Glossary › Analytics › Retention Curve
Analytics

Retention Curve

A retention curve is worth a thousand NPS scores. It shows you not just whether users are satisfied, but whether they're actually staying, and exactly when they leave. Learning to read a retention curve is one of the most valuable analytical skills in product management.

Updated October 7, 2026

Retention Curve: definition

A retention curve is a line chart that shows what percentage of a group of users (a cohort) is still active as time passes since they signed up. It starts at 100% on day 0 and falls; whether it keeps falling or flattens out is the quickest read on whether a product keeps its users.

Also called a retention chart, a retention graph or a cohort retention curve.

What is a Retention Curve?

A retention curve (also called a retention chart or cohort retention graph) plots the percentage of users from a cohort who remain active over time. The x-axis shows time since acquisition (day 1, day 7, day 30, day 90), and the y-axis shows the percentage of the original cohort still active at each point.

A 100% retention rate at day 0 (everyone just signed up) drops over time as users churn. The shape and terminal level of that curve tells you most of what you need to know about product-market fit, onboarding effectiveness, and long-term retention health.

If you have a statistics background, a retention curve is a close relative of the survival curve, the chart medical researchers use to show what share of a group remains after a given time, with "still active" standing in for "still surviving". Each point on it is a retention rate: the users from the cohort who were active at that moment, divided by everyone who started.

How to Read Retention Curves

Read a retention curve in this order; each step answers a different question:

  1. The first week. How far does the line fall between day 0 and day 7? This is the stretch onboarding controls most directly.
  2. The bend. Where does the drop start to slow? The earlier the curve bends, the sooner new users find a reason to come back.
  3. The floor. Does the line flatten above zero, and at what level? That flat stretch is your retained core.
  4. The comparison. Does the newest cohort's line sit above the older ones at the same age? One curve is a snapshot; the trend from cohort to cohort is the verdict.
ANATOMY OF A RETENTION CURVE 100% 75% 50% 25% 0% Share of the cohort still active Day 0 Day 1 Day 7 Day 14 Day 30 Day 60 Day 90 Days since sign-up FIRST WEEK shaped by onboarding Day 0: everyone in the cohort is active Flattens the retention floor Smiles users come back Trends to zero no lasting users Illustrative curves, not benchmarks

Three shapes a retention curve can take. The first week is mostly decided by onboarding, and the level where a curve flattens is the retention floor.

The shape of the curve is the signal:

  • Curves to zero: retention drops to 0% over time, meaning you have no lasting users. This is a product-market fit problem, not an onboarding problem.
  • Drops to a floor and flattens: retention drops steeply then levels off at a consistent percentage. This is a healthy retention curve. The floor level (10%, 20%, 40%) is your stable active user base.
  • Smiling curve: retention drops, then recovers. Usually indicates re-engagement campaigns working, or seasonal patterns in usage. Rare but very positive.
  • Sharp early drop-off: most users leaving in the first few days signals an onboarding or first-impression problem: users aren't reaching the aha moment before they leave.

For benchmarks, retention modelling and the fixes that bend each part of the curve, see our complete guide to reading and bending retention curves.

Retention Curve Terms

The words used around retention charts, in one place:

TermWhat it means
Day 0The day each user entered the cohort. It is always 100%, because everyone in the cohort was active on the day they started.
N-day retentionThe share of the cohort active on exactly day N. Day-30 retention is the most quoted example: the share of a cohort active on its 30th day.
Bracket (rolling) retentionThe share active at least once inside a window around day N, such as week 4. It smooths out the noise of products people don't use every day.
Unbounded retentionThe share active on day N or on any later day. It can only fall, so it reads higher and smoother than N-day retention.
Retention floorThe level where the curve stops falling and runs flat. The users above it are the cohort's retained core.
Smiling curveA curve that falls and then rises because inactive users come back. N-day and bracket curves can smile; an unbounded curve cannot.
Cohort retention curvesOne curve per cohort drawn on the same axes, so cohorts can be compared at the same age.

Which definition you choose changes the number you report, so always state it next to the chart. Our user retention guide compares the retention types in detail, and the cohort entry explains how the groups behind each curve are built.

