Cohort analysis

Cohort analysis is a type of analytics that looks at the behavior or performance of groups of people (cohorts) who share a common characteristic or experience. For example, a cohort could be people born in the same time period or those who started schooling in a specific year.

In ChartMogul, a cohort is a group of customers who start their first subscription in the same interval (day, week, month, quarter, or year). Cohort analysis allows you to better understand churn, retention, and conversion rates based on when customers subscribe and identify points in your customer lifecycle that result in contraction and churn.

Here’s what we cover in this article:

Resources and further reading:

Before you begin

Read more about how ChartMogul organizes and classifies MRR movements.

How cohorts are useful

To grow a SaaS business, you need customers to continually renew their subscriptions. So it’s important to identify points in your customer lifecycle that result in downgrades (contraction) or cancellations (churn).

Instead of looking at the overall churn rate for all customers, cohorts allow you to compare groups of customers based on when they subscribe so you can answer questions like:

Which month in the lifespan of a subscription is churn at its highest?
How did last month’s product release impact churn?
How is retention for customers who subscribed in February during our Valentine’s marketing campaign?
How is retention for customers who signed up using a discount?

Reading a cohort table

The results of a cohort analysis are visualized using a table that may be different from other charts and reports you're familiar with and can be a challenge to read and understand.

Screenshot of a cohort table with numbers for each element explained here.

The cohort analysis you select — and how you configure it — determines the calculation of Cohort Value and the metric ChartMogul displays in table cells.

Columns

The first and second columns (1) display each cohort’s name (identified by the time of conversion, e.g., Feb 2023) and its value. The remaining columns (2) represent each interval of the cohort’s existence (day, week, month, quarter, or year, depending on the cohort type).

Rows

Each row (3) is an individual cohort. Cohorts appear in chronological order. The last row (4) provides an average of the relevant metric across cohorts for each interval.

Cells

Cells (5) show the relevant metric depending on the cohort analysis you select (e.g., churn, expansion, contraction or reactivation) and how you configure it. Future intervals (6) are blank.

Colors

ChartMogul assigns a color to each metric to help you interpret the results of your analysis. Green indicates the metric with the best value (i.e., lowest churn or highest retention rate) and red the worst. The metric falling in the middle is yellow. The remaining metrics are colored using shades along a green-yellow-red gradient to indicate their value relative to the best (and worst) metrics in your current analysis.

Cohort curves

Every cohort report includes a chart that draws each cohort as a curve, making it easier to compare how cohorts evolve over time. The chart reflects the settings you choose for the report.

Use the checkboxes in the cohort table rows to choose which cohorts appear on the chart. Shift+click a row to select or deselect a range at once. Select the Average row checkbox to overlay the weighted average curve.

The chart is visible by default. Click the HIDE CHART button to hide it.

Screenshot of the cohort curves chart for Paid Subscriber Churn, shown above the cohort table with each cohort drawn as a separate line.

Cohort analyses in ChartMogul

ChartMogul offers the following cohort analyses, which you access by navigating to Reports > Cohorts.

Churn

Customer Churn — the percentage or number of customers who have canceled all of their subscriptions, offset by reactivation.
Net MRR Churn — the percentage or amount (in your primary currency) of MRR lost due to cancellations, offset by expansion and reactivation.
Quantity Churn — the percentage or number of subscription downgrades or cancellations, offset by expansion and reactivation.

Retention

Customer Retention — the percentage or number of customers who have one or more active subscriptions, including reactivation. Learn more.
Net MRR Retention — the percentage or amount in your primary currency of MRR from active subscriptions, expansion, and reactivation, minus churn and contraction. Learn more.
Quantity Retention — the percentage or number of subscriptions still active, including expansion, churn and contraction.

Conversion

Conversion of Non-Subscription Customers to Subscribers — the number or percentage of customers who previously made a non-recurring purchase and then purchased a subscription in the selected month. (Available on our Scale and Volume plans)

Configuring your analysis

Type

Use the Type drop-down to select an interval for your analysis:

Daily cohorts
Weekly cohorts
Monthly cohorts (default)
Quarterly cohorts
Annual cohorts

Screenshot of the configuration options for a cohort report: Type, Start, Show, Relative to, and Include current month. The Type drop-down open and shows Daily, Weekly, Monthly, Quarterly, and Annual cohorts.

