Count new customers, not revenue

A kidney-health supplement and education brand, the one from the flat-revenue case, came to me with a topline that had held roughly flat month to month, and paid and email both reporting healthy returns. Nothing in the dashboards said anything was wrong. Counted by first-purchase date, new customers had fallen by more than half since early 2024, from about 282 a month to about 90 (store order count, early 2024 to mid-2026).

The two facts fit together once you see the third. About 70% of orders were on a subscription (share of orders, from order line items, June to July 2026). The big 2024 and mid-2025 cohorts were still renewing, and their renewals filled the hole the missing new customers left. Revenue was a lagging number. The front door had been closing for a year and the back book had paid the difference.

Why the topline lags

Any business with repeat revenue has a back book: customers acquired in the past who keep paying. Subscriptions make it visible; replenishment, trade accounts and loyal regulars do the same thing more quietly. The back book is a good thing. It is also a delay line between a failure in acquisition and the month revenue shows it, and the longer the average tenure, the longer the delay.

At this brand tenure measured three ways gave three answers: a median of 4 to 5 months, a mean of 7 to 9, and about 12 if the current churn rate held. The cohort curve showed why: about 18% cancelled in month one, 35% by month three, half by month five. With a half-life around five months, a cohort of 282 keeps paying long enough to cover several months of 90s. By the time the topline fell, the brand would have been a year into the problem.

The number that does not lag

New customers per month, counted from the first-order date in the store. Not the platform's "new customer" flag, which resets with a new email, and not the ad platform's count, which cannot see a third of the buyers it is paid not to reach. The store's first-order date is the only record of when a person became a customer, and it moves the month the problem starts.

I put that line next to revenue for 24 months. Revenue: roughly flat. New customers: down more than half. One chart, two lines, and the question "is this stability" had its answer. The paid and email reports had both been true; they measured what they were pointed at, and neither was pointed at the front door.

The same lesson in a different store

A pharmacy and health ecommerce brand, from the recovery case, came out of a hard spring with revenue going from 31.3% down to 12.4% up year over year between April and August. True, and a relief. One split changed the story: everyday orders were still about 13% down, and bulk orders had carried the summer. By the first quarterly review, store revenue was up 12.5% year over year and revenue without bulk was up 0.3%, so the Q4 target was set on the second number.

Different brand, same shape. A line the headline hides, carrying the headline. At the subscription brand it was the back book; at the pharmacy it was bulk. The topline is a sum, and a sum can hold flat while its parts move in opposite directions. The split is where the decision lives.

What I built around it

The new-customer line went into the brand's monthly record as the first number, above revenue. The six subscription flows I wrote follow the cohort curve, with a save series that offers pause or skip before any discount, so the back book is protected while the front door is worked on. And the failed-payment flow, which was reaching 37 of 102 subscribers with a declined card because its check ran a moment before the store wrote the status, was fixed, because a back book also leaks, and that leak was a third of the people it should have caught.

Do this Monday

1. Export every order with its customer and date for the last 24 months.
2. Find each customer's first-ever order date and count customers by the month of that date. That is your new-customer line.
3. Put monthly revenue next to it on one chart, same months.
4. Find the month the lines diverge. New customers falling while revenue holds is a back book covering a fall; that month is when acquisition broke.
5. Find your sum's parts. Subscription against one-off, bulk against everyday, trade against web. Plan on the part that moves, not the sum.

In one line: revenue is a sum that lags; new customers per month from first-order dates is the number that moves first, so chart both and plan on the one that moved.