# Product-market fit: what it means and how to assess it

> Assess repeat use, renewal and perceived value without relying on one threshold.


Product-market fit (PMF) is the match between a solution and a group of customers who need it, use it over time and are willing to pay for it. A survey, one percentage or a short sales spike cannot establish PMF on its own. Assess it by combining evidence of customer need, repeat use, retention or renewal, and the business’s ability to serve that market sustainably.

**Hypothetical example.** Of 12 users who answer a survey, five say they would be “very disappointed” if they could no longer use the product: 5/12, or 41.67%. In a separate measure, eight of 20 accounts are active in week eight (8/20 = 40%); six of the 20 renew (6/20 = 30%). Within a defined segment of ten customers, five renew (5/10 = 50%), compared with one of ten customers outside it (1/10 = 10%). These invented figures do not prove PMF. They point to a useful research question: do those ten accounts share a need and usage pattern that might explain the difference? Check more cohorts, and define “active” and the renewal period before drawing conclusions.

## What finding fit means

A product shows signs of fit when an identifiable market gets enough value to return, stay or pay to solve a problem. Bring four questions together:

- **Need:** What specific problem leads customers to seek a solution?
- **Use and repetition:** Do they return when that need comes up again?
- **Payment and renewal:** Do they pay, renew or continue under known terms?
- **Capacity and economics:** Can the business serve this group well and sustainably?

Stated responses help explain perceived value, but they do not replace behavior. Activity is not the same as retention: someone may open or try a tool and never come back. Renewal is a stronger commercial signal, but its meaning depends on the buying cycle, measurement window, discounts, contract terms and reasons for cancellation. No single metric captures the whole fit.

## Read the example without mixing denominators

Each percentage answers a different question and uses an explicit base:

| Signal | Calculation | Interpretation |
| --- | --- | --- |
| Survey | 5 / 12 = 41.67% | Perception among respondents; eight invitees did not respond. |
| Active in week eight | 8 / 20 = 40% | Activity in that window; it is not the same as renewal. |
| Overall renewal | 6 / 20 = 30% | Continuation within this cohort; it does not explain motives or profitability. |
| Renewal by segment | 5 / 10 = 50%; outside, 1 / 10 = 10% | A descriptive difference; it does not establish causation or a persistent pattern. |

Do not add or average these rates: they measure different things, use different bases and may cover different windows. Do not swap “accounts” for “users” without stating the unit. One account can include several people; if a survey counts people and renewal counts contracts, the denominators do not describe the same population.

The segment difference can guide further research. Check how the ten accounts were selected, whether cohorts started at comparable times, how large their market is, and whether channel, price or support differed. With ten customers in each group, one account moves the rate by ten percentage points. This is a descriptive clue, not a benchmark or a stable estimate.

## Sean Ellis’s 40%: a heuristic, not a certification

In his original July 2009 post, Sean Ellis described asking existing users how they would feel if they could no longer use the product. He suggested looking at respondents who would be “very disappointed” and said the 40% threshold was arbitrary, based on comparing results across nearly 100 startups. In the same post, he qualified growth by saying those users had to represent a sufficiently large target market. [Read the original post, “The Startup Pyramid.”](https://www.startup-marketing.com/2009/07/)

That context matters. The post does not establish 40% as a universal law, a causal experiment or a PMF certification. The example’s 5/12 is 41.67%, but a small denominator and eight nonresponses make it a poor basis for a verdict. Use the question to understand perceived value among people who already know the product, then compare it with use, retention, payments and market size.

## A practical process before scaling

1. **Define the segment and unit.** Describe who buys, in what situation and what need they share. Distinguish users, accounts and contracts.
2. **Specify the expected behavior.** Define what counts as “active,” the value event and the window in which it should recur. Keep the definition consistent across cohorts.
3. **Measure comparable cohorts.** Track activation, repeat behavior and retention from a consistent start date. State exclusions, period and account count.
4. **Compare intent with commercial behavior.** Read open-ended responses alongside payment, renewal, cancellation and purchase reasons. Stated intent is context, not a substitute for a transaction.
5. **Check capacity and economics.** Include support, delivery, cost to serve, margin and acquisition channels. A satisfied group may be too small or expensive to serve sustainably.
6. **Choose the next experiment.** If a cohort or segment looks stronger, investigate which need, behavior or channel sets it apart. Adjust the product, segment or service, then measure again before committing to scale.

The decision is not “PMF: yes or no” based on crossing a percentage. It is whether the evidence justifies more research and fit improvement, or supports a limited expansion test. Scaling spend before understanding retention, market size and cost to serve can amplify a misleading signal.

For researching what drives a purchase, see [Jobs to Be Done](/en/guides/jobs-to-be-done/). For planning a limited expansion, read about [go-to-market strategy](/en/guides/go-to-market-strategy/).

## Plan your next post in HeyMark.

Keep the idea, review the draft with your team, and see how it performed in the accounts you connected.

[Start free](https://app.heymark.ai) · [See how it works](https://heymark.ai/en/#product)

## Developer and agent resources

- [HeyMark MCP documentation](https://heymark.ai/en/mcp/)
- [llms.txt](https://heymark.ai/llms.txt)
- [Full site content for language models](https://heymark.ai/llms-full.txt)
- MCP protocol endpoint: `POST https://mcp.heymark.ai`
- OAuth protected-resource metadata: [/.well-known/oauth-protected-resource](https://mcp.heymark.ai/.well-known/oauth-protected-resource)
- MCP server card: [/server-card](https://mcp.heymark.ai/server-card)
