Definition
Split similar cities or regions into treatment (ads on) and control (ads off), run both for a fixed period, then compare revenue movement. Because geography is hard to contaminate, geo tests produce cleaner causal reads than audience holdouts.
Why it matters
It is the most practical truth serum for scaling decisions - a channel that fails geo-lift at current scale rarely deserves next quarter's budget increase.
Benchmarks
4-8 week duration; match regions on prior-year sales curves
All glossary terms