Ecommerce Playbook: Numbers, Struggles & Growth Why “Perfect Attribution” Is a Lie (And What Actually Matters)
12 snips
Jan 13, 2026 In a thought-provoking discussion, Steve Rekook, Director of Data at Common Thread Collective, emphasizes that the pursuit of 'perfect attribution' is misguided. He argues that measurement should foster quick decision-making rather than aim for absolute accuracy. Steve highlights the importance of focusing on shared metrics that impact contribution margins and explains the interplay between marketing measurement methodologies. They also cover the hidden costs of precision and how to prioritize tests effectively for driving revenue.
AI Snips
Chapters
Transcript
Episode notes
Unit-Economics Mislead Without Context
- Simple unit-economics CAC targets ignore attribution fuzz, customer mix, and scale effects.
- True incremental CAC can differ dramatically from averaged cohort CACs.
Schedule Regular MMM Runs
- Run MMM regularly (quarterly to twice yearly) to understand channel impact and ad stock.
- Use MMM to inform allocation decisions for different channel mixes and scale.
Test Incrementality Continuously
- Run geo holdout incrementality tests every 4–6 weeks to normalize platform attribution.
- Use incrementality to separate new vs returning customer contributions and guide tactical buys.