The situation
Growth and efficiency were pulling in opposite directions.
The account had a large, diverse catalog and a demanding efficiency target. Leadership wanted 8x or better on non-brand, and they wanted to grow. The easy way to hit that target would have been to shrink the catalog and lean on proven products and repeat-customer demand. That could improve reported ROAS in the short term while quietly capping new-customer growth.
What we found
Blended targets were hiding the account's actual structure.
- Existing customers reordering consumables behaved differently from practices buying for the first time.
- Repeat-customer demand could carry blended ROAS and mask new-customer acquisition performance.
- Regional performance varied enough that one national target obscured meaningful differences in return and acquisition cost.
What changed
The structure created separate levers for efficiency and growth.
- 01
Split existing customers from prospecting and stopped judging both against the same target.
- 02
Organized campaigns into regional tiers so budgets, bids and targets could move independently.
- 03
Kept the catalog broad and used campaign structure to allocate spend rather than restricting coverage to proven products.
The result
Efficiency improved while monthly spend grew more than tenfold.
Monthly spend grew from a few thousand to roughly $50K. Fully attributed year-to-date non-brand ROAS remained above 10x. A fully attributed quarter at roughly $50K per month averaged 10.6x against an 8x target.
Constraint & learning
Scale came from separating different business jobs.
Existing-customer campaigns carried efficiency. Prospecting campaigns carried growth, measured on account creation at a defined cost rather than ROAS alone. Regional tiers let the team push acquisition in markets producing new customers efficiently and push return where the economics favored it. The structure followed the business instead of forcing every buyer, market and product into one blended target.