Churn Rate: Definition, Formula, and 5 Easy Steps

Churn rate looks simple - until you try to explain it to your board, slice it by plan, and fix it without guesswork. This guide gives you a straight, operator-ready path: what churn is, how to calculate it in five steps, the mistakes that skew numbers, and how to turn metrics into actions that actually reduce churn.

At its core, churn is how much of your customer base or revenue you lost over a period. If you sell subscriptions, it’s a survival metric. If you sell contracts, it’s a renewal metric. If you run memberships or retainer services, it’s your canary in the coal mine. But not all churn is the same, and mixing types leads to bad decisions.

There are two major angles: customer churn (also called logo churn) and revenue churn. Customer churn tells you what share of accounts left. Revenue churn tells you how much money walked out, which can move differently than logos if pricing varies. Losing one enterprise customer can dwarf losing ten small plans. That’s why smart teams track both.

Another split: gross vs net. Gross revenue churn looks only at lost revenue. Net revenue churn subtracts upgrades and expansions from the revenue you lost. If expansions outpace losses, you get negative net churn - the dream state where your existing base grows without new sales. Each version answers a different question, and leaders need all three views: logo, gross revenue, and net revenue.

Logo churn is simple to communicate: “We lost 3% of customers last month.” It’s useful for understanding product-market fit by segment and for forecasting support and onboarding loads. But it can understate financial risk if your small customers churn faster than enterprise - or overstate it if your loss was mostly low-ARPU trials.

Revenue churn adds weight to the number. “We lost 0.8% of MRR” is a different board slide than “We lost 3% of customers.” Finance, RevOps, and sales leaders care about revenue churn because it feeds LTV, payback, and growth capacity. It also avoids false alarms when you run a price ladder with a long tail of small plans.

Net revenue churn is the reality check. If your expansions and cross-sells cover downgrades and cancellations, your base is compounding. If they don’t, you need to fix retention or monetize more value. The healthiest companies publish both GRR (Gross Revenue Retention) and NRR (Net Revenue Retention) and isolate them by cohort and segment.

You don’t need a PhD or a BI team to calculate churn correctly. You need clear definitions, a clean time window, and consistent data. Here’s the workflow that works in any spreadsheet or CRM dashboard.

Before you start: decide if you’re calculating customer churn or revenue churn. Do both if you can. Use the same time window for both to compare apples to apples.

Pro tip: don’t include new customers acquired during the period in the starting base when you calculate logo churn. That’s the most common mistake that underreports churn.

Step 1 - Choose your period and cohort rules

Pick a clear window: monthly is standard for SaaS; quarterly for enterprise contracts; weekly or monthly for consumer subscriptions. More important than the window is the cohort rule: are you measuring churn among customers that existed at the start of the period, or among customers acquired in a given month (acquisition cohort)? For top-line reporting, use customers at the start of the period.

Decide how to treat trials: exclude free trials unless they automatically convert to paid and are counted as “customers” in your reporting. Decide how to treat pauses: if a pause stops billing, treat it as churn for revenue and as either churn or at-risk for logo, but be consistent across periods.

Document these rules once. Nothing breaks trust in numbers faster than changing definitions mid-quarter without a note.

Step 2 - Define your starting base

For logo churn: Starting Customers = number of active, paying customers at 00:00 on Day 1 of your period. Do not add new customers acquired during the period to this denominator.

For revenue churn: Starting Recurring Revenue = MRR or ARR at 00:00 on Day 1. If you bill annually, divide by 12 to get comparable MRR, or work in ARR consistently across all revenue metrics.

If you sell multiple products or have add-ons, build the starting base at the plan or product-family level too. It will make your later analysis actionable instead of generic.

Step 3 - Count churned customers and lost revenue

Logo churn numerator: customers who were in your Starting Customers and fully canceled by the end of the period. If they canceled on Day 3, they still count as churn for that period. If they downgraded but stayed, they do not count toward logo churn (they will impact revenue churn).

