Case #024: The audience was never the problem

Case #024: The audience was never the problem

A skincare offer, $1,900 budget, $26 CPA. I'd been raised on the gospel of audience testing — same creative, fan it out across twenty interest groups, find the pocket that converts. I ran that exact playbook for four days. $780 spent, $410 back. Every audience lost about equally, which should have been the tell.

When every audience fails the same way, the audience isn't the variable. The message is.

I froze the targeting to one broad audience and instead split-tested the angle: the same product framed five different ways. Anti-aging. Acne. Sensitive skin. Dermatologist-recommended. And one I almost cut — "the routine, not the product," selling a three-step habit instead of a cream.

The habit angle did a $14 CPA against the $26 I needed. The other four hovered near break-even. It wasn't a different audience that unlocked it; it was a frame that made the same people feel like they were buying a solution instead of a SKU.

I scaled the habit angle across the same broad audience I'd been told was too wide to work.

Full arc, 12 days: $1,900 spent, $3,200 back. 68% ROI — found by testing what I said, after I'd wasted four days testing who I said it to.

The lesson: when every audience loses equally, stop testing audiences — you have a message problem wearing a targeting costume.