Short answer: If new customers keep arriving but revenue stays flat, churn is very likely cancelling out your growth. Every business has a ceiling where the number of customers lost each month equals the number won, and you can calculate it: divide new customers per month by your monthly churn rate. If your current base is close to that number, more marketing will not move revenue until you reduce churn.
You fixed the funnel. Leads are coming in, the sales team is busy, and the board slides show new logos every month. Then finance closes the quarter and the revenue line has barely moved. Nobody did anything wrong this month, which is what makes it so frustrating. The problem is arithmetic, and it has been building quietly for a while.
This article explains that arithmetic, shows you how to diagnose it in your own numbers, and sets out what actually reduces churn, in software and beyond.
Key takeaways
- Churn creates a growth ceiling: new customers per month divided by monthly churn rate. Past that point, growth stops.
- Logo churn and revenue churn tell different stories. Track both.
- Involuntary churn from failed payments is the fastest kind to fix, because the customer never chose to leave.
- The median B2B SaaS company kept 84% of its revenue base before expansion in 2025, according to Benchmarkit. Expansion is what pushes net retention above 100%.
- Making cancellation hard is not a strategy. The FTC's click-to-cancel rule was vacated in 2025, but state laws and other regulations still apply, and friction only delays the loss.
What is the maths behind flat revenue despite new leads?
Flat revenue despite new leads usually means you have reached your churn ceiling. Here is how to find yours.
Growth ceiling = new customers per month ÷ monthly churn rate
Suppose you win 40 new customers a month and lose 3% of your customers each month.
| Current customers | Lost per month (3%) | New per month | Net change |
|---|---|---|---|
| 500 | 15 | 40 | +25 |
| 1,000 | 30 | 40 | +10 |
| 1,333 | 40 | 40 | 0 |
| 1,500 | 45 | 40 | −5 |
At 1,333 customers, you are running to stand still. The same marketing that produced healthy growth at 500 customers produces nothing at 1,333. Nothing changed in the funnel. The base simply got big enough that a steady percentage loss became a large absolute number.
This is why churn hides so well in early stages. When a company is small, new customers swamp losses. By the time the losses catch up, the team has usually scaled acquisition spend on the assumption that the curve would keep rising.
What is the difference between logo churn and revenue churn?
Logo churn counts customers who leave. Revenue churn counts the money they took with them, including downgrades.
They can tell very different stories. Losing ten small accounts may barely dent revenue. Losing one large account can wipe out a quarter. Track both, and add net revenue retention (NRR), which includes expansion from customers who stayed.
Benchmarkit's 2026 report, covering 342 B2B SaaS and AI companies in 2025, put median gross revenue retention at 84%, down from 88% the year before. Its 2025 report put median net revenue retention at 101%. Read together, those figures show how the typical company was losing around a sixth of its revenue base before expansion and only just making it back. If your expansion slows, the base starts to shrink.
Why does fast acquisition hide a leaky bucket?
Early growth hides churn because the numbers look good in aggregate. Fifty signups a month feels like momentum, even if twenty of them stop using the product within weeks. The losses only become visible when leads start drying up and new volume can no longer cover them.
How do you know if churn is eating your marketing budget?
Is your payback period longer than your customers stay?
This is the most important single test. If it takes 16 months of a customer's payments to recover what you spent acquiring them, and the average customer leaves after 12, every new customer loses you money.
Benchmarkit's 2026 report put median CAC payback at 16 months, an improvement from 18 months the previous year. Compare your own payback period with your median customer lifetime. If lifetime is shorter, scaling acquisition makes the problem worse, which is the core argument in our guide to whether to fix retention or acquisition first.
Is your marketing paying for turnover?
Look at the last four quarters and split new revenue into two buckets: revenue that replaced churned customers, and revenue that grew the business. If most of your new revenue is replacement, your marketing budget is funding turnover. That is when founders start to feel that business growth has stalled despite everyone working harder.
Are your attribution models flattering acquisition?
Most attribution looks at the first conversion and stops. It counts a customer as a win the moment they sign, even if they leave three months later. If you compare channels on cost per acquisition alone, a channel that brings in customers who churn fast will look better than it is. Compare channels on retained revenue after six or twelve months instead. Our piece on content marketing ROI explains how to measure this honestly.
[DIAGRAM BRIEF] Line chart showing customer count rising then flattening at the churn ceiling, with new customers per month held constant. Alt text: Chart showing how churn cancelling growth creates a ceiling on customer count.
How do you tell voluntary and involuntary churn apart?