Retention Curves and Onboarding

The first 7 days of the retention curve are almost entirely determined by onboarding. A steep drop in days 1–7 means new users aren't activating: they're not reaching the moment of value before they give up and stop returning.

The most impactful way to shift a retention curve is to shift the day-1 and day-7 numbers. Every improvement to onboarding (faster time to value, better interactive guidance, a more targeted first-run experience) lifts the early segment of the curve, which compounds into dramatically different 90-day retention.

That first-week zone is where in-app guidance earns its keep. A short product tour that hands over to a small onboarding checklist helps new users reach first value in their first session, and Kompassify's no-code product analytics reports activated, engaged and churning users, so you can see whether the change moved the curve.

Segmenting Retention Curves

Aggregate retention curves hide more than they reveal. The most useful analysis compares curves across segments:

  • Acquisition channel: do users from organic search retain better than paid? Different channels often bring users with different intent and expectations.
  • Onboarding completion: compare retention of users who completed key onboarding steps vs those who skipped. This quantifies the ROI of your onboarding investment.
  • Plan type: free users often have very different retention dynamics from trial or paid users.
  • Persona or segment: retention curves by company size or job role often reveal which market segments truly have PMF.

Go deeper on the Kompassify blog

Longer guides on the ideas in this entry.

GuideWhat Is a Retention Curve? Definition, Chart & ExamplesThe three curve shapes with examples, and how to make a retention chart in four steps. GuideRetention Curves: How to Read, Benchmark & Bend ThemThe cliff, the slope and the asymptote, benchmarks, retention modelling and the fixes for each part. MetricsUser Retention: Definition, Formula & Good BenchmarksN-day, rolling and unbounded retention compared, with the formula behind every point. PlaybookHow to Increase User Retention in SaaSThe practices that lift the curve once you know where it loses people.

Frequently Asked Questions

What is a retention curve and how do I read one?
A retention curve is a line chart showing what percentage of a cohort of users is still active over time since sign-up, starting at 100% on day 0. Read it in four steps: how far it falls in the first week, where the drop starts to slow, whether it flattens above zero and at what level, and whether newer cohorts sit above older ones at the same age. A curve that flattens is healthy; one that keeps sliding toward zero means users are not finding lasting value.
What is day-30 retention?
Day-30 retention is the share of a cohort that is active on the 30th day after signing up. If 1,000 users signed up in a week and 280 of them performed your return event on their day 30, day-30 retention for that cohort is 28%. It is one point on the retention curve, which is why the full curve tells you more than the single number.
What is a good retention rate at day 30?
It depends heavily on the product category, the price point and how often the product is meant to be used, so a generic benchmark makes a weak target. The comparison that matters most is your own trend: is each new cohort's day-30 point higher than the last one? If it is, your onboarding and product changes are working.
What is a retention floor?
The retention floor is the level where a retention curve stops falling and runs flat. The users above it are the cohort's retained core: they have found enough value to keep coming back. Raising the floor usually means getting more new users to that value earlier, which is why onboarding changes show up there.
Why does my retention curve drop so much in the first week?
Because the first week is when users decide whether the product is worth a second visit, and many of them decide before they have reached its value. A steep first-week drop is usually an activation problem rather than a product problem: shorten the path to the first useful result and guide the step where new users stall, and the early part of the curve, along with everything after it, moves up.
What tools can you use to build retention curves?
A spreadsheet works for a first chart: group users by sign-up week, count how many are active in each later week, and divide by the cohort size. To keep the curve live you need a product analytics tool that groups users into cohorts and tracks a return event over time. Kompassify's no-code product analytics reports activated, engaged and churning users alongside the product tours, checklists and surveys you use to act on what the curve shows.
How long should you track a retention curve?
For most SaaS products, 90-day retention is the most actionable window: it reflects real user behaviour without being distorted by too many confounding factors. 12-month retention is important for revenue forecasting and LTV modelling, but less useful for day-to-day product decisions.

Improve the most important part of your retention curve

Kompassify helps SaaS teams flatten the day-1 and day-7 drop-off with product tours, onboarding checklists and in-app announcements, then shows the effect in no-code product analytics.

Start for free →

Free up to 100 monthly active users · plans from $129/month · GDPR-compliant and EU-hosted

Keep exploring