Big question
How much of your video reach is the same person counted twice?
The position
Assume the plan overstates reach, because the same person is counted in more than one place. Report the planned figure and the measured figure separately rather than reconciling them into one number nobody can defend.
Why it is live
Every large plan adds linear, streaming and social reach together, and the joins between them are modelled rather than observed. The WFA restated in December 2025 that identity, exposure and campaign data remain separated, which turns a channel-level reach plan into a person-level duplication problem.
The case for it being solvable
Trade reportedA cross-media service with a published method and audited coverage is being built and partially deployed.
The case against
UnknowableNo published, independent study exists of how far combined video reach plans land from what happened, and producing one would need a single advertiser's raw logs from every platform for the same period.
Where the weight sits
Strongly against the plan number. The direction of the error is certain and the size of it is not, which is exactly what to say when somebody presents a combined reach figure.
What would change this
A measurement company publishing planned against measured reach for a set of real campaigns, or a cross-platform currency going live with an audited method.
Where this is taught
TV and Video
Sources
- WFA cross-media measurement, 3 Feb 2025secondary
- WFA next steps for Halo, 19 Dec 2025secondary