Ecommerce Playbook · 2026-06-30 · 22 min
Most brands forecast from gut feel and ROAS. CTC runs on four interlocked proprietary models that produce a daily P&L forecast with 3.1% accuracy across more than $3 billion in GMV. In Part 2 of the Canon series, Luke Austin breaks down each model, how they connect, and what a Profit Engineer actually does with them Topics covered: Why a single ROAS number can't run a business The Spending Power Model - ACoNS curves and what spending power actually tells you The Retention Model - forecasting returning revenue cohort by cohort The Event Effect Model - how to get daily precision without breaking monthly targets The Creative Demand Model (teased - full breakdown in the creative strategy episode) How these four models wire together into a connected forecast system What R-squared below 0.6 signals about your retention data The three optimization points every brand should know: max CM, max revenue, max LTV CM Key stat: 3.1% forecast accuracy across $3B+ in GMV managed - producing 30%+ revenue growth and 40%+ contribution margin growth.
The Measurement Gap: iROAS, Geo Holdouts, and Progressive Truth63 / 100Why Ritsu Switched Email Platforms Mid-Growth (And What It Actually Took)How We Forecast $4B in GMV to 3% AccuracyHow One Methodology Runs 170+ Brands12 Years of Agency Knowledge in One Document