Case #035: I blamed the offer. It was my fault.

Case #035: I blamed the offer. It was my fault.
A dating offer ran beautifully for nine days. $90/day, steady +35% ROI. Then on day 10 it cratered — CPA doubled overnight. My first move was to email the AM: "offer's dying, what's the replacement?"

He pulled his stats. The offer's network-wide numbers were flat and healthy. The problem was only on my account.

I'd run the same three creatives, untouched, for ten straight days to the same audience. The audience hadn't gotten worse — it had gotten bored. CTR had quietly decayed from 1.8% to 0.7% while my conversion rate held. Same offer, same users, fatigued ad. The algorithm responded to falling CTR by raising my costs, and I'd nearly thrown away a working offer over my own laziness.

The fix: I now treat creative as perishable inventory. Every campaign gets a refresh cadence written into the launch plan — new hook or new visual every 5 to 7 days, before fatigue, not after. Three fresh variants went live.

CTR jumped back to 1.6% within a day. $1,500 spent over the next two weeks, $2,250 back. 50% ROI, on the offer I'd already tried to fire.

— The lesson: when a winner dies on schedule, suspect your own ads before you blame the offer — fatigue looks exactly like a dying offer, and only one of them is your fault.