The one number that caps how big your company can get
Divide your new revenue each month by your churn rate. That number is the size your company stops growing at — and no marketing channel, however good, can push you past it.
There is a chart almost every subscription business eventually produces. Revenue climbs for eighteen months, then goes flat, and stays flat, while everybody keeps doing the same amount of work. Marketing is landing customers. Sales is closing. Nothing broke. The line simply stopped.
The explanation is one division, and it takes about ten seconds:
MRR ceiling = new MRR per month ÷ monthly churn rate
Here is Jason Cohen running it live on a business doing $3,000 a month in new recurring revenue at 9% monthly churn:
So 3K of MRR comes in, but at 9%, so divided by 9%, and you get about 30, 40k in MRR, which is about where you are. And that's why this chart was flat, because you brought in 3K, but 3K left.
Jason Cohen
The mechanism is worth stating slowly, because it is the part that surprises people:
If marketing is adding one K of MRR per month in new revenue, but cancellation is nine percent — as the company gets bigger and bigger, nine percent is nine percent of a bigger and bigger number. But marketing is still putting in 1K, 1K, 1K each month. So at some point, you're at a size where the number of customers or revenue that walks out the door in churn is equal to the amount that marketing is bringing in. And at that point the company is literally not growing.
Jason Cohen
Churn is a percentage of a growing base. New revenue is an absolute number produced by a team of fixed size. One of those scales with you and the other one doesn't, so they always meet. Where they meet is your ceiling, and you can compute it today.
Why this is not a health metric
Most dashboards file churn next to NPS and support response time — things you keep an eye on. It doesn't belong there. It belongs next to your headcount plan, because it decides the maximum size of the company you are building.
Below the ceiling, every hour of acquisition work compounds: customers stack up and the base grows. At the ceiling, the identical hour of acquisition work is a treadmill — you are replacing people who left, at full cost, forever. The work looks the same from inside. The outcome is completely different.
Which means the most important thing about the number is when you compute it. At forty customers, changing churn means changing a product that forty people have built habits around. At zero customers it means choosing a different cohort, which is free.
Run it on yourself
| New revenue per month | 8% churn | 5% churn | 3% churn | 2% churn |
|---|---|---|---|---|
| $1,000 | ~$12.5K | ~$20K | ~$33K | ~$50K |
| $2,000 | ~$25K | ~$40K | ~$66K | ~$100K |
| $5,000 | ~$62.5K | ~$100K | ~$167K | ~$250K |
Read across a row rather than down a column. The row is what your acquisition effort produces; the columns are what happens to it. Halving churn does more for the ceiling than doubling the sales team, and it is usually much cheaper.
Two more pieces of arithmetic fall out of the same number and are worth having in your head:
- Average customer lifetime, in months, is 1 ÷ monthly churn. At 5% that is twenty months. At 10% it is ten — and roughly 90% of a cohort is gone inside eight.
- Your payback period has to fit inside that lifetime. A four-month payback on a ten-month customer is a business; on a five-month customer it is a hobby with invoices.
What counts as bad
Rob Walling's published ladder for monthly logo churn in B2B SaaS is blunt and easy to remember:
| Monthly churn | Verdict |
|---|---|
| under 2% | great |
| around 7–8% | "company on fire" |
| over 10% | catastrophic |
The gap between "great" and "on fire" is six percentage points. It does not feel like much on a dashboard. It is the difference between a $100K/month ceiling and a $25K one at identical sales effort.
The reason to trust this over most business advice
Walling arrives at the same formula independently, in a different year, about a different company — $4,000 in new MRR ÷ 4% churn = a $100,000/month ceiling — and adds the line that removes the mystique from the whole phenomenon:
There is no magic revenue level at which companies plateau — only this arithmetic.
Two operators who built at very different scales, working from different data, reaching for the same instrument first. That is about as close to consensus as this field produces, and it is a much stronger reason to take the number seriously than any single person's opinion about it.
What to do this week
- Compute it. New MRR added last month, divided by last month's churn rate. One line. If you cannot produce both numbers, that is the finding — go and instrument the second one, because you already have the first.
- Compare it to your goal. If the ceiling is below the number you are planning around, no channel fixes it. Stop reading channel advice and go and look at who is leaving.
- Split churn by cohort before you react to it. A 10% blended rate that lives entirely in one segment is not a churn problem, it is a targeting problem wearing a churn costume — and the fix is the list, not the product.
- If you are pre-revenue, choose the cohort with the ceiling in mind. The cheapest possible moment to fix churn is before anyone has churned.
The uncomfortable version: if you are flat and busy, you are almost certainly at your ceiling, and every additional hour of prospecting is being spent to stand still. The way out is not more of it. The way out is a base that leaks less, which begins with picking different people to sell to.
Keep reading
- One channel is no longer a strategy The playbook that said pick a channel and go deep was written for a search engine that judged you on your own pages. Three things changed at once, and the minimum viable channel mix went from one to three.
- Treat the first call as a research interview you also happen to be selling on You will lose most first meetings. The ones you lose are worth something only if you left with an answer you did not have before — which requires deciding, before you dial, what you are trying to learn.
- How long to give an idea before you are allowed to conclude anything A week is not enough and six months is too long, and the number matters less than the thing it protects you from: quietly retreating into the work that feels productive while the question you were testing goes unanswered.