Analytics cohort: definition and retention analysis
Updated on February 22, 2026
Quick definition
A cohort is a group of users who share a common characteristic at a precise point in time — generally their acquisition or first-interaction date — whose behavior is tracked longitudinally to measure retention, churn rate, or generated value. The cohort is the central unit of analysis in any user-loyalty study.
How it works
In web analytics, a cohort is a group of users defined by a shared temporal characteristic. The most classic cohort groups users who made their first visit or sign-up during the same week or month. These users are then tracked over time to observe how many return the following week, the following month, and so on.
A cohort can also be defined on other criteria:
- Acquisition source
- Subscribed pricing plan
- First feature used
- Geographic segment
Cohort analysis is particularly powerful for SaaS products and mobile apps where retention is a critical business metric. It allows you to compare the loyalty of users acquired through different channels and measure the impact of product improvements on the retention of new cohorts.
Why it matters
Cohorts reveal a product's true long-term health that aggregated metrics conceal. A product can show growing user numbers while having catastrophic retention if acquisition offsets churn. Cohort analysis exposes that problem immediately.
It is also indispensable for computing CLTV, optimizing onboarding strategies, and assessing the real impact of product improvements on long-term user behavior.
How to improve or use it
- 1Precisely define the cohort-entry event (first visit, first sign-up, first purchase) and the retention event.
- 2Plot retention curves by monthly cohort to observe macro trends.
- 3Compare cohorts by acquisition source to identify channels generating the most loyal users.
- 4Use these insights to adjust your onboarding and re-engagement strategy.
- 5Measure the impact of each product improvement by comparing retention curves of cohorts before and after the change.
With Sublim
Sublim makes it possible to build cohort analyses based on custom events, without cookies and in GDPR compliance. By hosting data in Europe, Sublim guarantees precise and ethical retention analysis — ideal for SaaS teams driving growth through retention rather than acquisition at any cost.
Frequently asked questions
What is the difference between a cohort and a segment?
A segment is a group of users defined by static criteria (country, device, current behavior). A cohort is a group defined by a shared temporal characteristic, whose evolution is followed over time. The cohort is dynamic in that it measures how the group's behavior evolves longitudinally.
What is a good retention rate for a cohort?
Benchmarks vary considerably by industry. For a B2B SaaS application, a 3-month retention rate above 70% is generally good. For a consumer mobile app, exceeding 30% at 30 days is already strong. Comparing your own cohorts over time is often more relevant than industry benchmarks.
How do you read a cohort table?
A cohort table presents cohorts in rows (for example, each acquisition month) and retention periods in columns (week 1, week 2, etc.). Each cell shows the percentage of users from that cohort still active in that period. The diagonal of the table lets you compare the same retention stage across different cohorts.
Related terms
Cohort analysis is an analytical method that groups users sharing a co…
Churn rate (or attrition rate) is the percentage of customers or subsc…
CLTV (Customer Lifetime Value) is the customer lifetime value represen…
A segment is a filtered subset of your visitors or sessions, defined b…