DeepVu × leading QSR brand
Thereportlookedflat. Thebusinesshadn'tbeen.
How incrementality measurement separated the growth branding actually caused from the growth that would have happened anyway.
The challenge
An aggregate view said the campaign did nothing. That conclusion was about to cost real growth.
Averages hide movement. When strong performance in one segment sits next to weak performance in another, the aggregate lands somewhere in the middle — and looks like nothing happened at all. The brand was close to cutting a channel that was working.
What we did
- 1
Define the question
Not 'what did the campaign deliver?' but 'what would have happened without it?'
- 2
Build the counterfactual
DeepVu modelled a like-for-like baseline so exposed and unexposed behaviour could be compared honestly.
- 3
Measure by segment
Incrementality was calculated per audience, device and daypart instead of one blended number.
- 4
Reallocate
Budget moved toward the segments carrying genuine incremental orders and away from the ones riding existing demand.
An average is a decision waiting to be made badly.
YOptima measurement team
The outcome
The channel wasn't underperforming. The measurement was.
What marketers can learn
Before you cut a channel, check whether you measured it or just averaged it.
Readytoengineerbetter brandingoutcomes?
Every campaign leaves signals about where growth is being created — and where it's being lost. Let's uncover them together.