Why Your SaaS Churn Rate Looks Fine But MRR Is Still Shrinking
Your churn rate looks fine but MRR keeps shrinking. Here is why customer churn rate and MRR churn rate are not the same thing, what net revenue retention actually tells you, and how to fix the four hidden patterns draining your revenue.
Why Your SaaS Churn Rate Looks Fine But MRR Is Still Shrinking
Your churn rate is sitting at 3%. Maybe even 2%. You check it every Monday and it looks healthy. And yet somehow, every month, your MRR dashboard tells a different story. Revenue is flat. Or worse, quietly going down. You add new customers, lose roughly the same number, and the number just refuses to move.
This is one of the most confusing and demoralizing places to be as a SaaS founder. And it is more common than you think.
The problem is not your churn rate. The problem is that you are measuring the wrong thing.
Churn Rate and MRR Are Not the Same Thing
Most founders track customer churn rate, which is the percentage of customers who cancel in a given month. It is clean, easy to calculate, and gives you a single number to point at.
But customer churn rate treats a $29/month customer and a $299/month customer as identical. When the $299 customer cancels and three $29 customers sign up, your churn rate looks great. Your MRR does not.
This is the core disconnect. Customer churn rate measures headcount. Revenue churn rate measures money. And the two can move in completely opposite directions at the same time.
Customer Churn Rate Formula
| Metric | Formula | What It Tells You |
|---|---|---|
| Customer Churn Rate | Customers Lost / Customers at Start × 100 | How many accounts you are losing |
| MRR Churn Rate | MRR Lost / MRR at Start of Month × 100 | How much revenue you are losing |
| Net MRR Churn Rate | (MRR Lost minus Expansion MRR) / MRR at Start × 100 | Whether you are actually growing |
If you are only tracking the first one, you are flying blind on the other two.
The Real Reason MRR Shrinks When Churn Looks Fine
There are four patterns that cause this. Most founders are experiencing at least two of them simultaneously without realizing it.
Your High-Value Customers Are Leaving Quietly
When you look at churn by customer count, every cancellation weighs the same. But when you look at which customers are actually leaving, a pattern often emerges. Your biggest accounts, the ones on your $199 or $399 plan, are cancelling. And they are being replaced by customers on your $29 starter plan.
Your headcount stays stable. Your average contract value drops. MRR shrinks.
This is called revenue churn and it is invisible until you segment your cancellations by plan tier. If you have never done that analysis, do it today. Pull your last 90 days of cancellations and sort by MRR value. The answer is usually sitting right there.
Expansion Revenue Has Quietly Stopped
Healthy SaaS businesses grow MRR from two sources. New customers coming in, and existing customers upgrading, buying add-ons, or expanding their seats. That second source is called expansion MRR and it is what creates net negative churn, the holy grail where your existing base grows faster than you lose customers.
When expansion MRR dries up and you are relying entirely on new customer acquisition to grow, any slowdown in sales immediately shows up as flat or declining MRR even if your churn rate is perfectly normal.
Ask yourself honestly: when did a customer last upgrade their plan without you personally asking them to?
You Have a Downgrade Problem You Are Calling Something Else
Downgrades are the silent cousin of cancellations. When a $199 customer drops to your $49 plan, they do not show up in your churn rate at all. They are still a customer. But you just lost $150 in monthly recurring revenue.
Across a large enough customer base, downgrade churn can represent more lost MRR than actual cancellations. It is almost never tracked separately. Most CRMs and billing dashboards show you cancellations clearly and bury downgrade data somewhere you never look.
Involuntary Churn Is Rotting Your MRR From the Inside
Research consistently shows that between 20 and 40 percent of all SaaS churn is involuntary, meaning the customer did not intend to leave. Their card expired. A payment failed. Your dunning sequence sent a few emails, they missed them, and the account got cancelled.
These customers wanted to stay. They are the easiest wins you are leaving on the table. And unlike voluntary churners who made a considered decision to leave, involuntary churners often come back immediately if you catch them at the right moment.
If you are not running a proper dunning sequence with smart retries, in-app notifications, and SMS fallback, you are probably losing 5 to 8 percent of your MRR to this every single month.
The Metric You Should Actually Be Watching
Net Revenue Retention (NRR), sometimes called Net Dollar Retention, is the single most important metric for understanding whether your SaaS business is actually healthy.
NRR Formula
NRR = (Starting MRR + Expansion MRR minus Contraction MRR minus Churned MRR) / Starting MRR × 100
| NRR Score | What It Means |
|---|---|
| Above 120% | World class. Your existing base grows faster than you lose customers. |
| 100 to 120% | Healthy. Expansion is offsetting churn well. |
| 90 to 100% | Danger zone. You are losing ground even if headcount looks stable. |
| Below 90% | Revenue is actively shrinking. Needs immediate attention. |
Best-in-class SaaS companies like Snowflake and Datadog have historically reported NRR above 130%. That means even with zero new customers, their revenue would still grow. That is the compounding engine that makes SaaS so valuable when it works.
If your NRR is below 100%, you cannot grow your way out of the problem with more sales. You are filling a leaky bucket and the hole is bigger than you think.
How to Actually Diagnose What Is Happening
Before fixing anything, you need to know which of these problems you actually have. Here is how to find out in one afternoon.
Pull your last 90 days of MRR data and segment it into four buckets. New MRR from new customers. Expansion MRR from upgrades and add-ons. Contraction MRR from downgrades. Churned MRR from cancellations. If your billing tool does not give you this breakdown natively, export to a spreadsheet and do it manually. It is worth the two hours.
