A founder celebrates 200 new signups this month. Last month, 180 users quietly cancelled. Net growth: 20 users. But the dashboard shows signups, not exits, so the response is predictable: more ads, more content, more top-of-funnel spend.
This is backwards. Churn, meaning the percentage of customers who leave in a given period, is almost never a marketing problem. Marketing did its job. People showed up. They tried the product and decided it was not worth staying. That decision happened inside your product. That is where the fix lives.
Every dollar you spend acquiring users into a leaky product is a dollar spent renting customers you will never keep. Plug the hole first. Then pour.
Most founders calculate churn wrong, and the wrong number always looks better than the real one. Here is the honest version.
Customer churn rate. Take the customers you lost this month. Divide by the customers you had at the start of the month. Do not include new signups in the denominator. If you started with 100 customers, lost 8, and gained 30, your churn is 8%. Not 8 divided by 130.
Revenue churn rate. Take the monthly recurring revenue (MRR, your predictable subscription income) lost to cancellations and downgrades. Divide by MRR at the start of the month. Revenue churn matters more than customer churn because losing a $500 account hurts five times more than losing a $100 account. If the two numbers diverge a lot, find out which segment is leaving.
Gross vs. net. Net revenue churn subtracts expansion revenue from upgrades. Net negative churn, where upgrades outpace losses, is the holy grail. But track gross churn separately. Expansion revenue can hide a retention problem for months before it catches up with you.
Here are the benchmarks for early-stage B2B SaaS:
- Under 3% monthly churn. Strong. Your product is sticky. Consider pushing harder on growth.
- 3% to 5% monthly. Acceptable for early stage. Fixable with onboarding work.
- 5% to 8% monthly. Warning zone. Growth will stall around 1,000 customers no matter what you spend.
- Over 8% monthly. The product is broken for your market. Stop acquisition spend today.
The brutal math: at 8% monthly churn, you replace your entire customer base every year. At 3%, customers stick around for nearly three. Same acquisition effort, triple the business.
A signup is a promise. Retention is the proof. Plot the percentage of each month's cohort that is still active at week 1, week 4, week 8, week 12. That curve tells you everything.
Three shapes matter:
- The cliff to zero. Users try once and never return. The curve drops to zero within weeks. No product-market fit. No amount of marketing saves this.
- The flattening curve. Users drop early, then the curve levels off above zero. That flat line is your retained core. Even 15% flattening is a foundation you can build on.
- The smile curve. Retention drops, then climbs back up as users deepen usage or invite teammates. This is rare and it means you have something genuinely compounding.
Track curves by cohort, not blended averages. Blended numbers hide decay. If your January cohort retained better than your June cohort, something you shipped made the product worse, and only cohort analysis will show it.
For early B2B SaaS, a healthy target is 40% to 60% of a cohort still active at day 30. Below 30% means most users never experienced the value. That points directly at onboarding, not features.
Churned users are the most honest source of product feedback you will ever get. They already left, so they have nothing to protect. Ask them why.
Email every churned user within 48 hours of cancellation. Keep it short. Three questions work:
- What were you hoping the product would do for you?
- Where did it fall short?
- What would have made you stay?
Do not pitch a discount. Do not try to win them back during the interview. That converts a learning conversation into a sales call, and you will learn nothing.
Aim for 10 to 15 interviews, then read the answers together. Individual answers are anecdotes. Patterns across ten answers are your roadmap. You will usually find one of three causes: they never understood what the product did (onboarding failure), they understood but waited too long to see value (time-to-value failure), or a specific missing capability mattered more than you thought (an actual feature gap).
Most founders assume the third. The interviews almost always reveal the first or second.
When churn is high, the instinct is to build more. Resist it. Fix in this order, because each layer depends on the one before it.
First, onboarding. The majority of churn happens in the first 30 days, and most of that happens in the first session. Audit your signup-to-first-action path ruthlessly. Every field, every step, every screen that is not essential to delivering value is a place users quit. If 60% of new signups never complete the core action, that is your churn problem right there.
Second, time-to-value. Time-to-value is how long it takes a new user to experience the outcome your product promises. Minutes are good. Days are fatal. Slack figured out that teams sending 2,000 messages stuck around, so everything pushed toward that threshold. Define your equivalent milestone and shorten the path to it. If your product needs a week of setup before anything useful happens, no feature will save it.
Third, features. Only after onboarding works and value lands fast should you build what churned users asked for. Features added to a broken first experience just give users more things to be confused by.
The test is simple. Watch five strangers sign up and use your product without help. Wherever they hesitate is where your churn is born.
- Calculate your real churn rate. Customers lost last month divided by customers at the start of the month. No new signups in the denominator. Write the number where you can see it.
- Plot one retention curve. Take last month's cohort and mark who is still active today. If more than half are gone, your next sprint is onboarding, not acquisition.
- Email three churned users. Send the three questions from this post. Do not sell. Just ask, listen, and write down their exact words.
Signups make you feel good. Retention makes you a business. Founders in the 52Waypoint community share their churn numbers, their exit interview answers, and the onboarding fixes that actually moved the curve. Bring yours.
