Sign in

Retention

Who comes back,
and what brings them back

Group people by when they started, then follow each group period by period. Count a return as any visit or as the one behavior that matters, and split it by who people are to find the users who stay.

In this demo car app, about seven in ten new drivers return in their first week. The curve shows where the drop stops and the regulars begin.

Part of analytics

One question,
three honest answers

“Did they come back?” depends on what you count. Pick the reading that fits the decision. The people in each cohort stay the same; only the way a return counts changes.

  1. Did they come back that week?

    Classic

    active in that period, gaps allowed

    The everyday reading. Someone who skips a week and returns the next still counts for the week they came back.

  2. Did it become a habit?

    Full

    active in every period, without a break

    The strict reading. Only people who never missed a period count, so what is left are your real regulars.

  3. Who could we still win back?

    Unbounded

    active in that period or any later one

    The forgiving reading. It counts everyone who ever returned, which sizes the audience a win-back campaign can still reach.

From curve to decision

Four questions a retention
review should answer

Retention is one number in a deck and four questions in a planning meeting. Each has its own view, on one page.

Numbers you can defend

Ready for the questions
people ask in the meeting

A retention number is only useful if nobody can talk it down. The page answers the usual objections before anyone raises them.

“The newest week always looks terrible.”

Young cohortsA cohort that hasn’t reached week 8 has no week 8 value. The cell stays empty instead of counting as zero, so recent weeks never read as a collapse.

“One small week is skewing the average.”

Weighted curveWeighted lets larger cohorts count for more, and every row shows its size first, so a flattering rate from a handful of people is easy to spot.

“Is this better than last quarter?”

Comparison windowPick an earlier window and it appears behind the curve and the trend, and beside every cell in the grid.

“Can we keep this view?”

Shortcuts and dashboardsSave the query as a shortcut that always opens on the latest cohorts, put it on a dashboard, or export the grid.

Works with

Retention, with the
rest of Countly

Every feature reads the same live data, so an answer in Retention doesn’t stop there. These are the handoffs built in.

All features
Cohorts

Filter retention to a cohort, or split it by cohort to compare audiences side by side.

Open Profiles

Filter and split by profile properties and profile groups, your own custom properties included.

Open Dashboards

Add the query on screen to a dashboard as a retention tile.

Open

Data ownership

Every return counted
on data that stays yours

Retention follows each person over weeks and months. That history is computed where it is stored, never copied to another service to be measured.

Data privacy
AspectFlexEnterprise
EnvironmentDedicated single-tenant instanceYour cloud, your data center or air-gapped
Data residencyThe Google Cloud region you chooseWherever you deploy, including air-gapped networks
Vendor accessOperated by Countly; access limited, granted by you and loggedNo default access
ISO 27001Certified
ISO 27017Certified
ISO 27018Certified
SOC 2Type II examined
TISAX®Participant*

*Countly Ltd. is a TISAX® Participant. TISAX® is a registered trademark of and governed by ENX Association. Learn more at enx.com/tisax

Try it

Every feature.
One data layer.

Start free trialAll features