# Churn Rate: Complete Guide 2026 | Enrich Labs > Calculate churn rate, compare 2026 SaaS benchmarks from Recurly, CRV, and Lenny Rachitsky, and cut logo plus revenue churn without guessing. _Source: https://www.enrichlabs.ai/blog/churn-rate-complete-guide-2026_ --- #TLDR Churn rate is the share of customers or recurring revenue you lose in a period. Logo churn counts accounts. Revenue churn counts dollars, including downgrades. Net revenue retention (NRR) folds expansion back in, which is why investors treat NRR as the headline and logo churn as the supporting chart. [Recurly](https://recurly.com/research/churn-rate-benchmarks/) (July 2026 network data) reports **3.60% overall churn** across industries, 2.34% voluntary, and 1.25% involuntary. SaaS sits at **3.22%** total (2.16% voluntary, 1.06% involuntary). [CRV](https://www.crv.com/content/saas-churn-rate) (June 2026) treats seed-stage monthly logo churn of 3% to 5% as typical, and **100% NRR as the Series A baseline**, with 110% to 120% competitive. [Lenny Rachitsky](https://www.lennysnewsletter.com/p/monthly-churn-benchmarks) (ProfitWell slice) put good B2B SMB monthly churn at 2.5% to 5%, and good enterprise monthly churn at 1% to 2%. [Harvard Business Review](https://hbr.org/2014/10/the-value-of-keeping-the-right-customers) cites Bain: acquiring a new customer costs five to 25 times more than keeping one, and a 5% lift in retention can raise profits 25% to 95%. Pair that with [customer acquisition cost](https://www.enrichlabs.ai/blog/customer-acquisition-cost-complete-guide-2026) and [customer lifetime value](https://www.enrichlabs.ai/blog/customer-lifetime-value-complete-guide-2026). High churn wrecks LTV:CAC even when CAC looks fine. Split voluntary vs involuntary. Fix failed payments first. Then fix onboarding, product value, and [customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026). Helena at [Enrich Labs](https://www.enrichlabs.ai) can run the acquisition side so you stop buying customers you cannot keep. * * * ## What is churn rate? Churn rate is the percentage of a starting cohort that leaves in a defined window. [Recurly](https://recurly.com/research/churn-rate-benchmarks/) defines it as subscribers who cancel or fail to renew, including people who never meant to leave because a card declined. Two units, two stories: - **Customer (logo) churn:** accounts lost divided by accounts at the start of the period. - **Revenue churn:** recurring revenue lost to cancels and downgrades divided by starting MRR or ARR. [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters) treats revenue churn as MRR lost from cancels and downgrades. Gross revenue churn ignores expansion. Net revenue churn subtracts expansion MRR. Negative net revenue churn means existing customers grow faster than they leave. [Planhat](https://www.planhat.com/customer-success/churn-and-retention) frames logo churn as the inverse of customer retention rate. Retention = 1 minus logo churn for the same cohort and window. Do not mix monthly logo churn with annual NRR on one slide. Churn is a product, pricing, billing, and [go-to-market strategy](https://www.enrichlabs.ai/blog/go-to-market-strategy-complete-guide-2026) problem. A leaky ICP shows up here first. If [demand generation](https://www.enrichlabs.ai/blog/demand-generation-complete-guide-2026) fills the pipe with tire-kickers, [customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026) spends the next year catching them on the way out. [Recurly](https://recurly.com/research/churn-rate-benchmarks/) also separates **voluntary churn** (the subscriber cancels) from **involuntary churn** (payment fails). That split is the operational cut to make this quarter. * * * ## How to calculate churn rate Pick a period. Use the same start-of-period denominator every time. Do not put new logos in the denominator for the month they arrived if your definition is beginning-of-period customers. State the definition in the dashboard and in the board pack. ### Customer churn formula Customer churn rate = (customers lost in the period / customers at the start of the period) x 100 Example: 1,000 customers on