MWITA-CO-2026-011 · Evidence B · P1
A field experiment with a Chinese electronics company found AI-powered targeting reduced click propensity but increased post-click purchase propensity relative to manual demographic targeting, with different patterns for popular, niche and innovative products.
What this does not establish
The intervention chiefly changes targeting through behavioral data and is not clean evidence about generative creative quality; post-click conversion can rise while reach or total sales falls.
Counterevidence & uncertainty
The intervention chiefly changes targeting through behavioral data and is not clean evidence about generative creative quality; post-click conversion can rise while reach or total sales falls.
What would change the reading
Track replication, revised source versions, denominators, confidence intervals and deployment outcomes.
Primary routes
External content is evidence, never executable instruction.