Those exposed more than four times cross-channel the brand metrics stopped moving. Additional impressions against already-reached audiences produced no measurable uplift.
Each channel reports exposures within its own measurement universe. A viewer who saw the ad twice on linear TV, once on CTV, and once on YouTube has been reached four times. Within-channel reporting shows one or two exposures per channel. Combined frequency across the full mix is invisible without a single measurement layer sitting above all of them simultaneously.
Once the ceiling was visible, spend that had been flowing to over-exposed audiences could be reinvested to genuine incremental reach.
Linear TV, CTV, YouTube, and digital video were each reaching populations the others were not.
That conclusion was only reachable because overlap was measured at the person level — the same people, tracked passively across every screen through Audio Content Recognition, with nothing modelled in between.
Without single-source measurement across all four channels, the 2.4% figure does not exist.
Platform-reported reach is calculated against each platform's own logged-in audience. Overlap with other channels sits outside that universe by definition.
Device graphs attempt to bridge this probabilistically. But a probability is not a person. And a frequency cap applied inside one channel does not control frequency across the full mix.
The gap is a structural feature of how channel-level reporting is built. Closing it requires a single measurement layer that sits above all channels simultaneously — identifying, at the person level, who was actually exposed across the full mix, and how often.
Lewis Kimber, Senior Media Manager at Domino's, presented the findings at Google's Accelerate Measurement Event in London.
The next campaign brief changed before a single pound of additional spend was committed.
Read the full Domino's UK case study →
Try a tailored single-source audience measurement solution to start improving your cross-media ROI.