A supplements brand scaled Meta spend into a collapsing account. CPC rose roughly tenfold, from 11p to over £1. CTR fell from 10.2% to under 2%, and website ROAS dropped below break-even before product costs. Meanwhile the email platform claimed a far bigger sendable audience than actually existed. Nobody independent of the spend was checking any of it.
What we found
The 90-day data told a clear story: impressions flat while clicks cratered. Classic creative fatigue and audience saturation, amplified by scaling budget into the decline. Google was quietly carrying the account at 2.55 blended ROAS, invisible in the platform-by-platform view. The email audit found nearly half the claimed audience was unsendable without re-permission, and funnel analysis showed a steep checkout drop-off leaking paid spend from both channels.
What changed
Budget moved to the channel the evidence supported. The failing channel was restructured around a new objective rather than fed more spend, and the brand now knows its real audience size. The decisive shift: performance conversations start from one agreed set of numbers instead of three platforms' competing claims.
When platform ROAS lies, run the business on profit
Retailers undercutting RRP made revenue-based ROAS actively misleading. The measurement was rebuilt and decisions moved to profit on ad spend.