five operators, · five curves.
Selected cases from the 22-client active book. Names used with permission.
// problem
$4.2M ARR, stalled at 1.9x blended ROAS for nine consecutive months. Founders considering shutting down paid social entirely.
// fix
Server-side CAPI rebuild caught a 28% pixel-undercount on Meta. Lifecycle re-architecture pulled win-back revenue from negligible to 12% of revenue.
// problem
Furniture brand with strong first-purchase economics but no second-purchase model. Repeat rate stuck at 8% for 14 months.
// fix
Built SKU-cross-sell propensity model. Refactored post-purchase flows to 18-month cadence. Repeat rate climbed to 27% by month nine.
// problem
iOS 14 attribution destruction had hidden Meta's real impact. Founders had de-prioritized the channel for a full year.
// fix
CAPI deployment + holdout study revealed Meta was carrying 41% of new-customer acquisition, not the 8% the platform reported. Restored channel weighting in spend mix.
// problem
Subscription pet product. Strong subscribe but punishing churn — 38% by month four. CFO had stopped trusting the LTV slide.
// fix
Built churn-stage model; identified day-67 cancellation cliff tied to a packaging issue. Lifecycle save-flow recovered 11pts. Honest LTV slide back in board deck.
// problem
Supplement brand at $18M, ready to push to $50M but blocked by single-channel concentration. 84% of acquisition was Meta.
// fix
Diversified into Applovin + YouTube + influencer-as-paid-media. Six quarters later, Meta is 47% of acquisition, total spend up 6.4x, blended CAC down 42%.
