1. Your subscriptions export
Stripe Dashboard → Billing → Subscriptions → Export (all columns). Needs a start or created date and an amount; canceled or ended dates, interval, quantity, customer and plan columns are used when present. Other tools' exports work if the headers are similar.
Choose a file or paste CSV text, then click Analyze.
Result
MRR waterfall by month
| Month | Start MRR | New | Churned | End MRR | Customers | Logo churn |
|---|
Cohort retention (share of each start month still subscribed)
Rows are the month a subscription started. Columns are months since. Green is retained; a cell is blank until that month has happened.
What would a price rise do?
Uses the active subscriptions and each plan's own recent churn from your file. The two sliders are your assumptions, not predictions: nobody can know how many customers a rise will lose until it happens.
What is lower churn worth?
Projects MRR 12 months out at your recent churn and recent new MRR, then again with churn lower. The slider is your assumption, not a prediction.
The model, and what it leaves out
- MRR per subscription is the export's amount normalised to a month (yearly ÷ 12, weekly × 52/12) times quantity when present, held constant for the subscription's life.
- Waterfall. Start MRR + new − churned = end MRR every month. A subscription that starts and ends in the same month counts in both, so the table always reconciles.
- Logo churn is subscriptions ended in a month divided by subscriptions active at its start; the summary uses the last three complete months.
- Lifetime is 1 ÷ monthly logo churn and revenue LTV is average revenue per subscription × lifetime. It is revenue, not profit, and is unreliable when churn is near zero.
- Price-rise simulator. Takes the subscriptions active now on one plan, applies your rise, removes the share you say cancels at once, then decays the rest at that plan's own recent monthly churn (plus any lift you add) for 12 months, and compares with no rise. Break-even is the cancelled share at which the two totals are equal. It is arithmetic on your assumptions, not a forecast; it ignores grandfathering, annual contracts that renew later, and customers who downgrade instead of leaving.
- Not modelled: upgrades and downgrades, coupons, tax, refunds, failed-payment recovery, usage billing. A customer who cancels and resubscribes shows as one churn and one new subscription.
Common questions
Can I get churn and cohort retention out of Stripe without a paid analytics tool?
Yes. Export your subscriptions from the Stripe Dashboard (Billing, Subscriptions, Export) as CSV and drop it here. The tool computes MRR, new and churned MRR, monthly churn and a cohort retention table from the start and cancel dates in the file. Nothing is uploaded; the file is read in your browser.
How is MRR calculated from the export?
Each subscription's amount is normalised to a monthly figure (yearly divided by 12, weekly times 52/12) and multiplied by quantity if the file has a quantity column. Each subscription's MRR is held constant at that value, because a subscription export has no history of upgrades, downgrades or discounts.
What does the tool not see?
Expansion and contraction between plans, coupons, tax, refunds, failed payments and usage billing. A customer who cancels and resubscribes appears as one churn and one new subscription. Treat the numbers as an estimate from the dates and amounts in your file.
What is the difference between logo churn and revenue churn?
Logo churn is the share of customers who left in a month; revenue churn is the share of MRR they represented. If small plans leave more than large ones, revenue churn is the lower of the two.
Is this financial advice?
No. It applies arithmetic to the file you provide and does not replace your own records or a qualified adviser.