What churn is and how to reduce it in your SaaS startup

Executive overview

High churn makes growth nearly impossible — replacing your entire customer base every year is unsustainable. Revenue churn (% of MRR lost per month) is the metric that matters; customer count is a distraction. Before product-market fit, no tactical fix works — you must first build something people want.

Net negative churn is the cheat code: expansion revenue outpacing cancellations means the business grows even with zero new customers.

The two types of churn

• Revenue churn: % of MRR lost in a given month — the primary metric
• Customer churn (logo churn): % of customers lost — less meaningful than revenue
• Gross churn measures raw losses; net churn subtracts expansion revenue
• Expansion revenue comes from customers upgrading plans or adding seats
• Net negative churn: expansion outweighs cancellations — business grows without new customers

Benchmarks and when to track

• Above 5% monthly churn is generally a problem for SaaS
• Consumer or low-price products can tolerate 5–7%; 9% means replacing the full base annually
• Churn numbers are unreliable below ~50–100 customers
• Track in real time via dashboard or payment-provider integration
• Check weekly (or daily) in early stage; monthly is acceptable for mature, stable businesses
• Segment by pricing tier, customer type, and cohort to find patterns

How to calculate churn

• Revenue churn: MRR canceled this month ÷ MRR at start of month
• Example: $500 canceled ÷ $10,000 starting MRR = 5% gross revenue churn
• Customer churn: customers canceled ÷ customers at start of month
• Segmenting by tier almost always shows lower-priced plans churn most
• Cohort analysis reveals whether specific acquisition batches (e.g. ad campaigns) underperform

Before product-market fit

• High churn pre-PMF usually means the product isn't solving a real problem
• Standard retention tactics don't work until PMF exists
• No single blueprint — customer development (talking to users) helps but isn't sufficient
• Avoid building new categories without significant funding and runway
• PMF is a spectrum, not a binary; churn should improve as fit improves

Fixable vs unfixable churn

Fixable:

• Never got set up — add in-app onboarding, welcome email sequences, proactive customer success outreach
• Missing features — build if on the roadmap and volume of cancellations justifies it
• Price too high — first check if it's a pattern or a one-off; target customer segments that don't complain about price
• Poor UX or performance — prioritise engineering and design resources if it's causing measurable churn
• No perceived value — explore additional features that increase utility for that customer type

Unfixable:

• Internal company politics at the customer
• Customer goes out of business
• Customer no longer needs the product (e.g. stopped doing social media)