March 1, 40 cancels in March. Monthly logo churn is 4%. ### Gross revenue churn [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters): (churned MRR / MRR at the end of the previous month) x 100 Example: $100,000 starting MRR, $5,000 lost to cancels and downgrades. Gross revenue churn is 5%. ### Net revenue churn [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters): (\[churned MRR minus expansion MRR\] / starting MRR) x 100 Same example plus $2,000 expansion. Net revenue churn is 3%. NRR is the cousin of net revenue churn. NRR above 100% means expansion more than offsets churn and contraction. [CRV](https://www.crv.com/content/saas-churn-rate) uses NRR in diligence more than logo counts. ### Monthly vs annual Do not annualize with x12 if you need a true annual rate. Compounding matters. [CRV](https://www.crv.com/content/saas-churn-rate) notes that 3% to 5% monthly logo churn at seed annualizes to roughly 30% to 45%. A 1% monthly rate is closer to 11% to 12% annual if you care about the remaining base, not a naive 12%. Report both. Sales-led annual contracts need annual logo churn. Self-serve monthly billing needs monthly logo and monthly revenue churn. [Revenue operations](https://www.enrichlabs.ai/blog/revenue-operations-complete-guide-2026) should own the definition so finance, CS, and marketing stop arguing about which number is real. [Marketing attribution](https://www.enrichlabs.ai/blog/marketing-attribution-complete-guide-2026) then ties channel to cohort churn. That is how you learn that a [PPC agency](https://www.enrichlabs.ai/blog/ppc-agency-complete-guide-2026) campaign can look cheap on CAC and expensive on 90-day logo loss. * * * ## Logo churn vs revenue churn vs NRR Metric What it answers Blind spot Logo churn How many accounts left A $200 SMB cancel equals a $20k enterprise cancel Gross revenue churn How many dollars left, including downgrades Expansion can hide a leak Net revenue churn / NRR Did the starting book of business grow Can look healthy while logos collapse [CRV](https://www.crv.com/content/saas-churn-rate) is blunt: NRR outweighs logo churn. Expansion, contraction, upsells, and seat adds sit in NRR. They set Series A NRR at 100% baseline, 110% to 120% competitive, 120%+ premium. Median NRR for $1M to $5M ARR sits near 104% in the figures they cite. Peak-cycle 2021 numbers no longer apply. [Planhat](https://www.planhat.com/customer-success/churn-and-retention) keeps logo churn and retention as inverses. Use both. A company can post 5% annual logo churn and 122% NRR if remaining accounts expand. That is a mid-market or enterprise pattern, not an SMB pattern. Segment before you average. [CRV](https://www.crv.com/content/saas-churn-rate) flags blended SMB plus enterprise churn as a diligence red flag. SMB below $10k ACV can sit near 4.1% monthly logo churn in the sources they cite. Enterprise above $100k ACV often sits below 1% monthly logo churn. [Customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026) and [revenue operations](https://www.enrichlabs.ai/blog/revenue-operations-complete-guide-2026) need one definition of churn in the CRM, billing system, and board deck. [Sales enablement](https://www.enrichlabs.ai/blog/sales-enablement-complete-guide-2026) should teach AEs the same language so a retained logo is not a $1 account on a zombie seat. [Account based marketing](https://www.enrichlabs.ai/blog/account-based-marketing-complete-guide-2026) teams feel this in reverse: one logo is the whole book. Logo churn of 1% can still be a disaster if that 1% was the named account. * * * ## 2026 benchmarks: what is a good churn rate? A good churn rate depends on segment, ACV, contract length, and stage. ### Recurly network (July 2026) [Recurly](https://recurly.com/research/churn-rate-benchmarks/) publishes network medians: - **Overall:** 3.60% total, 2.34% voluntary, 1.25% involuntary - **SaaS:** 3.22% total, 2.16% voluntary, 1.06% involuntary - **Business and professional services:** 3.44% total - **Ecommerce:** 4.25% total - **Education:** 4.99% total Recurly also writes that below 2% is strong in almost any segment, 2% to 4% is the zone most well-run