Traditional is the brand's own Meta prospecting over the same days, retargeting excluded. Both sides use Meta's purchase attribution, so the comparison is like for like.
How we got there
Relay runs one loop: test many angles, keep what buys, rebuild everything on it. The comparison window above is the scale phase.
Aug 13 · Test
Two lanes, every ad with its own lander
An interest lane of audience stacks and a persona lane of eight buyer profiles, each ad paired to a lander written for that audience.
Aug 18 · Offer
The bundle enters the mix
A two-device bundle at a set price joins the persona lane alongside single-device angles and the brand's own best performers.
Aug 24 · Judge
Cut to what converts
Skincare-brand audiences bought at two to three times the rate of editorial, retail and wellness stacks. Every other lane was paused, not resurrected.
Aug 25 · Scale
Proven audience, proven offer, varied looks
The remaining budget runs the bundle to the proven audience in several distinct looks, each with a message-matched lander, so Meta does not collapse them.
Day by day
Relay (purple) vs. traditional (amber), cumulative within each phase. Both lines reset at the start of the scale phase. Dashed lines are each side's scale-window blended value.
Metric
Scale-phase Relay creatives
Relay ads live in the scale window, by share of impressions. Each links to its own message-matched lander on the brand's domain.
Run this test on your own brand.
Same budgets, measured side by side. See where your spend is leaking.