Case #040: The campaign that lost money on purpose

Case #040: The campaign that lost money on purpose

The spreadsheet said stop. $11,000 spent, $8,200 back, a 25% loss on first purchase. By every standard media-buying rule, this campaign should have been killed on day four. I let it run for sixty days, and that decision was the entire point.

The offer was a subscription supplement with a rebill. The CPA cost was real — $48 to acquire a customer whose first order paid $36. On the front end, every sale lost twelve dollars. A pure CPA buyer would have shut it down immediately and been right by their own math.

But the offer rebilled monthly at $36, and the data we had on retention was the hidden asset. Average customer stayed 3.8 months. That changed the unit economics from a $12 loss to a real number.

The arc over 60 days, tracked to lifetime value not first sale:

— $11,000 spent acquiring ~228 customers at $48 each
— first-order revenue: $8,200 (the visible 'loss')
— rebill revenue by day 60: $14,900 and still climbing
— blended revenue: $23,100 against $11,000 spent
— true ROI: 110%, after looking like a 25% loss for the first week

The customers who looked unprofitable on day one were the most profitable in the account by day sixty.

The lesson: when an offer rebills, first-purchase ROI is a liar — buy to lifetime value, or you'll kill your best campaigns on day four.