months later.

Analyze cohorts by signup date, segment, and use case. A global average can hide a small, highly retained population behind a majority that leaves.

What to observe

  • Does the curve eventually flatten?
  • Are newer cohorts retaining at least as well as early ones?
  • Which early action separates retained users?
  • Does retention exist in usage, accounts, and revenue?

Watch out: High first-week activity is not retention. Match the window to the natural frequency of the problem.

Y Combinator — Retention and product-market fit

2. Users would genuinely miss the product

Sean Ellis’s question asks how users would feel if they could no longer use the product. The famous 40% “very disappointed” benchmark is a useful reference, not a universal law.

Survey only users who experienced core value. Study who answers “very disappointed,” which benefit repeats in their words, and what holds back users who seek that same benefit.

What to observe

  • Are respondents active and qualified?
  • Does one core benefit repeat spontaneously?
  • Is the score improving within a coherent segment?
  • Do stated answers match behavior?

Watch out: A tiny, fan-heavy sample can clear 40%. Use the score to identify the segment and benefit that resonate, then track movement.

First Round Review — Superhuman’s PMF engine

3. The problem is urgent before you present the solution

Near PMF, customers do not merely like the idea. They already tried to solve the problem with a spreadsheet, agency, script, hire, or another tool, and can explain what the old way costs.

Ask about past events rather than intentions. “We spent six hours every Monday” is stronger evidence than “we would probably use this.”

What to observe

  • What triggered the search?
  • What solution was already in use?
  • What cost or risk was already accepted?
  • Why did the decision have to happen now?

Watch out: Compliments, waitlists, and positive demo feedback measure interest. PMF begins with behavioral change and real tradeoffs.

4. Customers pay without every sale requiring an exception

Payment shows that perceived value beats another use of budget. The signal strengthens when customers renew, expand usage, or upgrade without permanent discounts or custom development.

Watch revenue retention, expansion, renewal, and churn reasons. A company can grow before PMF by selling bespoke services or subsidizing every account.

What to observe

  • Do customers renew at the intended price?
  • Does usage expand naturally?
  • Does support and onboarding time fall per account?
  • Do margins remain healthy with volume?

Watch out: Revenue alone is not PMF. If every contract requires a different roadmap or founder intervention, you may have profitable consulting rather than a repeatable product.

Y Combinator — The Real Product Market Fit

5. Users create part of the growth

When a product solves an important problem well, users recommend it to others in the same situation. New prospects repeat customer language and understand the promise quickly.

Track invitations, unsolicited recommendations, brand search, mentions, and accounts attributed to word of mouth—even without a built-in viral loop.

What to observe

  • How many new accounts cite an existing user?
  • Do customers share without incentives?
  • Is recommendation language consistent?
  • Does acquisition cost fall in the best-retained segment?

Watch out: A viral launch creates attention, not necessarily PMF. Check whether referred users activate and retain.

6. One segment loves the product far more than others

PMF often appears first in a narrow pocket defined by role, maturity, workflow, or context. The average can look weak while one segment retains, pays, and recommends much better.

Describe this group by situation and behavior, not demographics alone. They usually feel the problem most intensely or reach the core benefit with the least friction.

What to observe

  • Which segment retains best?
  • Who reaches value fastest?
  • Who would be most disappointed without the product?
  • Which context makes the promise obvious?

Watch out: Do not slice data until a good number appears. The segment must be identifiable before purchase, large enough, and reachable through a repeatable channel.

First Round Review — Superhuman’s PMF engine

7. Demand starts putting pressure on the team

When the market pulls the product, problems change: onboarding fills up, inbound grows, support increases, customers request integrations, and the team struggles to keep pace. This can indicate demand no longer driven solely by founder outreach.

The signal only matters with retention and sound economics. Aggressive paid acquisition can overwhelm a team while hiding weak retention or negative margins.

What to observe

  • Do prospects arrive with formed intent?
  • Are sales cycles shortening for the right segment?
  • Do requests cluster around the same benefit?
  • Is growth healthy after support and infrastructure?

Watch out: Being busy proves nothing if growth is bought at a loss. Operational pressure must accompany retention, value, and sustainable economics.

Y Combinator — The Real Product Market Fit

False signals that look like PMF

Many signups

They may come from a launch or temporary channel. Check activation and retention.

Highly enthusiastic users

Confirm that they use, pay, and belong to a reachable segment.

Fast revenue growth

Separate product revenue, bespoke services, discounts, and subsidized acquisition.

Many feature requests

They may indicate an overly generic product. Look for one shared core benefit.

Investor and media interest

They evaluate a story and potential; customers demonstrate repeated value.

How to read the seven signals together

  1. 01Start with retention: is there a cohort that stays?
  2. 02Identify the segment and benefit shared by that cohort.
  3. 03Verify urgency through behavior and interviews.
  4. 04Confirm economic value through renewal, expansion, and margin.
  5. 05Observe whether recommendation and inbound make growth more organic.
  6. 06Repeat for every new segment: PMF can be lost while expanding.

What Onny can monitor

By connecting Supabase, Stripe, support, and product tools, Onny can track activation, cohorts, usage, renewals, churn, recurring requests, and segment changes—then deliver a weekly brief explaining what changed instead of listing isolated metrics.

You approach PMF when the evidence starts reinforcing itself

Users with the strongest problem reach value, stay, pay, recommend, and ask for more. Data confirms their stories, and their words explain the curves.

While signals conflict, do not seek a flattering verdict. Find the segment where they begin to converge, then become indispensable to it before expanding.

Sources and further reading

Talk to your product data

Onny can analyze Supabase cohorts, connect them with Stripe, and surface the signals that deserve a decision.

Explore the Product usage scenario