Quick Answer

Churn rate is calculated by dividing the number of customers (or the amount of recurring revenue) lost during a period by the number (or revenue) you started with, then multiplying by 100. The exact formula shifts depending on whether you are measuring customer churn, gross revenue churn, or net revenue churn, and each answers a different retention question.

Introduction

Churn rate looks like a simple percentage until you start querying real subscription data and discover mid-period cancellations, plan upgrades, annual contracts, and refunds all pulling the number in different directions. The industry-standard churn rate formula divides lost customers or lost MRR by the starting count for the period, but that baseline hides critical decisions about how upgrades, downgrades, and new signups get treated. Data engineers and growth operators building this logic in SQL or dbt often inherit a number from a dashboard without knowing which of those decisions was made. That gap is exactly where board reports drift from reality and why two teams inside the same company can report churn figures that differ by 40 percent.

Key Takeaways:

• Customer churn rate divides lost customers by starting customers; revenue churn divides lost MRR by starting MRR, and the two rarely match.

• Net churn subtracts expansion revenue from lost revenue, which can produce a negative number when upsells outpace cancellations.

• Most calculation errors come from including new customers in the denominator or counting downgrades as full churn events.

The Core Churn Rate Formula

The churn rate formula answers one question: what percentage of what you had at the start of a period did you lose by the end of it? That framing matters because it forces you to exclude anything you gained during the period from the denominator, which is where most miscalculations begin.

Customer Churn Rate Step by Step

To calculate customer churn rate correctly, you need a clean count of customers at period start and a count of those same customers who canceled before period end. New signups within the period should not appear in either number.

• Define the period: monthly, quarterly, or annual, and hold it consistent across reports.

• Count starting customers: active paying customers on day one of the period, excluding trials and free users.

• Count churned customers: those from the starting cohort who canceled or failed to renew before period end.

• Apply the formula: (churned customers / starting customers) × 100.

• Sanity check the denominator: confirm no new signups from the period were pulled in by your JOIN logic.

For a worked example, say you started July with 500 paying customers and 15 of them canceled by month-end. Your customer churn rate is 3.0 percent, regardless of how many new customers signed up during the same window. This is where a well-structured retention analytics and churn pipeline pays off, because it isolates the starting cohort cleanly.

Revenue Churn and the MRR Version

Revenue churn measures lost recurring revenue rather than lost logos, which matters more when your pricing is tiered and customer size varies. The gross revenue churn formula is (MRR lost from existing customers during the period / MRR at start of period) × 100, and it excludes new MRR from acquisitions. According to authoritative churn rate definitions, this separation between existing-customer movement and new business is what keeps the metric interpretable over time.

Net Churn, Gross Churn, and Real-World Edge Cases

Once upgrades, downgrades, and mid-period cancellations enter the picture, a single churn number no longer tells the full story. Splitting the calculation into gross and net views is the standard fix, and each one is used for different reporting purposes.

Comparing Churn Metrics Side by Side

The table below shows how the same period of activity produces different numbers depending on which churn definition you use. Assume a starting MRR of $100,000, $8,000 in canceled MRR, $3,000 in downgrades, and $6,000 in expansion revenue from upgrades.

The takeaway is that net churn can look flattering because expansion revenue masks real cancellation problems, while gross churn exposes the retention issue underneath. Serious SaaS unit economics metrics reporting almost always shows both alongside each other, because using only net churn hides the leak that a customer retention team actually needs to fix.

Handling Upgrades, Downgrades, and Partial Periods

Upgrades should never reduce your churn number directly; they belong in expansion MRR, which then flows into the net churn calculation as a separate input. Downgrades are partial churn events, so a customer moving from a $500 plan to a $200 plan contributes $300 of churned MRR, not a full logo churn. Mid-period cancellations are trickier: the industry convention, reinforced by churn calculation methodology, is to count the customer as churned in the period their cancellation takes effect, not when they submitted the request. TrackRaptor readers building this in a data warehouse should model each subscription state change as an event with an effective date, then aggregate by period, rather than trying to derive churn from a nightly snapshot table alone.

Conclusion

Calculating churn rate correctly comes down to defining the period rigidly, excluding new business from the denominator, and choosing the right variant (customer, gross revenue, net revenue) for the decision you are trying to inform. Report gross and net side by side, treat downgrades as partial revenue churn, and time cancellations by their effective date rather than their request date. Once the baseline is trustworthy, everything downstream (cohort analysis, LTV modeling, and churn prediction models) becomes far more reliable. The teams that get this right treat churn as an engineered pipeline output, not a spreadsheet formula. That discipline is what separates a defensible board number from one that gets picked apart in the next diligence call.

Want to see how top data teams model this end-to-end? Explore the TrackRaptor customer retention playbook for warehouse-native patterns, SQL references, and edge-case handling used by production SaaS teams.

Frequently Asked Questions (FAQs)

What is the industry standard churn rate formula?

The industry standard is (customers lost during the period/customers at the start of the period) × 100, with new signups excluded from the denominator.

How do you calculate churn for subscription-based products?

For subscription products, calculate both customer churn (logos lost) and MRR churn (recurring revenue lost) over a fixed period, since a single metric hides plan mix effects.

How do you calculate churn rate when customers upgrade or downgrade?

Treat upgrades as expansion MRR (not a reduction to churn) and downgrades as partial revenue churn equal to the dollar difference between the old and new plan.

Is net churn a better indicator than gross churn?

Net churn is better for overall revenue health but worse for diagnosing retention problems, so mature teams report both alongside each other rather than choosing one.

How does churn rate affect customer lifetime value?

LTV is inversely proportional to churn, so a churn rate of 5 percent monthly implies an average customer lifespan of 20 months, which caps how much you can profitably spend on acquisition.

What role does cohort analysis play in churn reduction?

Cohort analysis reveals whether churn is concentrated in specific signup periods, plans, or segments, which is impossible to see in a single blended monthly churn number.

How should GDPR-compliant companies calculate churn rates?

GDPR-compliant companies should aggregate churn on pseudonymized subscription identifiers and avoid retaining raw personal data solely for retrospective churn recomputation.

About the Author

Noah Richardson is a SaaS Metrics Advisor who writes about KPIs, retention analysis, and revenue-focused analytics for growth-stage software companies. He specializes in helping data and finance teams build defensible churn, LTV, and unit economics reporting directly inside their data warehouse. His work centers on turning noisy subscription data into board-ready metrics that hold up to investor scrutiny.