FoundersGPT Weekly · Issue #1Web edition · not yet sent by email
What 5% monthly churn does to your growth
One founder lesson, one mechanism, one thing to do this week.
This is the full text of the first issue. It has not been sent by email yet, so the replies and the weekly schedule it mentions have not started.
Welcome to FoundersGPT Weekly.
One founder lesson every Tuesday. One mechanism worth understanding. One thing you can do about it this week. Five minutes, no filler.
That's the whole format. Let's start.
THE LESSON
Sometimes a growth problem is a retention problem.
When an early-stage company stalls, a common diagnosis is "we need more customers" and a common treatment is more top-of-funnel. If churn is the constraint, that treatment moves less than expected.
Here is a simple model that shows why.
THE MECHANISM
Why 5% monthly churn caps your company
Take a company that starts with 100 customers. Each month it loses 5% of its existing customers and adds 20 new ones. Hold both numbers constant.
Run it forward and the count climbs, then flattens toward 400.
Why 400: each month you lose 5% of your customers and gain 20. The count stops moving when the two are equal, when customers × 0.05 = 20. So customers = 20 ÷ 0.05 = 400. Below 400 you add more than you lose and the count rises; the gap to 400 shrinks by 5% a month, so the count gets closer every month without reaching it.
That level, new customers ÷ churn rate, is the ceiling under these assumptions. It holds only while both inputs stay constant. Change either one and the ceiling moves.
Now cut churn to 2%. Same 20 new customers a month. The ceiling becomes 20 ÷ 0.02 = 1,000.
You did not add a single extra new customer a month. You changed one input and the level the company converges on became 2.5 times larger. (At 2% churn it also takes longer to get there.)
So "we need more leads" is not automatically the right answer. Doubling acquisition doubles the ceiling. Halving churn doubles it too. Which is cheaper depends on your business, but it is worth costing both before defaulting to acquisition, and your existing customers can tell you something about why they left.
The uncomfortable version: if churn stays at 5% and you keep adding 20 customers a month, a plan to reach 1,000 customers does not add up. Under those inputs the count settles at 400. Reaching 1,000 means changing an input.
THE FRAMEWORK
The exit interview nobody runs
Before a retention strategy, find out why people leave, and when.
Four questions, sent by you personally, to every customer who cancels:
- What were you trying to do when you signed up?
- What did you end up doing instead?
- Was there a specific moment you decided to stop?
- What would have had to be true for you to stay?
Question 3 is often the most useful. Churn can feel like slow drift, but customers can sometimes name a moment — a bug, an invoice, a competitor's demo, a week they did not log in and nothing bad happened.
When they can, that moment tells you where to look, and it costs one email.
Keep any discount out of this email. You are asking for information, and an offer in the same message makes the answers harder to read.
THIS WEEK
One thing to do
Calculate your ceiling. New customers per month ÷ monthly churn rate.
If that number is smaller than the company you are describing to investors, the plan needs a different input: lower churn, more new customers, or both.
If you do not have enough data to calculate it yet, that is the answer too: instrument it this week, before the number matters.
That's issue one.
Reply and tell me what your ceiling came out at. I read every reply, and right now there are few enough of you that I mean that literally.
See you Tuesday.
— FoundersGPT
Intelligence for builders.
Get the next issue by email
Free. Unsubscribe in one click. Or read the library.