Case #024: The campaign I refused to scale

Case #024: The campaign I refused to scale

A lead-gen offer for home services was printing: $1,500/day, 70% ROI, stable for two weeks. Every bone in my body said raise the budget. I didn't raise it by a dollar.

I'd been burned before by vertical scaling — pushing one campaign's budget up until the algorithm chokes on bad inventory. This time I went horizontal instead.

Instead of taking the $1,500/day winner to $4,500, I cloned it into three separate campaigns, each capped at $1,500, each targeting a different state with the same proven creative and funnel. The winning campaign stayed exactly where it was, untouched and happy.

The four campaigns running in parallel:

— original (TX): held at 70% ROI, never disturbed
— clone (FL): settled at 58% ROI by day 5
— clone (OH): 64% ROI
— clone (NC): 51% ROI

Blended across all four, week three:

— $39,000 spent
— $61,400 revenue
— 57% ROI at 4x the volume

The original winner never saw its ROI dip, because I never asked it to spend more than it could spend well. I multiplied the campaign instead of inflating it.

The lesson: when something works, don't ask it to do more — ask three copies of it to do the same.