Revenue churn numerator (gross): total MRR/ARR lost during the period due to cancellations and downgrades among your Starting Recurring Revenue. Don’t include revenue you failed to collect from new customers who joined mid-period. Keep delinquent churn (payment failures) tagged so you can treat it with recovery workflows.

Optional, but useful: track “save” events separately - customers who signaled intent to cancel but stayed after a discount/upgrade. This doesn’t change the formula; it just highlights which interventions work.

Step 4 - Calculate the rates

Customer (logo) churn rate (%) = (Churned Customers during period / Starting Customers) × 100.

Gross revenue churn rate (%) = (Gross MRR/ARR lost from Starting Recurring Revenue during period / Starting Recurring Revenue) × 100.

Net revenue churn rate (%) = [(Gross MRR/ARR lost − Expansion MRR/ARR within Starting Recurring Revenue) / Starting Recurring Revenue] × 100. If the result is negative, you have negative net churn (good).

Step 5 - Interpret, segment, and set targets

One number won’t fix your business. Break churn down by: acquisition channel, plan, industry, company size, region, and tenure. A 2% overall churn hides a 7% churn on your entry plan - and that might be fine if it’s a feeder to mid-tier expansion.

Set targets that reflect your motion. A self-serve SMB tool can be healthy at 3–5% monthly logo churn if acquisition is cheap and expansion is strong. An enterprise platform with 12-month contracts should sit near 0.5–1% monthly logo churn or better, with GRR ≥ 90% and NRR ≥ 110–130% depending on upsell motion.

Create a short weekly report: logo churn %, gross revenue churn %, NRR %, net adds, and saves. Trend it. If a change breaks your trend, check definitions and data sources before you celebrate or panic.

Example 1 - SaaS monthly: You start April with 1,000 paying customers (Starting Customers). During April, 40 cancel. Logo churn = 40 / 1,000 = 4%. For revenue, you start April with $100,000 MRR. You lose $6,000 from cancellations and downgrades. You expand $4,000 via upgrades. Gross revenue churn = $6,000 / $100,000 = 6%. Net revenue churn = ($6,000 − $4,000) / $100,000 = 2%.

Example 2 - Annual contracts: You start Q1 with $1.2M ARR. During Q1, one $60k account churns; two $20k accounts expand to $30k each (+$20k ARR). Gross revenue churn = $60k / $1.2M = 5%. Net revenue churn = ($60k − $20k) / $1.2M = 3.33%. Your logo churn might be 1/80 = 1.25%, but the revenue story is different and should guide where you act.

Example 3 - Membership with tiers: 5,000 members at start; $250,000 MRR. 300 churn from the basic tier; 20 churn from premium; 100 upgrade from basic to standard; 15 downgrade from premium to standard. Segmenting shows basic-tier onboarding is leaking. Logo churn = 320 / 5,000 = 6.4%. Gross revenue churn counts lost dollars from cancellations and downgrades; net churn recognizes upgrades offsetting some loss.

Snapshot churn (the five-step method above) tells you what happened this period across your whole base. Cohort churn tells you how a specific group behaves over time - for example, customers acquired in January and their churn by month 3, 6, and 12. If your product or onboarding evolved recently, cohort churn is how you prove the new experience is working.

Cohorts are essential when your acquisition mix changes. If you switch from organic to paid, or add a reseller channel, the average churn can look stable while the quality of new customers drops. Segment cohorts by channel and persona to avoid false comfort.

Use both views in leadership meetings. Snapshot churn for operational cadence; cohort churn to validate initiatives and forecast LTV and payback by acquisition source.

Gross Revenue Retention (GRR) answers: “How much of our starting revenue did we keep, ignoring expansions?” Healthy GRR for mid-market SaaS is often 85–95% annually; enterprise targets 90–97% depending on stickiness. If GRR falls, you have a core retention problem, not a pricing problem.

Net Revenue Retention (NRR) answers: “After churn, downgrades, and expansions, did our base grow?” Best-in-class enterprise SaaS publishes 120–140% NRR; strong mid-market sits 105–125%. If NRR is high while GRR is flat, your expansion engine works but your core experience still leaks - don’t mask it with upsell incentives alone.

