Context · Business model
Marketplace
Two sides, where an intervention on one moves the other.
What changes here
The same four questions, and what each one runs into in this context.
Attribution
Supply and demand need separate measurement. Blending them hides which side is actually constrained.
Incrementality
Interference is the default rather than the exception. A treated market can pull inventory away from a held-out one.
Allocation
Only spend on the constrained side grows the market, and which side that is changes by geography and season.
Decisioning
Seller and buyer policies need separate holdouts. Treating a seller campaign as a buyer campaign with different filters builds spillover in.
Written for this context
Incrementality
- Planned
The Diagnostics That Earn Belief
Balance, overlap, pre-trends, placebo tests, and sensitivity analysis. Which to run, in what order, and what each one can and cannot rule out.
- Planned
When the Experiment Leaks
Noncompliance, contamination, and spillovers. Why intent-to-treat is usually the right estimand even though it answers a slightly different question.