Product market fit is the most cited concept in startup advice and is defined imprecisely enough that founders disagree about whether they have it.
The original framing
Being in a good market with a product that can satisfy it.
Which is descriptive rather than measurable.
The observation was that you can feel it happening rather than needing to measure it.
Behavioural signals
Usage growing without proportional marketing, retention flattening rather than declining, and customers referring others.
Which are observable in data.
Retention curves flattening at a meaningful level is the most commonly cited indicator.
The retention curve
The proportion of a cohort still using a product over time.
Which declines and either flattens or approaches zero.
A curve that flattens indicates a group for whom the product is genuinely useful.
Survey approaches
Asking how disappointed users would be if the product disappeared.
Which produces a threshold proportion cited as an indicator.
It is a heuristic rather than a validated measure and is widely used anyway.
Market matters more than product
The original argument was that market pull dominates product quality and team.
Which is contested and has substantial supporting anecdote.
Excellent products in markets that do not want them fail regularly.
Fit is not permanent
Markets change and competitors arrive.
Which means fit can be lost.
Companies that achieved it once are not immune to losing it.
Premature scaling
Investing in growth before fit is established.
Which is identified in startup failure research as a leading cause.
Spending to acquire customers who churn is a reliable way to run out of money.
What to actually watch
Cohort retention, organic growth and whether customers would be genuinely inconvenienced without you.
Measuring retention properly
Cohort analysis tracking each group of new users over time.
Which reveals whether the product retains anyone.
Aggregate active user counts hide a churning base entirely.
Defining the action
What counts as a retained user depends on the product's natural frequency.
Which differs enormously between categories.
Weekly measurement for a monthly product produces misleading curves.
Segment fit
Fit within a specific segment before broader appeal.
Which is the pattern most successful companies describe.
Looking at retention by segment frequently reveals fit that aggregate data hides.
Qualitative signals
Customers using the product in unexpected ways and objecting when it breaks.
Which is informative and does not appear in dashboards.
What to do before fit
Talk to users, ship changes quickly and resist hiring ahead of evidence.
Sales-led signals
Shorter sales cycles, fewer objections and inbound enquiries.
Which indicate pull rather than push.
Sales teams describe the change as things becoming noticeably easier.
Pricing power
Ability to raise prices without losing customers.
Which is a strong indicator of genuine value delivery.
Customers who would not switch away at a higher price are demonstrating something.
Word of mouth
Customers referring others without being asked.
Which is measurable through source attribution.
It is also the cheapest acquisition channel available.
Losing fit
Competitors, market shifts and changing customer needs.
Which means monitoring continues after the milestone.
The practical summary
Cohort retention flattening at a meaningful level, growing organic acquisition, and customers who would be genuinely inconvenienced without you.
The retention curve
What percentage of a cohort is still active after several months.
Which either flattens at some level or continues declining toward zero.
A flattening curve indicates a group of people for whom the product genuinely works.
Usage depth
How frequently and how thoroughly active users engage.
Which distinguishes habitual use from occasional use.
Depth generally predicts retention better than any survey response.
Survey measures
Asking how disappointed users would be if the product disappeared.
Which produces a rough signal and is frequently over-interpreted.
The threshold commonly quoted has no strong empirical basis.
Segment-specific fit
Fit frequently exists for a narrow group before it exists broadly.
Which is useful information about where to concentrate.
Averaging across all users hides it.
Why the concept is slippery
It is described as something you know when you feel it.
Which is unhelpful as a decision criterion.
Measurable proxies are imperfect and are considerably better than intuition.
Engagement against acquisition
Growing sign-ups with flat engagement is not fit.
Which paid acquisition can disguise for a long time.
Turning acquisition spending off is a blunt and revealing test.
Qualitative signals
What customers say when asked what they would use instead.
Which is more informative than satisfaction scores.
Answers involving spreadsheets and manual work indicate genuine need.
Premature scaling
Hiring and spending before the signals are present.
Which is the most expensive misdiagnosis available.
The cost of waiting an extra quarter is generally much lower.
Measuring it consistently
Pick a definition, apply it every month and watch the direction.
Which is more useful than any single reading.
Changing the definition to make the number look better is a recognisable pattern.