Loading…

Which customer segments add value when contacted?

Learn a segment targeting policy and evaluate it on a separate randomized sample. Declare the observation window or study design before running. The result includes denominators and leaves unidentified quantities uncomputed.

Required data: Assignment unit ID · Treatment (0/1) · Outcome · Pre-treatment segment · Train/evaluation role

A predeclared segment mean-difference model. Individual effects are unobserved; model selection on evaluation data contaminates evaluation.

Next step

Do new users stay and pay?

Common next stepPrepare a new dataset
CUPED experiment adjustment
Find other analyses