eecommerceconsultant
Resources/Paid Media

Beyond ROAS: an incrementality-first Meta setup

Why last-click ROAS is quietly bankrupting DTC brands, and the incrementality tests we run first.

May 20269 min readBy Ecommerce Consultant

The ROAS trap

Reported ROAS on Meta is not a measure of profit. It is a self-reported claim by the ad platform that a click preceded a purchase. For most brands, 30–60% of that credit would have happened regardless.

Optimising to that number pushes budget toward branded and retargeting audiences that were already going to convert.

The first three tests we run

A geo holdout, a conversion lift test, and a spend step-change on a single campaign. Together they triangulate a defensible incrementality curve.

None of them require a data science team — but they do require the discipline to not touch the account for two weeks.

What we do with the curve

The output is a marginal-return chart that tells us what the next dollar actually does. Media plans are built from that chart, not from ROAS goals pulled out of a spreadsheet.

Key takeaways

  • 01Reported ROAS overstates incremental profit for almost every DTC brand.
  • 02A geo holdout is the cheapest incrementality test you can run.
  • 03Budget from a marginal-return curve, not a ROAS target.
Talk to us

Want this applied to your business?

Book a strategy call and we'll turn this framework into a written 90-day plan for your brand.

Book a strategy call →
Services that put this into practice