What the engine delivered
The challenge
A major DIY retailer's €1M outdoor category had never been touched by a structured paid experiment. Tracking was broken—inconsistent UTM tagging hid paid social revenue inside 'direct'—so no one could judge channel performance. Five messaging angles existed but only the blandest ran, and there was no lead-capture mechanism for the high-consideration, multi-week purchase journey.
Our approach
- Stretch launched a 13-week e-commerce growth sprint beginning with a unified UTM tracking structure and a single agreed attribution source of truth. Nine experiments across 16 campaigns then tested five message angles, five retargeting audiences, three creative formats, two new channels, and a product-match quiz—each given a fixed test budget and a binary kill-or-scale verdict on return, not opinion. A lead-capture layer (quiz + email exchange) was added to convert browsers not yet ready to buy.
The result
Two channels (Pinterest and catalog retargeting) were scaled and one (Bing) was killed by week 9, with ~100x platform ROAS on under €10K of test spend. Analytics visibility of paid-social revenue grew roughly 18x after the tracking fix. The sprint generated 2,000+ owned emails at ~€1 each, with €40K+ in revenue proven via 1-to-1 email-to-order matching, and delivered a documented experiment method ready to scale across the rest of the shop.
A category doing a million a year, and nobody had ever run a structured paid test on it. Not one message A/B, not one channel comparison. Everything everyone believed about it was an assumption.
Want the full case study?
Download the complete version including methodology, experiment details and full results.
