Why ROAS Is a Lie (And What to Measure Instead)

ROAS tells you revenue divided by spend, but it does not tell you whether that revenue would have happened anyway. Without a counterfactual, the number is storytelling, not measurement.

The attribution problem

Most platforms credit themselves for conversions they touched, even if the real decision was made days earlier through organic research or word of mouth. When every channel claims the same sale, ROAS inflates across the board.

What to measure instead

Start with incrementality: run holdout tests, compare treated vs untreated audiences, and measure the true lift your spend generates. Pair this with blended CAC and LTV to understand whether your program is building a business or just buying clicks.