Case #009: Revenge buying, and the hole I dug deeper

Case #009: Revenge buying, and the hole I dug deeper
A promising offer turned on me. $700 spent, $300 back over three days, a 57% loss, and I took it personally. So I did the dumbest thing in the playbook: I tripled the budget the next morning to "make it back." If it just got more volume, I told myself, the law of averages would save me.

It didn't. More volume of a losing campaign is just a bigger loss. The fundamentals — the CPA, the conversion rate, the angle — were all still broken; I'd only made them break faster and louder. By the end I was $1,900 spent, $1,000 back. A 47% hole, dug with both hands, chasing a sunk cost like it owed me money.

The leak wasn't the original loss. Losing tests is the job. The leak was letting emotion size my next bet. Revenge spending treats a losing campaign as a slot machine that's "due," when a losing campaign is just a losing campaign at any budget.

The fix: a hard rule that any budget increase must be triggered by a metric crossing a threshold — never by a feeling, and absolutely never by a loss. Down days shrink my spend or stop it. They never grow it. The size of my next bet is decided by the data, with the emotion left out of the room.

I walked away, cooled off, and rebuilt the offer cold a week later — new angle, breakeven math done first, scaling tied strictly to CPA. $1,600 spent, $2,420 back. 51% ROI, made by the version of me that wasn't trying to get even.

— The lesson: a loss is data, not a debt the next campaign owes you — when budget size follows emotion instead of metrics, you don't recover the loss, you multiply it.