Logo retention is the flip side of logo churn. If your monthly logo churn is 2%, your monthly logo retention is 98%. Annualize carefully: 98%^12 ≈ 78.5% annual logo retention - a very different story than reading “98%” in isolation. This is why leaders demand both the rate and the math behind it.

Raw churn won’t tell you what to do Monday. Segmentation will. Start with four cuts: by acquisition channel (paid search, referral, outbound), by plan/tier, by company size/industry, and by tenure (0–30 days, 31–90, 91–365, 365+). Each cut should highlight a different operational lever (onboarding, product gap, pricing, ICP drift).

If early-tenure churn is high, fix activation: time-to-value, onboarding communications, setup friction. If late-tenure churn is high, look for competitive displacement or value plateau: missing features, stale content, or pricing misalignment as customers scale.

Don’t forget country/region. Payment failures, tax quirks, or local compliance can masquerade as product churn when it’s really operational drag.

You can’t calculate churn accurately without a single source of truth. For most growth-stage teams, that’s your CRM plus your billing tool. Align your CRM “Lifecycle” or “Subscription Status” with your billing events: trial started, converted, active, paused, delinquent, canceled, reactivated. Tag downgrade/upgrade separately from cancel.

Build a simple churn dashboard: start-of-period customers and MRR, churned logos and dollars, downgrades, expansions, NRR/GRR, and saves. Add filters for segment and cohort month. Automate the extraction from billing to CRM daily, not weekly, so you can catch issues in time to act.

If you want help wiring this into HubSpot, Salesforce, or a clean spreadsheet workflow, our CRM implementation playbooks include cohort reporting, NRR/GRR dashboards, and save reasons your team can act on. See how we build it here: CRM and RevOps implementation.

Most churn has an early warning: no logins, low feature usage, payment retries piling up, support tickets about the same blocker. Automate your saves. Trigger a human task when product usage drops below a defined threshold for 7 days. Fire a contextual in-app nudge with a one-click “Need help?” flow into success. Start dunning sequences at the first failed charge, not the third.

Build win-back paths that match the reason for leaving: price sensitivity (offer a downgrade path with preserved data), missing feature (capture the request and notify when shipped), bad fit (exit interview and a clean offboarding email that invites future return). Track win-back conversion so you don’t guess which playbooks pay off.

If you lack the workflows, you don’t need to hire a big team or a traditional agency. You can stand up high-signal automations in weeks using Make/n8n, your CRM workflows, and a few crisp triggers from product analytics. If you want proven systems rather than experiments, explore our marketing automation service.

Your acceptable churn depends on acquisition cost (CAC), gross margin, and expansion potential. If CAC payback is 4–6 months and gross margin is 80%+, you can tolerate slightly higher churn as long as expansions push NRR above 100%. If CAC payback stretches past 12 months, churn must be low and expansion strong or growth will stall.

Rough guideposts many operators use: monthly logo churn of 1–2% is healthy for SMB SaaS; 0.5–1% for mid-market; <0.5% for enterprise. Annual GRR of 85–95% is common; NRR of 105–130% signals strong expansion. Don’t copy-paste these - validate against your price points, contract terms, and ICP.

Translate churn to cash: a 1-point improvement in monthly logo churn (from 4% to 3%) on a 1,000-customer base is 10 extra customers retained per month. If ARPU is $100, that’s $1,000 MRR retained monthly, compounding. It’s easier to keep $1,000 than to acquire it at a $300 CAC per customer.

Including new customers in the denominator. This makes churn look magically better as you grow. Always use the starting base for the period. If you want to study new customers, use a cohort view with its own denominator.

Mixing voluntary and involuntary churn. Payment failures (involuntary) belong in your revenue churn numerator, but you should tag them separately and measure recovery rate. Cancellation reason “no time to use it” requires a different fix than “card expired.”

