Unlocking Retail Media Measurement Science and the Future of Performance with Andrew Covato
12 snips
Feb 18, 2026 Andrew Covato, founder of Growth by Science and former measurement lead at Google, Facebook, Netflix, and Snap, breaks down modern ad measurement. He contrasts platform-driven attribution with incrementality and MMM. He explains designing geo holdouts, calibrating models with experiments, and what retail media networks must provide for robust measurement.
AI Snips
Chapters
Transcript
Episode notes
Ad Platforms Optimize Revenue Over Advertisers
- Ad platforms optimize their own revenue not advertiser outcomes. Andrew Covato explains platforms will support any measurement approach that drives business even if it's not fully objective.
MMM Needs Experiments To Become Causal
- Marketing Mix Models (MMM) use aggregate time series to correlate inputs like spend, promotions, weather to sales. Andrew warns MMMs are correlational by default and become more causal when calibrated with experiments like geo tests.
Offer Bespoke Measurement For Nonstandard Brands
- Build bespoke measurement for advertisers whose businesses don't fit out-of-the-box tools. Growth by Science helps brands with custom MMMs, incrementality tests, and standing up in-house measurement when templates fail.