What is retention rate?
Retention rate is the share of users still active after a period. The formula, N-day versus rolling retention, and how cohorts make it readable.
What is retention rate?
Retention rate is the share of users from a starting group who are still active at the end of a defined period. It is measured per cohort, usually the people who signed up in the same week or month, and reported at intervals such as day 1, day 7 and day 30. It is the inverse of churn rate.
The formula.
Retention rate = (users active at end of period / users in the starting cohort) x 100
- starting cohort
- Everyone who joined in the same window, for example all signups in the week of 3 August.
- active
- Your product's definition of a meaningful action, not merely logging in. State it explicitly.
New users acquired during the period never enter this calculation. Retention is measured within a fixed cohort, which is what stops growth from hiding a leak.
A worked example.
The August signup cohort of a SaaS product.
| Signups in the cohort | 1,240 |
|---|---|
| Still active on day 7 | 471 |
| Still active on day 30 | 298 |
| Day 30 calculation | 298 / 1,240 |
Day 7 retention = 38%, day 30 retention = 24%
The steep part is between signup and day 7, which points at onboarding rather than at long-term value. Fixing the first week moves every later number with it.
N-day, unbounded, and rolling retention
Three definitions circulate under one name, and they give materially different numbers on the same data.
- N-day retention: was the user active on exactly day N? Strict, spiky, and the right choice for products people are meant to use daily.
- Unbounded retention: was the user active on day N or any day after? More forgiving and better for weekly or monthly products.
- Rolling or bracket retention: was the user active in a window around day N, for example days 5 to 9? Smooths out weekday effects.
Pick the one that matches how often your product is genuinely meant to be used. A B2B reporting tool used every Monday will look catastrophic under N-day retention and healthy under a weekly bracket, and the second reading is the truthful one.
Why retention is read as a cohort grid
A single retention percentage is an average across cohorts that were acquired differently, onboarded differently and shipped different versions of the product. The grid separates them.
Read it two ways. Along a row: how one cohort decays over time, which tells you whether the product holds people. Down a column: how the same week-since-signup number moves across successive cohorts, which tells you whether recent changes to onboarding or acquisition are working.
The shape that matters most is whether the curve flattens. A curve that decays and then goes horizontal means you have found a group for whom the product is genuinely durable. A curve that keeps sliding toward zero means nobody sticks, and no amount of acquisition spend will fix it.
See it on real data
The live demo is a fully populated dashboard with real demo traffic. No signup and no card required.
Open the live demoGetting retention right
Define active honestly
Counting logins overstates retention badly. Count the action that delivers the value: a report run, a message sent, a deploy shipped.
Never mix acquisition into it
If new signups enter the denominator, growth masks decay. Retention is measured inside a closed cohort for exactly this reason.
Match the interval to the product
Daily for consumer habit products, weekly for team tools, monthly for anything invoiced monthly. Using the wrong interval makes a healthy product look broken.
Look for the flattening point
The week the curve stops falling is your real retained base. Everything above it is people who were never going to stay.
Retention in Mrkr
Mrkr builds cohorts from any tracked event, so the starting group can be signups, first purchases, or whoever completed onboarding, and the active definition can be any later event you care about.
Cross-session retention needs identity that persists beyond a day, which cookieless mode deliberately does not provide. For retention over weeks you either enable Mrkr's optional per-site cookie mode, which does require a consent banner for that site, or you send a stable user id with your events from your own backend once someone has an account.
That trade-off is worth stating plainly rather than hiding: a cookieless identifier that expired daily could not honestly report 30-day retention, and a tool that claimed otherwise would be measuring something else.
Questions, answered.
Related terms.
- What is churn rate?
Churn rate is the share of customers, users or revenue lost during a period.
- What is cohort analysis?
Cohort analysis groups users by something they share, usually the week or month they signed up, then tracks each group separately over time.
- What is dau/mau ratio?
DAU/MAU is daily active users divided by monthly active users, expressed as a percentage.
- What is north star metric?
A north star metric is the single number a company chooses to represent the core value its product delivers to customers.
- What is event?
An event is any recorded action beyond loading a page: a button click, a form submitted, a video played, a subscription upgraded.
Where this shows up in Mrkr
- See what Mrkr measures in retention analysis
Stickiness, new versus returning, and a straight answer on what needs cookie mode.
- Group users into cohorts with Mrkr cohort analysis
- Browse the full analytics glossary
Every metric, method and privacy term, defined in one place.
Your first visitor is already here.
Drop in the script and watch them land. It takes about a minute.