TG3 47 SaaS brands scaled and $84M+ in client pipeline generated. See the proof → Free: 20 SaaS calculators, no signup. CAC, LTV, churn, Rule of 40. Open the tools → We rebuild attribution at the warehouse so every channel gets honest credit. See how →
TG3 SaaS/Glossary/Cohort analysis
SaaS metrics glossary

Cohort analysis.

The single best way to see whether your product is actually getting stickier. Here is what cohort analysis is and the retention truths it reveals that averages hide.

Definition
Cohort analysis groups users by when they joined and tracks each group over time, so you can see retention and behaviour that blended averages hide.

A blended retention number lies by mixing new and old users together. Cohort analysis pulls them apart, so you can see whether the users you acquired last month retain better than the ones from a year ago. It is how you tell real improvement from a flattering average.

How cohort analysis works

How cohort analysis works.

Group by start date  →  track behaviour over time
Cohorta group of users who started in the same week or month
Trackinghow each cohort retains, expands or churns across following periods

Read cohorts to judge whether your retention work is landing, see net revenue retention.

Benchmarks

What cohort analysis reveals.

The shape of the curve is the insight. A retention curve that decays to zero means no product market fit. One that flattens means a core of users who stick. Whether newer cohorts flatten higher than older ones tells you if your product is genuinely improving.

Blended metrics cannot show any of this. They average a great recent cohort with a poor old one and call it stable. Cohort analysis is the only honest way to see whether the changes you ship actually move retention.

How to improve it

How to use cohort analysis well.

01

Watch the curve shape

A flattening curve means stickiness. One decaying to zero means a fit problem.

02

Compare new to old cohorts

Newer cohorts retaining better than older ones is proof your product is improving.

03

Cohort by acquisition source

Some channels bring users who stick. Cohorts reveal which ones are worth more.

04

Tie changes to cohorts

Judge a retention initiative by the cohorts after it, not a blended average.

Common questions

Questions about cohort analysis.

What is cohort analysis?+

Grouping users by when they started and tracking how each group behaves over time, to reveal retention that averages hide.

Why is cohort analysis better than blended metrics?+

Because blended numbers mix new and old users and hide the truth. Cohorts separate them so you can see whether retention is actually improving.

What does a cohort retention curve tell you?+

A curve that flattens means a sticky core and likely product market fit. One that decays to zero means a fit problem.

How do you use cohort analysis for marketing?+

Cohort by acquisition source to see which channels bring users who stick and judge retention changes by the cohorts that follow them.

Net revenue retentionActivation rateProduct market fitChurn reduction

Trusting a blended retention number?

The 30-minute audit includes whether your cohorts say your retention is improving or just averaging out. No sales sequence.

Book the 30-minute audit →
Response inside 4 business hours · We turn down 1 in 3