How to calculate churn rate (and which formula to use)
Calculate churn rate: the formula, what it tells you about product health, and why it matters more than growth for sustainable businesses.
How to calculate churn rate (and which formula to use)
Quick answers
What is churn rate? Churn rate is the percentage of customers (or revenue) lost in a given period. It's one of the most important metrics in SaaS and one of the most inconsistently calculated — two teams at the same company can produce different numbers from the same data, both technically correct.
What is the difference between customer churn and MRR churn? Customer churn counts the percentage of customers who stopped paying — it treats a $10/month customer and a $10,000/month customer the same. MRR churn measures the percentage of revenue lost. A company losing five small customers while expanding three large ones could have worsening customer churn and improving MRR churn simultaneously.
What is the formula for customer churn rate? Customer Churn Rate = (Customers lost in period / Customers at start of period) × 100. Measure it monthly for most SaaS products. For annual contract products, measure at the renewal period — monthly churn on annual contracts is misleading.
What is net MRR churn and why does it matter? Net MRR churn = (MRR lost from cancellations and downgrades − MRR gained from expansion) / MRR at start of period. It can be negative, which is what you're aiming for. Negative net churn means existing customers are expanding faster than others are leaving — that's the engine of compounding SaaS growth.
Why does my churn rate look fine but the business doesn't feel fine? Because aggregate churn is an average — and averages hide variation. A 5% monthly churn could be 2% from enterprise customers and 12% from SMB, or 1% from long-tenured users and 20% from users in their first 30 days. The number looks acceptable; the breakdown shows a problem. Break churn down by cohort, plan, and segment in Metabase using dashboard filters.
How do I make sure everyone uses the same churn definition? Define it once in Metabase Data Studio — including which formula, which window, and what counts as churned. Every dashboard that references the churn metric draws from the same definition. The CEO and the product team get the same number.
Where can I find a SQL template for calculating churn rate? The Metabase LTV with SQL guide includes a churn rate CTE you can adapt directly to your schema. The Metabase retention analysis event shows how to break churn down by cohort inside Metabase.
---
Churn rate is one of the most important metrics in SaaS and one of the most inconsistently calculated. Two teams at the same company can produce different churn numbers from the same data — both correct by their own definition.
Here's how to calculate it, which version to use, and how to make sure everyone's looking at the same number.
The two types of churn
Customer churn: The percentage of customers who stopped paying in a given period. Also called logo churn.
Revenue churn (MRR churn): The percentage of revenue lost in a given period from cancellations and downgrades, after accounting for expansion from existing customers.
These tell different stories. A company losing five small customers while expanding three large ones might have positive revenue churn (net revenue retention) and negative customer churn. Both numbers are true. Both matter.
The customer churn formula
Customer Churn Rate = (Customers lost in period / Customers at start of period) × 100
The MRR churn formula
Gross MRR Churn = (MRR lost to cancellations + downgrades) / MRR at start of period × 100
Net MRR Churn = (MRR lost − MRR gained from expansion) / MRR at start of period × 100
Net MRR churn can be negative — which is what you want. Negative net churn means your existing customers are expanding faster than others are churning. That's the engine of compounding SaaS growth.
The Metabase learn guide on calculating LTV with SQL walks through a churn rate CTE you can adapt directly for your schema.
The measurement window problem
Churn rate is only comparable if measured over the same window. Monthly churn and annual churn are not the same — a 2% monthly churn compounds to roughly 22% annual churn. Make sure you're consistent.
For most SaaS products, monthly churn is the standard. For products with annual contracts, monthly churn measurement can be misleading — measure at the renewal period instead.
What churn hides
An overall churn rate is an average — and averages hide variation. A 5% monthly churn could be:
- 2% from your enterprise customers and 12% from your SMB segment
This is why cohort analysis matters. Looking at churn by acquisition month, plan type, or user segment reveals patterns that the aggregate number obscures. The Metabase retention analysis event covers exactly this — breaking down retention data in Metabase to find which cohorts are the problem.
Putting it in Metabase
Save your churn query in Metabase, visualize it as a line chart, and pin it to your core metrics dashboard. Use dashboard filters to let your team slice by plan, segment, or acquisition channel without writing new queries.
Define the metric — including which formula you use — once in Data Studio. That way, when the CEO asks for churn and the product team asks for churn, they get the same number.
---