Once you have those four numbers, calculate your NRR for the last three months. If it is below 100%, you have a retention problem regardless of what your customer churn rate says. If it is above 100% but growing slowly, you probably have an expansion problem, your existing customers are not growing their accounts.
Then look at cancellations by plan tier for the last 60 days. Are higher-tier customers churning at a higher rate than lower-tier customers? If yes, you have a value delivery problem at the top of your plan structure, and that is a product and onboarding problem more than a sales problem.
The Part Most Founders Get Wrong When Trying to Fix This
The instinct when MRR is shrinking is to push harder on acquisition. Spend more on ads. Hire another salesperson. Launch on Product Hunt again.
That is the wrong move if your NRR is below 100%. Every new customer you acquire will eventually churn at the same rate as your current base. You are accelerating the treadmill, not getting off it.
The highest-leverage thing you can do when MRR is shrinking despite reasonable churn numbers is to fix the moment of cancellation itself.
Most SaaS products treat the cancel button as a dead end. User clicks cancel, a confirmation dialog appears, they confirm, they are gone. No conversation. No understanding of why. No last attempt to save them or at minimum learn what broke.
That 30-second window when someone is about to cancel is the most valuable moment in your entire customer relationship. It is the one moment where the user is actually thinking about your product, has a specific reason for leaving, and is still on your platform.
A guided exit conversation at that exact moment, not a survey email three days later, captures things no analytics tool will ever show you. The integration they needed that you never built. The price concern they never raised with support. The bug that annoyed them for three weeks until they finally gave up.
Companies that add a real-time exit conversation to their cancel flow typically see two things happen. They save 15 to 25 percent of users who were about to churn. And they collect enough product feedback in 30 days to reprioritize their entire roadmap.
What a Healthy SaaS Revenue Retention Stack Looks Like
| Layer | What to Fix | Expected Impact |
|---|---|---|
| Involuntary churn | Smart dunning with card retries, in-app alerts, SMS | Recover 30 to 50% of failed payments |
| Cancellation flow | Real-time exit conversation at the cancel button | Save 15 to 25% of churning users |
| Downgrade prevention | Show value before renewal, offer pause option | Reduce contraction MRR by 10 to 20% |
| Expansion revenue | Behavior-triggered upgrade prompts based on usage | Add 5 to 15% NRR from existing base |
| At-risk detection | Health scoring to flag drifting users before cancel | Proactive saves before the cancel click |
The order matters. Fix involuntary churn first because it is the fastest win with the least effort. Then fix your cancellation flow because it catches everyone else. Then work upstream on health scoring and expansion.
Most founders do this in reverse order. They spend months building health score dashboards while losing 8% of MRR every month to failed payments and a cancellation flow that is just a confirmation dialog.
The Numbers That Should Make You Act Today
If you have 500 customers paying an average of $99 per month, your MRR is roughly $49,500. At 5% monthly churn, you lose about 25 customers and $2,475 in MRR each month.
If 30% of that churn is involuntary, that is $742 per month in MRR that is recoverable with a proper dunning sequence. Over a year that is $8,900 in revenue you are just leaving behind.
If a cancellation flow with a real exit conversation saves 20% of your remaining churners, that is another 5 customers per month kept. At $99 average and 8 months average lifetime, each saved customer is worth $792 in retained LTV. Five per month is $3,960 in recovered LTV monthly.
Combined, those two fixes alone, better dunning and a real cancellation conversation, are worth over $50,000 in annual revenue impact at that scale. Not from new sales. From customers you already had.
Frequently Asked Questions
What is the difference between customer churn rate and MRR churn rate?
Customer churn rate counts the percentage of accounts that cancelled. MRR churn rate measures the percentage of revenue that was lost. They can move in opposite directions when higher-value customers churn at different rates than lower-value ones. MRR churn rate is always the more important number for understanding business health.
What is a good net revenue retention rate for SaaS?
Above 100% NRR means your existing customers are growing their spend faster than you lose revenue to churn and downgrades. Above 120% is considered world-class. Below 100% means revenue is shrinking from your existing base regardless of new sales.
What causes SaaS MRR to shrink even with low churn?
The most common causes are high-value customers churning at above-average rates, expansion MRR drying up, downgrades that do not show in customer churn metrics, and involuntary churn from failed payments. Any one of these can cause NRR to drop below 100% while your customer churn rate looks healthy.
How do I calculate net revenue retention?
Take your MRR at the start of the month. Add any expansion MRR from upgrades. Subtract contraction MRR from downgrades. Subtract churned MRR from cancellations. Divide by your starting MRR and multiply by 100. A result above 100% means your existing base is growing.
What is the fastest way to improve MRR retention?
Fix involuntary churn first with smart payment retries and in-app card expiry alerts. Then add a real conversation to your cancellation flow to understand and save voluntary churners. Together these two changes typically improve NRR by 5 to 15 percentage points within 60 days.
What is a SaaS cancellation flow?
A cancellation flow is the designed experience between a user clicking cancel and their account actually closing. A good cancellation flow opens a real-time conversation to understand why the user is leaving, offers relevant alternatives like a pause or downgrade, and collects structured feedback even when the user decides to leave anyway.
Flidget catches users at the exact moment they click cancel and opens a real conversation — so you finally know why they are leaving and have one last chance to keep them. Drift detection also flags at-risk users before they ever reach the cancel button. Start free at flidget.com