The Attribution Window Mismatch: How Standard BI Infrastructure Killed a 4.2x ROAS Channel
Why we stopped optimizing media buying and started optimizing our measurement stack. A deep dive into fixing B2B attribution for considered purchases.
Marketing Engineer
I'm a Marketing Engineer who builds systems at the intersection of data engineering and growth strategy. After being laid off in December 2024, I've been contracting and contributing to open source. I'm passionate about MMM, AI Agents, and privacy-first server-side tracking.
What fascinates me most isn't the innovation itself — it's the adoption curve. That's where I've chosen to focus my career.
Currently open to contract work or full-time opportunities.
Debug and validate consent management platform implementations in real time.
The media planning toolkit that lives inside Excel.
Why we stopped optimizing media buying and started optimizing our measurement stack. A deep dive into fixing B2B attribution for considered purchases.
High ROAS is often a signal of selection bias, not efficiency. In this case study, I break down how a luxury retailer’s "profitable" retargeting campaign was actually a negative-leverage tax on their most loyal customers—and how we proved it with a simple RCT.
When ad channels scale together, standard regression models fail to distinguish cause from effect. This post explores how correlated spend leads to random credit assignment and why you need randomized experiments—not just better models—to untangle the signal