A founder celebrates 200 new signups this month. Last month, 180 users quietly cancelled. Net growth: 20 users. But the dashboard shows signups, not exits, so the response is predictable: more ads, more content, more top-of-funnel spend.
This is backwards. Churn, meaning the percentage of customers who leave in a given period, is almost never a marketing problem. Marketing did its job. People showed up. They tried the product and decided it was not worth staying. That decision happened inside your product. That is where the fix lives.
Every dollar you spend acquiring users into a leaky product is a dollar spent renting customers you will never keep. Plug the hole first. Then pour.
Most founders calculate churn wrong, and the wrong number always looks better than the real one. Here is the honest version.
Customer churn rate. Take the customers you lost this month. Divide by the customers you had at the start of the month. Do not include new signups in the denominator. If you started with 100 customers, lost 8, and gained 30, your churn is 8%. Not 8 divided by 130.
Revenue churn rate. Take the monthly recurring revenue (MRR, your predictable subscription income) lost to cancellations and downgrades. Divide by MRR at the start of the month. Revenue churn matters more than customer churn because losing a $500 account hurts five times more than losing a $100 account. If the two numbers diverge a lot, find out which segment is leaving.
Gross vs. net. Net revenue churn subtracts expansion revenue from upgrades. Net negative churn, where upgrades outpace losses, is the holy grail. But track gross churn separately. Expansion revenue can hide a retention problem for months before it catches up with you.
Here are the benchmarks for early-stage B2B SaaS:
- Under 3% monthly churn. Strong. Your product is sticky. Consider pushing harder on growth.
- 3% to 5% monthly. Acceptable for early stage. Fixable with onboarding work.
- 5% to 8% monthly. Warning zone. Growth will stall around 1,000 customers no matter what you spend.
- Over 8% monthly. The product is broken for your market. Stop acquisition spend today.
The brutal math: at 8% monthly churn, you replace your entire customer base every year. At 3%, customers stick around for nearly three. Same acquisition effort, triple the business.
A signup is a promise. Retention is the proof. Plot the percentage of each month's cohort that is still active at week 1, week 4, week 8, week 12. That curve tells you everything.
Three shapes matter:
- The cliff to zero. Users try once and never return. The curve drops to zero within weeks. No product-market fit. No amount of marketing saves this.
- The flattening curve. Users drop early, then the curve levels off above zero. That flat line is your retained core. Even 15% flattening is a foundation you can build on.
- The smile curve. Retention drops, then climbs back up as users deepen usage or invite teammates. This is rare and it means you have something genuinely compounding.
Track curves by cohort, not blended averages. Blended numbers hide decay. If your January cohort retained better than your June cohort, something you shipped made the product worse, and only cohort analysis will show it.
For early B2B SaaS, a healthy target is 40% to 60% of a cohort still active at day 30. Below 30% means most users never experienced the value. That points directly at onboarding, not features.
Churned users are the most honest source of product feedback you will ever get. They already left, so they have nothing to protect. Ask them why.
Email every churned user within 48 hours of cancellation. Keep it short. Three questions work:
- What were you hoping the product would do for you?
- Where did it fall short?
- What would have made you stay?
Do not pitch a discount. Do not try to win them back during the interview. That converts a learning conversation into a sales call, and you will learn nothing.
Aim for 10 to 15 interviews, then read the answers together. Individual answers are anecdotes. Patterns across ten answers are your roadmap. You will usually find one of three causes: they never understood what the product did (onboarding failure), they understood but waited too long to see value (time-to-value failure), or a specific missing capability mattered more than you thought (an actual feature gap).
Most founders assume the third. The interviews almost always reveal the first or second.
When churn is high, the instinct is to build more. Resist it. Fix in this order, because each layer depends on the one before it.
First, onboarding. The majority of churn happens in the first 30 days, and most of that happens in the first session. Audit your signup-to-first-action path ruthlessly. Every field, every step, every screen that is not essential to delivering value is a place users quit. If 60% of new signups never complete the core action, that is your churn problem right there.
Second, time-to-value. Time-to-value is how long it takes a new user to experience the outcome your product promises. Minutes are good. Days are fatal. Slack figured out that teams sending 2,000 messages stuck around, so everything pushed toward that threshold. Define your equivalent milestone and shorten the path to it. If your product needs a week of setup before anything useful happens, no feature will save it.
Third, features. Only after onboarding works and value lands fast should you build what churned users asked for. Features added to a broken first experience just give users more things to be confused by.
The test is simple. Watch five strangers sign up and use your product without help. Wherever they hesitate is where your churn is born.
- Calculate your real churn rate. Customers lost last month divided by customers at the start of the month. No new signups in the denominator. Write the number where you can see it.
- Plot one retention curve. Take last month's cohort and mark who is still active today. If more than half are gone, your next sprint is onboarding, not acquisition.
- Email three churned users. Send the three questions from this post. Do not sell. Just ask, listen, and write down their exact words.
Signups make you feel good. Retention makes you a business. Founders in the 52Waypoint community share their churn numbers, their exit interview answers, and the onboarding fixes that actually moved the curve. Bring yours.