subscription businesses occupy, and above 5% deserves an investigation. For software on their network they cite 3.04% median overall for software businesses, with top quartile at 1.78% or below. In the $250+ ARPC cohort, median sits at 3.54% with involuntary churn at 0.18%. Treat Recurly as a billing-network view. It is not the same as private B2B SaaS ARR surveys. ### CRV investor view (June 2026) [CRV](https://www.crv.com/content/saas-churn-rate): - **Seed (below $1M ARR):** 3% to 5% monthly logo churn is typical - **Series A ($1M to $10M ARR):** they cite 12.50% median annual revenue churn for private B2B SaaS (2025), with the top 25% below 5.48% - **Growth ($10M+ ARR):** monthly logo churn tightens; contract length does most of the work - **NRR:** 100% Series A baseline; 110% to 120% competitive; 120%+ premium CRV also notes AI-native products can show much weaker GRR than traditional B2B SaaS, with higher-priced AI products (above $250/month) closer to 70% GRR in the ChartMogul figures they cite. Trajectory beats a single snapshot. ### Lenny Rachitsky / ProfitWell [Lenny Rachitsky](https://www.lennysnewsletter.com/p/monthly-churn-benchmarks) (February 2022, still the most cited monthly-churn ladder): - **B2C SaaS:** 3% to 5% monthly is good; under 2% is great - **B2B SMB + mid-market:** 2.5% to 5% good; under 1.5% great - **B2B enterprise:** 1% to 2% good; under 0.5% great He stresses cohort retention and NRR over blended monthly churn, because monthly churn mixes new and old users. Month-1 churn of 5% to 50% is often activation failure or paid acquisition of the wrong ICP. ### Why the numbers disagree Recurly is a processor network. CRV is a VC diligence frame. Lenny is an operator plus ProfitWell slice. Use the frame that matches your motion. A $29/month [product-led growth](https://www.enrichlabs.ai/blog/product-led-growth-complete-guide-2026) tool should not copy an enterprise NRR target. An enterprise [SEO agency](https://www.enrichlabs.ai/blog/seo-agency-complete-guide-2026) client on annual contracts should not copy Recurly ecommerce. [Content marketing for SaaS](https://www.enrichlabs.ai/blog/content-marketing-for-saas-complete-guide-2026) that attracts researchers will show different 30-day churn than [B2B lead generation](https://www.enrichlabs.ai/blog/b2b-lead-generation-complete-guide-2026) that books demos with budget owners. Benchmark the cohort, not the company average. * * * ## Voluntary vs involuntary churn [Recurly](https://recurly.com/research/churn-rate-benchmarks/) splits the two. Voluntary is an active cancel: price, product, competitor, priority change. Involuntary is a failed payment: expired card, bank decline, limit hit. The subscriber did not choose to leave. [Recurly's involuntary churn docs](https://docs.recurly.com/recurly-subscriptions/docs/involuntary-churn) name expired cards, bank changes, and failed payments as the usual causes. On Recurly's July 2026 all-industry mix, involuntary is 1.25% of 3.60% total. That is about one-third of network churn. SaaS involuntary is 1.06% of 3.22%. [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters) attacks involuntary churn with Smart Retries, dunning email, card account updater, and more payment methods. [Recurly's failed-payment recovery write-up](https://recurly.com/blog/failed-payment-recovery-data-based-strategy/) makes the same point: involuntary churn is a recovery operations problem, not a product pep talk. Fix this before you hire another CSM. Voluntary churn needs product, onboarding, packaging, and [customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026). Exit surveys help. Usage drop-off helps more. [AI customer service](https://www.enrichlabs.ai/blog/ai-customer-service-complete-guide-2026) can catch ticket sentiment before the cancel click. [Email](https://www.enrichlabs.ai/blog/best-ai-for-email-marketing-2026) dunning and save flows still do the unpaid work that a human CS team cannot scale at SMB volume. [Recurly's churn management playbook](https://recurly.com/blog/strategies-to-reduce-and-prevent-subscriber-churn/) keeps the same split: measure both, then run different