Every scale attempt spiked CAC within days. A 5-10 ad / week creative pipeline could not outrun how fast the Meta algorithm exhausted winners on beauty PDPs.
The strategic frame every angle traces back to - the villain we name, the mechanism that beats it, and the proof no competitor can copy.
Meta exhausts beauty creative in days - 5-10 ad / week shops can't outrun the algorithm, so scale spikes CAC every time
150+ variant / week engine + bid-cap laddering + 48-hour kill/promote discipline
Creative volume as a compounding moat - winning hooks recycled into 2nd / 3rd generation variants
Stood up a 150+ creative variant / week pipeline across 6 angle families. Throughput high enough that no single winner had to carry the cohort - the next-generation variant was always already in test.
Laddered bid caps so spend stepped up only when CPA proved out at the prior step. Avoided the classic 'double the budget and watch CAC double' beauty-Meta pattern.
Every creative cohort got a 48-hour read. CPA-positive: scale within the bid-cap ladder. CPA-negative: cut, no exceptions. Decision cycles measured in days, not weeks.
Recycled winning hooks into second- and third-generation variants - new visuals, new openers, same proven hook. Creative lifespan extended 3.1× because the hook didn't die when the visual fatigued.
Percentage of new creatives reaching CPA-positive at 48 hours tripled versus the previous in-house pipeline - volume created the surface area for winners to emerge.
Spend more than doubled while win-rate climbed - bid-cap laddering kept efficiency intact through every scale step.
Winning hooks survived the visual fatigue cycle by ladder-recycling into 2nd / 3rd generation visuals - the cohort that would have died at week 2 ran six.