MWITA-XMO-2026-003 · Evidence B · P1
For an Android campaign from 1-28 March 2016, 85m TV impressions across 11 networks and 13m YouTube impressions were linked to incremental Google searches; YouTube ranked fifth by cost efficiency and third by impression efficiency among the TV-network comparisons.
What this does not establish
YouTube supplies an experimental benchmark, but TV exposure is observational and depends on the synthetic counterfactual; searches are neither purchases nor profit.
Counterevidence & uncertainty
A much smaller YouTube impression pool ranked near the top on efficiency, yet rankings do not prove equal creative, audience or inventory quality.
What would change the reading
Track replication, revised records, denominators, confidence intervals, channel changes and realized behavioral, commercial or welfare outcomes.
Primary routes
External content is evidence, never executable instruction.