User Retention
Share of users active at period start who are still active at period end, measured as ending active users divided by starting active users.
◆ Ratio
Formula
Built from
What it measures
User Retention tells you what fraction of your user base stayed active from one period to the next. It is the raw all-in ratio — before you carve out churn, downgrades, or expansion — and the denominator is who you started the period with, not the average or the ending count.
Why it matters
User Retention is the simplest read on whether your product keeps the people it has. Ops uses it to tune onboarding and support, product uses it to find where friction lives, and execs use it to model runway and the acquisition pressure needed just to stand still. A high retention number means growth compounds on a stable base; a low one means you are constantly refilling a leaking bucket. It answers one blunt question: did we keep our users, or did we have to replace them?
How to read it
Read User Retention as a trend across periods, not a single snapshot, and always pin down what "active" means before you trust it. A value of 1.0 means you held steady — new activations exactly offset losses. Below 1.0 you are shrinking; above 1.0 the gross count grew, but that can be new signups papering over heavy churn underneath, so it is not proof of stickiness. Compare against Gross User Retention (GUR) to separate real, mechanical user loss from net movement, and benchmark against your own cohort curves and stage rather than a single industry number.
What good looks like
Good
User Retention holds high and steady or climbs across consecutive periods, with the gross gains driven by genuine activation rather than churn being masked by new signups.
Watch
Retention drifting down period over period, or sitting flat only because heavy new activation is offsetting rising churn — check Gross User Retention plus onboarding and support signals to find the friction.
Bad
Retention falling well below your historical band, signaling systemic problems — weak product-market fit, broken onboarding, pricing pressure, or a competitor pulling users away — and warranting an immediate root-cause audit.
Watch-outs
- Using the wrong denominator. User Retention divides by opening active users, not the average or the closing count — a late-month spike in signups can inflate the ratio if you anchor it to the wrong base.
- Leaving 'active' undefined. Logins, API calls, paid seats, and not-yet-deleted accounts give wildly different numbers — lock the definition to your product model and apply it consistently every period.
- Comparing cohorts born in different months. Churn curves steepen early and flatten late, so a young cohort and an old one are not comparable at the same calendar point — use cohort-aligned retention curves instead.
- Reading any value above 1.0 as growth. It only means gross activations outnumbered net departures; heavy churn can hide beneath it, so confirm with Gross User Retention, Net User Retention, or your actual logo count.
Worked example
Hypothetical
You close August with 1,200 active users, so you reset starting active users to 1,200 on September 1. By September 30 you have 1,100 active users. User Retention is 1,100 / 1,200 = 0.917, or 91.7%. If September brought 250 new activations, churn must have been roughly 350 to net the 100-user decline.