Dema

Wasted Ad Spend

Wasted ad spend is budget that produced no incremental profit. Note that definition carefully: not budget that produced no conversions. A campaign can convert well and still be waste, if those conversions would have happened without it.

The four kinds worth separating

  • Non-incremental spend. Budget capturing demand you already had. Branded search is the classic case — the customer was going to find you anyway, and the click is charged regardless. Only an experiment can tell you how much of it is real.
  • Spend on products that cannot convert. Traffic sent to items that are out of stock in the sizes people want, or where a broken size curve has quietly gutted conversion rate.
  • Spend on unprofitable products. Campaigns that convert well on low-margin, high-return items. Revenue looks fine; Gross Profit 2 does not.
  • Spend on unprofitable customers. Acquisition that works but brings in one-time discount buyers whose lifetime value never covers their CAC.

Why ROAS hides it

ROAS divides revenue by spend, which means it cannot distinguish any of the four cases above. Two campaigns can report identical ROAS while one sells high-margin products to new customers and the other sells discounted, high-return products to people who would have bought anyway.

Steering on ROAS therefore tends to increase waste over time, because the easiest way to improve it is to buy more of the demand that already exists. epROAS — profit rather than revenue over spend — closes part of that gap, and incrementality testing closes the rest.

How to find it

Three measurements, in order of how much they tell you:

  1. Margin-level performance. Rebuild campaign reporting on contribution margin instead of revenue. This alone usually reclassifies a meaningful share of "winning" campaigns.
  2. Product-level availability. Cross-reference spend against stock and size coverage to find budget pointed at pages that cannot convert.
  3. Incrementality testing. The only method that identifies non-incremental spend, because it is the only one that observes what happens when the spend stops. Branded search and retargeting are usually where the surprises are.

What "eliminating" it realistically means

Some waste is unavoidable — prospecting is a search process and a proportion of it will always fail. The goal is not zero waste but knowing which portion is structural and which is fixable, then reallocating the fixable part rather than cutting budget across the board. Cutting evenly removes good spend along with bad.

See unified measurement for how attribution, MMM and incrementality testing combine to identify non-incremental spend.

Turn data into decisions.