MWITA-XMO-2026-005 · Evidence B · P1
Immediately after TV spots, online sessions rose significantly but sales evidence was smaller and less stable: treated-region 30-minute means moved from 21.5 to 22.1 sessions (+3%) and $721 to $773 sales (+7%, not significant).
What this does not establish
Regions were matched rather than individually randomized; one travel platform/market and short windows do not establish durable customer value or campaign profit.
Counterevidence & uncertainty
Browsing lift was consistent, whereas sales weakened with the window and mobile purchase effects were null; on-air gains were partly offset off-air.
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.