What the engine delivered
The challenge
The client had a €1M+ seasonal product category (garden houses) with zero prior paid experiment data, broken UTM tracking, and no knowledge of which channels, creatives, or messaging angles worked. GA4 captured almost none of ad-attributed revenue due to tracking gaps, and all revenue depended on direct conversion with no lead generation layer in place.
Our approach
- A structured 9-experiment sprint was run over 13 weeks across Meta, Pinterest, and Bing, covering UTM fixes, messaging A/B tests (5 angles), creative format tests, retargeting, lookalikes, and a quiz lead engine. A context-rich landing page quiz captured emails matched manually against sales exports to produce pixel-free 1-to-1 attribution. Bing was killed at week 9 (0.72x ROAS); Meta and Pinterest were scaled.
The result
Meta delivered 101x average ROAS and €803K platform revenue on €7,950 spend. Pinterest achieved the highest ROAS of the sprint at 189x (€0.75 per purchase) on just 6% of budget. The quiz generated 2,182 leads at €1.18 CPL and 22 verified sales worth €42,546 in proven revenue — no pixel required. Static creative won on ROAS (263x) while video scaled on purchase volume.
One product category, zero experiment history, and a baseline of €1M in prior-year revenue. The brief was simple: find out what works. The condition was uncomfortable: we had to start with nothing.
Want the full case study?
Download the complete version including methodology, experiment details and full results.
