← All terms·Digital attribution
Reference library · Digital attribution
Brand Lift
Brand lift is the percentage-point gap in a brand metric (aided awareness, favorability, purchase intent) between a group exposed to the campaign and a matched control group that was not, measured by post-campaign survey. It isolates the campaign's effect from category trends and seasonality that would move both groups equally. It is reported per metric, not as one blended number.
Formula
Worked example
A snack brand runs a YouTube campaign in Bulgaria. Post-campaign survey: exposed group (n=1,800) has 630 respondents with aided awareness = 35%. Control group (n=900) has 261 aware = 29%. Lift = 35 - 29 = 6pp, relative lift = 6/29 = 20.7%. The campaign moved awareness 6 points above where it would have sat without exposure, a real but modest gain given the media spend.
How it is used
Planners run brand lift alongside upper-funnel campaigns to justify spend beyond clicks and views, using platform tools (YouTube/Meta Brand Lift) or an independent panel (Kantar, Dynata) with a pre-assigned randomized split. The recurring mistake is building the control group from people who simply didn't see the ad organically rather than a randomized holdout, which lets self-selection into exposure inflate the lift.
The common mistake
Use a randomized control group assigned before the campaign, not people who happened not to see the ad, or the lift number is inflated by self-selection.