Counting pauses and reactivations inconsistently. If your pause policy stops billing, it’s revenue churn now - even if they come back. Track reactivations in a separate metric; don’t quietly net them out of the period’s churn or you’ll mask volatility.

Agencies write decks about churn. Operators wire systems that change it. At Anton Chisliuk · AI Marketing Ops, the model is simple: one senior operator builds your CRM dashboards, content engine, outbound, and automations end to end - no account-manager layer. Proof matters: for Cleverence (B2B SaaS), we grew from 0 to ~80K organic visits/month and ~2K inbound leads/month; retention wins followed because activation and qualification improved. If you want to see who’s behind the work and how the operator model runs, start here: about the operator model.

Want a deeper look at how sustained pipeline improves retention quality? See the Cleverence case detail - the lift in qualified demand lowered bad-fit starts and reduced early churn: Cleverence case study.

Lead with the three-number slide: Logo churn %, GRR %, NRR %. Then show a one-line trend for the last 6–12 periods. If something changed (pricing, onboarding, a big feature), annotate the timeline so the conversation is about causes, not disbelief in data.

Follow with segmentation that suggests action: “Churn is concentrated in SMB self-serve within first 45 days; activation tasks take 3–5 hours; we’re shipping an in-app checklist and two concierge slots per week.” You’ll get faster decisions when the slide connects metric to lever.

End with the saves pipeline: at-risk accounts, open save tasks, expected retained MRR, and a short list of experiments. Churn is not a static number - it’s a queue you can work like sales.

Churn becomes manageable when you measure it the same way every time, cut it by the segments you can influence, and assign real owners to saves and fixes. The five-step workflow here gives you a clean base; the rest is instrumentation and habits.

If you need the dashboards, billing/CRM sync, and save workflows stood up fast, skip the agency tax and get an operator to wire it once and teach your team. Start with a practical RevOps build: CRM and churn reporting setup.

Author: Anton Chisliuk - Leads, sales, content, and systems - one senior operator.

- What is a good churn rate for SaaS?

For SMB-focused SaaS, 1–3% monthly logo churn is common; for mid-market, 0.5–1.5%; for enterprise, <1% monthly. Annual GRR often targets 85–95%, and strong NRR sits between 105–130% depending on upsell motion. Always calibrate to your pricing, margin, and contract terms.

- Should I use MRR or ARR for revenue churn?

Use the unit that matches how you report other revenue metrics. MRR is standard for monthly cadence; ARR for board-level annual views. Be consistent within a dashboard. If you bill annually, you can convert to MRR (ARR/12) for comparability across teams.

- Do I include new customers in churn calculations?

No. For snapshot churn, the denominator is customers at the start of the period. New customers affect net adds, not the churn percentage. If you want to study how new customers behave, build acquisition cohorts and track their retention over time.

- How do I handle downgrades vs cancellations?

Downgrades do not affect logo churn; they reduce revenue and belong in gross revenue churn. Cancellations hit both logo churn and revenue churn. Track downgrade reasons separately - they often point to packaging or value-communication issues instead of product failure.

- What about failed payments (involuntary churn)?

Count lost revenue from failed payments in gross revenue churn, but tag it as involuntary. Then measure recovery rate from dunning flows. Many teams recover 20–50% of failed charges with better retries, reminders, and payment method updates.

- How do I calculate net revenue retention (NRR)?

NRR (%) = [(Starting Recurring Revenue − Churned/Downgraded Revenue + Expansion Revenue within the same cohort) / Starting Recurring Revenue] × 100. If the result is over 100%, your base is growing without new customer adds.

- Can I average monthly churn to get annual churn?

Not by simple addition. To annualize monthly logo retention, compound it: if monthly logo retention is 98%, annual logo retention is 0.98^12 ≈ 78.5%. The same compounding applies in reverse for churn.

- How often should I report churn?

Weekly for ops (quick trend checks and saves), monthly for leadership, quarterly for the board with cohort and segment deep dives. Make sure all reports use the same definitions to avoid confusion.