You are evaluating whether retail media spend is driving incremental conversions, but exposed users are also likely seeing paid social, search, and CRM touches during the same period. A simple exposed-vs-unexposed comparison would be confounded by cross-channel overlap, and spend in one channel may spill over into outcomes attributed elsewhere.
How would you design an experiment to measure the incremental impact of retail media spend when there is likely cross-channel overlap? Explain how you would isolate causal lift, define success, and decide whether the spend is truly incremental.