Churn Rate
Also known as: customer churn, attrition rate, customer attrition, subscriber churn, net revenue retention, voluntary churn
Churn rate (also called attrition rate) is the percentage of customers who stop using a product or service within a defined time period, typically monthly or annually.
For subscription businesses (SaaS, streaming services, telecommunications), churn rate is one of the most important metrics: high churn means the business must constantly acquire new customers just to maintain revenue, while low churn allows compounding growth. Monthly churn rate of 3% might sound small but means 31% of customers leave over a year. Understanding why customers churn (product gaps, competitive alternatives, price sensitivity, service failures) drives retention strategies.
Churn Rate, explained at three levels
At home
Churn rate tells you what percentage of customers cancel in a given time period. If a streaming service has 1,000 subscribers and 30 cancel in a month, its monthly churn rate is 3%. This means the company needs to add at least 30 new subscribers each month just to stay at 1,000 total. The lower the churn rate, the healthier the business: customers stay longer, spend more over their lifetime, and the company doesn't constantly need to replace lost customers with expensive new customer acquisition.
At work
Churn rate is crucial for subscription businesses because it determines customer lifetime value and how much you can spend to acquire a customer. If customers stay for an average of 24 months and pay $100/month, their lifetime value (LTV) is approximately $2,400 (minus costs). If churn is too high and customers stay only 6 months, LTV is $600: dramatically reducing what the business can profitably spend on customer acquisition. Reducing churn is often more economically efficient than increasing new customer acquisition: retaining an existing customer is commonly estimated to cost 5 to 7 times less than acquiring a new one.
At home
Customer churn vs. revenue churn: customer churn counts the number of customers who cancel; revenue churn measures the revenue lost to cancellation. These differ when customers pay different amounts: a high-value customer churning has more revenue impact than many low-value customers. Revenue churn is sometimes negative (net revenue expansion): if the revenue from upgrades, upsells, and expansion among remaining customers exceeds the revenue lost to churn, the company has negative net revenue churn. Net Revenue Retention (NRR) above 100% means the customer base is growing in revenue even if some customers leave: a common feature of the best-performing enterprise SaaS companies.
At work
Measuring churn rate: the standard formula is (customers who churned in period) / (customers at start of period). Variations create different numbers: using the average of start and end customers in the denominator; excluding customers who signed up and cancelled within the same period (new churn vs. existing churn); measuring at different granularities (monthly vs. quarterly vs. annual). Comparing churn rates across companies or benchmarks requires using the same definition. Annual churn is not simply monthly churn x 12: a 3% monthly churn is approximately 31% annual churn (using compound rates).
At home
Voluntary vs. involuntary churn: voluntary churn is customers who actively cancel (dissatisfied, no longer need the product, found a competitor). Involuntary churn is customers who are lost due to payment failures (expired cards, insufficient funds, bank declines) without intending to cancel. Involuntary churn can represent 20-40% of total subscription churn. Dunning management (automated email sequences and card retry logic to recover failed payments) can recover a significant portion of involuntary churn. Separating these in churn measurement is important: the interventions differ completely: involuntary churn is addressed with payment infrastructure; voluntary churn requires product, customer success, and pricing interventions.
At work
Predictive churn modelling: rather than analyzing who has already churned, predictive churn models identify customers who are at high risk of churning in the near future: enabling proactive retention intervention. Feature engineering for churn models typically includes: product usage metrics (declining login frequency, reduced feature usage), support history (increased ticket volume, escalations), account changes (downgrading plan, removing users), and engagement signals (not opening emails, not responding to outreach). Logistic regression and gradient boosted tree models are common approaches. Model evaluation for churn prediction prioritizes precision-recall trade-offs: too many false positives waste customer success resources; too many false negatives miss high-risk customers.
Real-world examples
At home
Jerome used four subscription services simultaneously: a gym ($50/month), a streaming service ($15/month), a podcast app ($8/month), and a language learning app ($13/month). He cancelled the podcast app after 3 months and the language learning app after 6 months when he realized he wasn't using them enough to justify the cost. Both cancellations were voluntary churn: he had tried the products and found they didn't fit his life well enough. The gym and streaming service, he had been using for three years without interruption. The difference was whether the product delivered obvious, regular value.
At work
Aisha was the CEO of a B2B SaaS company with 5% monthly churn. An investor presented her with a benchmark showing similar companies averaged 2% monthly churn. She built a churn analysis: 60% of churned customers cited lack of specific features; 25% cited price; 15% cited switching to a competitor. She prioritized a product roadmap focused on the most-cited missing features, implemented a 60-day pre-cancellation save campaign (proactive outreach when usage dropped), and introduced annual plans with a 15% discount (reducing payment frequency and securing revenue). Over 12 months, monthly churn decreased from 5% to 2.5%.
Common confusions
- Many people assume a low churn rate means customers are happy with the product — Low churn can reflect customer satisfaction, but it can also reflect lock-in, high switching costs, lack of viable alternatives, or contractual commitments. Enterprise software with multi-year contracts has low churn by construction, not because customers are necessarily satisfied, and customers who cannot cancel often reduce their usage and decline to renew at contract end. Net Promoter Score (NPS), customer satisfaction surveys, and product usage data provide information about customer happiness that churn rate alone does not reveal. A company with low churn but low NPS and declining usage is at risk of mass non-renewal when contracts expire.
- Some assume that reducing churn is always more important than increasing new customer acquisition — The relative priority depends on the company's current churn rate and growth stage. Early-stage companies with small customer bases often need acquisition growth more than churn optimization: even high churn can be tolerated if the market is large and the product is improving rapidly. At scale (larger customer base), the compound economics of churn become more significant and churn reduction's value typically exceeds acquisition scaling. Additionally, some churn is healthy: churning customers who were a poor fit (wrong use case, incorrect expectations) reduces support burden and improves cohort data about the right customer profile. Not all churn reduction efforts are worth their cost.
For analysts
Churn analysis is a high-value engagement for any subscription or recurring revenue business. The diagnostic starts with the right churn measurement (customer churn vs. revenue churn, monthly vs. annual, voluntary vs. involuntary) to ensure the client and their stakeholders are measuring what matters. Subsequent analysis: what are the most common reasons for voluntary churn (exit survey data)? At what tenure point does most churn occur (is there a 30-day cliff, a 90-day cliff, or ongoing linear churn)? Which customer segments have the highest churn rates (does the problem cluster in a specific product tier, company size, or use case)? These questions direct retention investment more precisely than simple aggregate churn rate improvement goals.
History & origin
The concept of customer attrition and its measurement is as old as subscription services. Monthly churn measurement became a standard SaaS metrics practice in the 2000s as the software-as-a-service model matured. David Skok's SaaS metrics writing and Bessemer Venture Partners' cloud benchmarks (late 2000s and 2010s) popularized churn as a core SaaS metric for investors and operators. Net Revenue Retention (NRR) / Net Dollar Retention (NDR) emerged as a related metric that captures both churn and expansion revenue: a company can have positive NRR even with some customer churn if remaining customers expand their spending. Investor benchmarks for SaaS companies now treat churn and NRR as central health metrics, alongside measures such as the 'Rule of 40' for growth plus profit margin.