Start

Select a time frame for your analysis using the Start date picker. Depending on your Type setting, select the starting day, week, month, quarter, or year.

Show

Use the Show drop-down to choose Percent or, depending on the cohort analysis you select, Customers (labeled Paid Subscribers for accounts with free subscription support), MRR, Quantity, or Conversion.

Relative to

If you select Percent from the Show drop-down, you can use the Relative to drop-down to choose Previous day/week/month/quarter/year or Starting day/week/month/quarter/year.

With Previous day/week/month/quarter/year, ChartMogul calculates the metric as a rate relative to the previous interval. With Starting day/week/month/quarter/year, ChartMogul calculates it relative to Cohort Value.

Include current day/week/month/quarter/year

Select Include current day/week/month/quarter/year to get the most up-to-date insights by including data for the current interval in your analysis.

Understanding the data

Cohorts

A customer joins a cohort when they become a subscriber for the first time, i.e., when their status changes from one of the lead statuses (New Lead, Working Lead, Qualified Lead, or Unqualified Lead) to Active Subscriber. Each customer remains in their original cohort regardless of whether they sign up for or purchase a second subscription (expansion) or cancel (churn) and re-subscribe (reactivation).

Intervals

Follow the development of each cohort by interval (day, week, month, quarter, or year), starting with interval 0 (when the customers subscribed) and continuing for each interval of the cohort’s existence.

Let’s take a monthly cohort analysis as an example. For a customer who subscribed on April 19, month 0 is April 19–April 30, month 1 is May 1–May 31, month 2 is June 1–June 30, and so on.

While the actual dates for a specific month (e.g., month 3) are different between cohorts, what makes them valuable from an analysis perspective is their distance (in months) to sign up. Knowing, for example, that churn peaks in month 3 is an important insight.

Cohort Value

Depending on the cohort analysis you select, Cohort Value represents one day, week, month, quarter, or year of:

Total MRR (in your primary currency)
Number of paid subscribers
Number of subscriptions acquired

In the following example, the May 2025 cohort had 42 paid subscribers.

Screenshot of a cohort analysis of the percentage of paid subscribers churned relative to the previous month with the Cohort Value column highlighted.

Metrics

Depending on the cohort analysis you select, ChartMogul displays one of the following metrics in table cells:

Conversion — the number of non-subscription customers who started a subscription
Customers — the number of active customers after churn, expansion, and reactivation
MRR — the amount of MRR after churn, expansion, and reactivation
Quantity — the number of active subscriptions after churn, expansion, and reactivation
Percent — the percentage in change, either relative to the previous month or the starting month

In our example, 9.52% of paid subscribers in the May 2025 cohort had churned in month 1.

Screenshot of a cohort analysis of the percentage of paid subscribers churned relative to the previous month, with the cell for month one of the May 2025 cohort showing 9.52%.

Select the cell to see a complete list of churn activities.

Screenshot of the Activity table containing a list of MRR movements, each with a description, value, type, and date.

Average

The Average row shows the average Cohort Value across all cohorts, and the average of the selected metric across all cohorts for each interval. ChartMogul accounts for cohort size (i.e., the number of customers) when calculating these averages.

For example, in a given month, a cohort of two customers experiences 50% churn while a cohort of 100 experiences 25%. The average churn rate (across the two cohorts, weighted by the number of customers) is 25.49% (total number of customers lost / total number of customers) and not 37.5% (the average of the two percentages).

Learn more about weighted averages.

Exporting your analysis

Export your cohort analysis by selecting the Export as CSV button.

Screenshot of a cohort analysis table with the Export as CSV button highlighted.

Segmenting cohorts

Using segmentation as part of your cohort analysis in ChartMogul helps you answer even more complex questions like:

When does a particular pricing plan experience its highest churn?
Which sales representative has the highest retention rate?
How does NPS score correlate with churn rate?