plays. * * * ## Why churn rate matters for unit economics Churn sets lifetime. Lifetime months is roughly 1 divided by monthly churn in a steady state. Double monthly churn, cut LTV in half. [CRV](https://www.crv.com/content/saas-churn-rate) walks a $2,000 ARPU, 80% gross margin example: monthly churn from 1.5% to 3.0% cuts LTV from $106,667 to $53,333 and LTV:CAC from 8.9 to 4.4. CAC payback stays the same because payback is a short cash metric. The long-term business is not the same. [HBR](https://hbr.org/2014/10/the-value-of-keeping-the-right-customers) (Amy Gallo, 2014) cites Frederick Reichheld / [Bain](https://media.bain.com/Images/BB_Prescription_cutting_costs.pdf): a 5% increase in customer retention rates increases profits 25% to 95%, and new-customer acquisition costs five to 25 times more than retention. [CAC](https://www.enrichlabs.ai/blog/customer-acquisition-cost-complete-guide-2026) without churn is a vanity ratio, and [CLV](https://www.enrichlabs.ai/blog/customer-lifetime-value-complete-guide-2026) is incomplete without a churn assumption. Paid acquisition that ignores retention is a treadmill. If [Meta ads](https://www.enrichlabs.ai/blog/best-ai-for-meta-ads-2026) or [Google Ads](https://www.enrichlabs.ai/blog/best-ai-for-google-ads-2026) bring in the wrong ICP, month-1 churn spikes. ChenLi Wang, quoted in [Lenny's piece](https://www.lennysnewsletter.com/p/monthly-churn-benchmarks), told companies that cutting paid spend in half can drop churn because the remaining mix is healthier. Map churn into [GTM](https://www.enrichlabs.ai/blog/go-to-market-strategy-complete-guide-2026) and [PLG](https://www.enrichlabs.ai/blog/product-led-growth-complete-guide-2026) reviews. Marketing owns the promise. Product owns time-to-value. CS owns expansion and save plays. [RevOps](https://www.enrichlabs.ai/blog/revenue-operations-complete-guide-2026) owns the math. [Product marketing](https://www.enrichlabs.ai/blog/product-marketing-complete-guide-2026) owns the packaging story that either matches the invoice or invites a downgrade. Helena at Enrich Labs sits on the acquisition and lifecycle-content side: ads, SEO, email, and social that match the ICP you can retain, not the cheapest click. Compare that motion to hiring an [AI marketing agency](https://www.enrichlabs.ai/blog/helena-vs-ai-marketing-agency-2026) if you want a full team vs one operator. * * * ## How to reduce churn rate Work the stack in this order. Tools sit at the bottom of this section. **Enrich Labs (Helena) sits first** because most SaaS churn we see starts as a bad acquisition mix, not a missing CS seat. ### 1\. Instrument the definition One start-of-period denominator. Separate logos, gross revenue, net revenue, voluntary, involuntary. Cohort tables by month of start, plan, ACV band, and channel. [CRV](https://www.crv.com/content/saas-churn-rate) treats inconsistent definitions as a trust problem in a data room. ### 2\. Kill involuntary churn Retries, dunning, account updater, backup payment methods. [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters) documents this path. Recurly's $250+ ARPC cohort shows how little involuntary churn remains when payment quality is high (0.18%). [Recurly on failed payments](https://docs.recurly.com/recurly-engage/docs/failed-rebill) is the operational checklist. ### 3\. Fix activation, not late-stage saves Month-1 to month-3 churn is usually onboarding. [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters) puts customized onboarding first on the reduce list. Time-to-first-value beats a longer academy. [PLG](https://www.enrichlabs.ai/blog/product-led-growth-complete-guide-2026) teams should gate the aha moment, then measure it. [Conversion rate optimization](https://www.enrichlabs.ai/blog/conversion-rate-optimization-ecommerce-complete-guide-2026) on the first-run experience is a churn project, not only a signup project. ### 4\. Staff customer success against NRR, not ticket volume [Customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026) owns health scores, QBR cadence, and expansion. Support owns incidents. Mix them and you get polite churn. ### 5\. Watch