Voluntary churn is when a customer decides to leave. Involuntary churn is when they did not decide anything: a card expired, a payment failed, retries ran out, and the subscription lapsed.
Recurly's churn benchmarks report the two separately for exactly this reason. The fixes have nothing in common. Voluntary churn is a product, value or fit problem. Involuntary churn is a billing problem.
Can failed payments be recovered automatically?
Often, yes. Most modern billing platforms offer three tools that work together:
- Smart retries that reattempt failed charges at times more likely to succeed.
- Card account updaters, which pull new card details from the card networks when a card is replaced.
- Dunning emails that tell the customer plainly what failed and give them a one-click way to fix it.
If you use Stripe, Chargebee, Recurly or similar, check that all three are switched on. It is common to find that one of them was never configured.
How do you spot billing problems before they become churn?
Watch for webhooks failing silently between your billing system and your product. If a renewal fails and nobody's system notices, the customer can lose access without anyone telling them. Set an alert for failed renewals and review it weekly.
What does churn look like outside software?
Churn is easiest to see in subscription software because billing makes it visible. In other businesses it hides inside "customers we never heard from again". RankedTag's measured result is in B2B SaaS only, so the table below is a measurement guide rather than benchmark data.
| Business type | What churn looks like | How to measure it |
|---|---|---|
| Professional services and law firms | Clients who finish a matter and are never contacted again | Clients with no new matter or contact in 12 months |
| Home services | Emergency customers who call a different firm next time | Share of past customers who rebook within 12 to 24 months |
| Clinics | Patients who miss recalls and do not return | Recall attendance rate by treatment type |
| Ecommerce | First-time buyers who never place a second order | Share of buyers who reorder within your normal repurchase window |
The fix in each case is the same shape: stay in touch between transactions with something useful. A quarterly legal update, a seasonal maintenance reminder, a recall message, a replenishment email. Our guide to growing revenue from existing customers covers each of these in practical detail.
Get the 52-check Founder Review
Submit your domain and Bhushan personally runs 52 checks across crawlability, schema, AI readiness, content depth and conversion path. You get a personal reply on LinkedIn within 48 hours, including when the answer is no.
Apply for a free Founder Review →
What actually reduces churn?
Map every cancellation to a reason
Add a required reason code to every cancellation in your CRM, and review the reasons monthly. Then look at what churned accounts did in their first 90 days. Patterns usually appear quickly: customers who never completed setup, never invited a teammate, or never used the feature they bought the product for.
Set honest expectations before the sale
Many early cancellations are mismatched expectations. Marketing that overpromises brings in customers who were never going to stay. Clear positioning, honest pricing pages and content that explains who the product is not for all reduce churn before the customer signs. This is also why we publish what we cannot claim on our own research hub.
Be careful with heavy introductory discounts
Steep first-year discounts can attract customers who value the price more than the product. When the full-price renewal arrives, many leave. If you discount, discount for commitment (annual plans) rather than for signing up.
Review your pricing model
Benchmarkit's 2026 report found median NRR of 108% for usage-based companies against 98% for seat-based companies. Pricing that grows with the value a customer gets gives them less reason to leave. Read our guide to product pricing strategy before changing anything.
Make cancellation simple and learn from it
The FTC's "click-to-cancel" rule, finalised in October 2024, was vacated by the Eighth Circuit on 8 July 2025. That does not make difficult cancellation safe. The Restore Online Shoppers' Confidence Act and many state automatic-renewal laws still apply in the US, and other markets have their own subscription rules. More importantly, a customer trapped by a hard cancellation flow still leaves eventually, usually angrier and less likely to recommend you. A simple flow with a short exit survey gives you something far more valuable: the reason.
Keep your analytics history
GA4 keeps detailed event and user data for Explorations for a maximum of 14 months on standard properties, and the default is only two months. Standard reports keep aggregated data, but cohort analysis needs the detail. Set retention to 14 months and turn on the free BigQuery export so you can study churn over longer cycles.
Your 10-step churn diagnosis
- Calculate your growth ceiling: new customers per month ÷ monthly churn rate.
- Compare your current customer count with that ceiling.
- Track logo churn, gross revenue retention and net revenue retention separately.
- Compare CAC payback with median customer lifetime.
- Split last year's new revenue into replacement and growth.
- Separate voluntary and involuntary churn.
- Switch on smart retries, card updaters and dunning emails.
- Add a required cancellation reason code and review it monthly.
- Set GA4 retention to 14 months and enable the BigQuery export.
- Compare acquisition channels on retained revenue, not first conversion.