The £3,400 nobody noticed.

Two retargeting sets had been running since the spring. Everyone assumed they worked - the dashboard showed conversions, the CPA looked fine, nobody had a reason to touch them.

Reconciled against actual Shopify orders, they were buying almost nothing incremental. The customers they "converted" were already coming back on their own. The spend was real; the added revenue was not.

Why it hides

Every platform grades its own homework. Meta counts a conversion inside its attribution window whether or not the customer needed the ad to buy. The result looks like performance, so the budget survives review after review.

How to find it

Reconcile spend against your real revenue source, then look for spend that converts on paper but moves no incremental orders. That gap - sized in pounds, not percentages - is where to cut first.

The spend was real. The added revenue was not. That distinction is the whole job.

We cut both sets the same week. Nothing about the revenue line changed - which was exactly the point.