usage, then intervene [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters): declining logins, fewer features, plan downgrades, and negative support tone. Trigger a play: training, a cheaper plan, or a save offer. Flexible packaging retains accounts that would otherwise full-cancel. ### 6\. Annual plans where the economics work Longer contracts cut monthly logo churn. They also hide dissatisfaction until renewal. Pair annual billing with usage health so you do not discover a dead account at month 11. ### 7\. Stop buying the wrong customers Channel mix is a churn lever. If paid social or a [branding agency](https://www.enrichlabs.ai/blog/branding-agency-complete-guide-2026) campaign inflates trial volume from non-ICP traffic, logo churn rises even if the product is fine. Tighten the promise on the landing page. Align [email marketing](https://www.enrichlabs.ai/blog/email-marketing-agency-complete-guide-2026) onboarding with the job the buyer hired you for. [B2B marketing automation](https://www.enrichlabs.ai/blog/b2b-marketing-automation-2026-guide) should suppress junk leads, not nurture them forever. ### 8\. Product and packaging Voluntary churn is often price vs value. Downgrade paths, add-on unbundling, and clearer seats beat a generic discount. Community and education help when the product is a workflow, not a one-click utility. [Social media](https://www.enrichlabs.ai/blog/social-media-marketing-agency-complete-guide-2026) and [UGC](https://www.enrichlabs.ai/blog/ugc-marketing-complete-guide-2026) can show real workflows so the wrong buyer self-selects out. ### Tools that actually touch churn - **Enrich Labs (Helena):** acquisition, SEO, ads, and lifecycle content aimed at the ICP you can keep. Start here if paid mix is the leak. [Helena vs an AI marketing agency](https://www.enrichlabs.ai/blog/helena-vs-ai-marketing-agency-2026). - **Stripe Billing:** retries, dunning, portal, analytics for involuntary and plan changes. Source: [Stripe revenue churn](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters). - **Recurly:** network benchmarks plus recovery. Source: [Recurly benchmarks](https://recurly.com/research/churn-rate-benchmarks/). - **A CS platform (example: Planhat):** logo vs revenue metrics in one place. Source: [Planhat](https://www.planhat.com/customer-success/churn-and-retention). Helena can run the content, ads, and lifecycle programs around those plays so CS is not the only team talking to the customer after day 30. * * * ## Churn rate by motion: SaaS, PLG, sales-led, DTC **Self-serve SaaS / PLG.** Expect higher monthly logo churn than enterprise. Use Lenny's SMB band. Cohort curves matter more than last month's blended rate. [PLG](https://www.enrichlabs.ai/blog/product-led-growth-complete-guide-2026) should report activation and week-4 retention next to churn. **Sales-led B2B.** Annual contracts. Report annual logo churn and NRR. [CRV](https://www.crv.com/content/saas-churn-rate) Series A frame applies. Multi-thread the account so a champion exit is not a logo exit. [LinkedIn advertising](https://www.enrichlabs.ai/blog/linkedin-advertising-campaign-complete-guide-2026) that books the wrong title will show up as 12-month logo churn, not as a bad CTR. **DTC subscriptions.** Recurly ecommerce at 4.25% is a different sport. Involuntary churn and impulse signup both run hotter. Dunning and win-back email do more than a CSM. [Shopify SEO](https://www.enrichlabs.ai/blog/shopify-seo-complete-guide-2026), [Amazon DSP](https://www.enrichlabs.ai/blog/amazon-dsp-complete-guide-2026), and [abandoned cart email](https://www.enrichlabs.ai/blog/abandoned-cart-email-complete-guide-2026) change who enters the subscription, which changes who stays. **Local and services.** If you sell software to [local businesses](https://www.enrichlabs.ai/blog/pest-control-marketing-complete-guide-2026) or agencies, SMB death and seasonality show up as unavoidable logo churn. Judge NRR the way Shopify-ecosystem operators do in Lenny's quotes: merchants die, graduates expand. The same pattern hits [HVAC](https://www.enrichlabs.ai/blog/hvac-marketing-complete-guide-2026) and [plumber](https://www.enrichlabs.ai/blog/plumber-marketing-complete-guide-2026) software: winter logos are not summer logos. **Agencies as the customer.