The revenue arrives monthly. The acquisition cost arrives all at once.

In a subscription business, the money spent to win a customer lands in one month and the revenue comes back over many. That single fact makes almost every in-month efficiency report misleading: judged inside the window it was spent in, acquisition always looks unprofitable, and companies that manage to a monthly return target systematically underbuy growth they could afford.

What matters instead is payback period — how many months of subscription revenue it takes to recover the cost of acquisition — and whether the customer stays past it. The second mechanic is what the media is being asked to sell. A self-serve signup and a sales-led demo are different products with different economics, and running both from one budget against one blended cost per conversion hides which of the two is actually working.

The conversion an ad platform can see is a trial start, a signup or a form. None of them are money. Between that event and revenue sits activation, conversion to paid, and retention past the payback point — and a channel that produces cheap trials which never convert will look like the best performer in the account for as long as nobody checks. Optimizing toward signups reliably finds the least committed users available.

We pass paid conversions and retained revenue back from the billing and CRM systems so bidding aims at customers rather than at trials, and we report cohorts: what a month’s acquisition cost, what it has returned since, and where it sits against payback. The honest version of that report sometimes says a channel is not profitable yet and needs more time. That is a different statement to saying it is working.