Ecommerce Scaling Secrets · Free operator guide · Keyword: MRR
I got this DM this week: “Do you have a sheet or template you use to track subscription numbers for ecom? MRR, active subscribers, churn, retention, projected revenue?”
Short answer: yes. Long answer: the sheet is the easy part. The hard part is knowing which of those numbers is lying to you, and what to do the Monday after you see it.
That's this guide. The exact metrics, the math behind each one, the four-tab sheet structure, where the same numbers already live inside your subscription app, the seven fixes that actually move them, and a working tracker at the bottom you can duplicate.
Your app shows you snapshots. Active subs today. MRR today. Churn this month. What it does NOT show you is movement, and movement is the whole business.
Two examples.
Aggregate churn hides the cliff. A blended 7% monthly churn can be a 25% drop-off at order one sitting on top of a 5% leak from the mature base. Two different problems, two different fixes, and one number hides both.
MRR × 12 is not next year's revenue. Ecom subscription math is not SaaS math. Blended DTC monthly churn runs 6.5–8.5%. At 7%, the next twelve months collect about MRR × 7.7. Multiply by 12 and you overstate next year by 30–40%. Plan inventory or a raise on that number and you will feel it.
Old advice still circulating vs. what the numbers say in 2026
| Old advice | 2026 reality |
|---|---|
| Track one churn % for the whole base | Track churn by ORDER NUMBER. A 4% cliff at order 2 hitting 4,200 subscribers costs more than a 6% leak at order 9 hitting 900 |
| MRR × 12 = next year | At 7% monthly churn, the next 12 months ≈ MRR × 7.7 |
| Blanket 50% off in the cancel flow | Reason-matched offers. On 3,000 cancel attempts, 50% off saved 27% and a free guide saved 22%. Five people of difference. The discount went to win-back instead |
| Failed payment = lost customer | 20–40% of all churn is a card that didn't go through. The best brands recover about half of it, most of it before the customer notices |
| Monthly is the default | Quarterly or prepaid at first purchase. 20–25% take it in month one, and renewal decisions drop from 12 a year to 4 |
| Metric | How to calculate it | Why it matters |
|---|---|---|
| Active subscribers | Count of live contracts. Exclude cancelled and expired. Decide once whether paused counts as active, then never change it | Headcount health |
| Gross MRR | Sum of the monthly-normalized value of every active contract (8-week, quarterly and annual all converted to a monthly number) | The headline number |
| Live MRR | Gross MRR minus the value of paused and skipped subscriptions | Cash you will actually collect |
| Customer churn | Cancels this month ÷ subscribers at the start of the month | Volume leak |
| Revenue churn | Churned MRR ÷ starting MRR | Dollar leak. The one that matters more |
| Retention / cohort | % of a signup month still active at order 2, 3, 6 and 12 | Where people actually drop |
| Projected revenue | Scheduled orders next 30 / 60 / 90 days × your real collection rate from payment-recovery history | Cash forecast |
Five more columns the operators who run this well keep next to those seven:
✅ The MRR waterfall — new / expansion / contraction / churned, so growth and leak show up separately
✅ Involuntary vs voluntary churn — a failed card is a different problem from “I don't want it”
✅ Cancel-flow save rate — saves ÷ cancel attempts
✅ Payment recovery rate — recovered ÷ failed
✅ Churn by order number — order 1 vs order 2 vs order 9