** An [SEO agency](https://www.enrichlabs.ai/blog/seo-agency-complete-guide-2026) or [digital marketing agency](https://www.enrichlabs.ai/blog/digital-marketing-agency-complete-guide-2026) on a seat plan churns when their own book of business shrinks. Price on value delivered to their clients, not on seats they will cut first. * * * ## Common mistakes - **Blended SMB and enterprise** into one churn number ([CRV](https://www.crv.com/content/saas-churn-rate) red flag). - **Logos only** when large accounts expand, or **NRR only** when logos are in free fall. - **x12 annualization** of monthly churn, then calling it annual. - **New customers in the denominator** in the same month they join, without saying so. - **Ignoring involuntary churn** because it feels not the product. - **Optimizing CAC** while churn eats LTV. Read the [CAC guide](https://www.enrichlabs.ai/blog/customer-acquisition-cost-complete-guide-2026) next. - **Waiting for a save desk** instead of fixing onboarding. - **Changing the definition** before a fundraise. - **Buying non-ICP traffic** through a [web design agency](https://www.enrichlabs.ai/blog/web-design-agency-complete-guide-2026) rebuild or a paid test that never checks 60-day retention. - **Treating [influencer](https://www.enrichlabs.ai/blog/influencer-marketing-agency-complete-guide-2026) spikes as quality** when 30-day logo churn says otherwise. * * * ## FAQ ### What is a good SaaS churn rate in 2026? Match your band. Recurly's SaaS network median is 3.22%. Lenny's good B2B SMB monthly range is 2.5% to 5%. CRV treats 3% to 5% monthly logo churn as normal below $1M ARR, and 100% NRR as the Series A floor. Enterprise should aim closer to 1% monthly logo or better. ### Is churn rate the opposite of retention rate? For the same logos and window, yes: retention = 1 minus logo churn. Revenue retention is not 1 minus logo churn. Gross and net revenue retention include contraction and expansion. [Planhat](https://www.planhat.com/customer-success/churn-and-retention) is explicit on this inverse for logos only. ### Should I track monthly or annual churn? Both. Monthly for self-serve and leading indicators. Annual for contracted ARR and board reporting. Never compare a monthly Recurly figure to an annual CRV revenue-churn figure without converting carefully. ### What is negative churn? Net revenue churn below zero: expansion MRR exceeds churned MRR. NRR above 100% is the same idea in retention form. [Stripe](https://stripe.com/resources/more/revenue-churn-101-how-to-calculate-it-and-why-it-matters) calls this out in the net revenue churn formula. ### How does churn affect CAC payback? Payback can look stable while LTV collapses, which is CRV's unit-economics point. Payback is cash recovery. LTV is the life of the customer. High churn still forces you to replace the base, which raises fully loaded CAC over time. See [customer lifetime value](https://www.enrichlabs.ai/blog/customer-lifetime-value-complete-guide-2026). ### How much of churn is involuntary? On [Recurly's July 2026 all-industry mix](https://recurly.com/research/churn-rate-benchmarks/), involuntary is 1.25% vs 3.60% total, about one-third. SaaS involuntary is 1.06% of 3.22%. Higher-ARPC cohorts run much lower involuntary (0.18% in Recurly's $250+ group). * * * ## Conclusion Churn rate is the leak in the bucket. Calculate logos and dollars. Split voluntary and involuntary. Benchmark against the motion you actually run, not a blended internet average. Recurly, CRV, and Lenny disagree in the details and agree on the job: NRR, cohorts, and payment recovery beat a single blended percentage. If acquisition is feeding the leak, fix targeting and onboarding with the same energy you spend on save offers. [Enrich Labs](https://www.enrichlabs.ai) built Helena so marketing teams can run ads, SEO, and lifecycle work that attract customers worth keeping. Start there, then let [customer success](https://www.enrichlabs.ai/blog/customer-success-complete-guide-2026) expand the ones who stay.