Definition
Brand lift is the difference in a brand measure between an exposed group and a control group. The measure is an answer to a survey question: do you remember seeing an ad for this brand recently, have you heard of it, would you consider it, do you intend to buy it. The exposed group saw the campaign; the control group was comparable in every way that matters but did not, usually because it was held back at random.
The lift can be stated two ways. Absolute lift is the exposed group's share minus the control group's share, in percentage points. Relative lift is that difference divided by the control group's share, which makes small bases look dramatic. The absolute figure, quoted with the base, is the one to compare across studies.
Because the two groups differ only in exposure, the gap is attributed to the campaign. That makes brand lift a form of IncrementalityConversions caused by marketing. measurement, applied to memory and attitude instead of to actions.
In practice
Most brand lift studies are run by the advertising platforms on their own inventory. The brand awareness and lift guide describes what Google Ads Help and Meta's Business Help Centre document: an audience split into people who can see the ads and a held-back Holdout GroupA randomly chosen part of an audience kept from exposure to measure what exposure caused., short surveys served to samples of each, and results reported as lift, subject to minimum budgets and set-up conditions the platforms publish. The studies attach mainly to video and awareness campaigns, which the YouTube Ads section and the Meta Ads set-up for brand awareness cover from the campaign side.
Three choices shape what a study can say. The question. Ad recall moves most easily, because it asks only whether the ad was noticed; consideration and intent move less and matter more. The sample. Survey answers from a few hundred people per group carry wide uncertainty, so small gaps are noise. The exposure. A study run at low ReachUnique people who saw an ad. or low frequency tests a thin version of the campaign, and a disappointing result may say more about the plan than the creative; the reach and frequency calculator helps check the plan before the study starts.
An oat drink brand reads recall and consideration
Sorrelperch Oat Drinks spent £45,000 over four weeks on a video campaign and ran a platform brand lift study alongside it. The campaign reached about 1.8 million people at an average frequency of 3. Ad recall was 27% in the exposed group against 18% in the control group, an absolute lift of 9 points. Consideration was 32% against 31%, a 1-point gap that the study marked as not significant. The ads were being noticed, but they gave nobody a reason to switch from dairy. For the next flight the team kept the budget and the reach plan and changed the message, replacing a brand montage with a side-by-side taste and froth comparison in the first five seconds. The follow-up study showed a 4-point absolute lift in consideration, with recall broadly unchanged.
What brand lift can and cannot show
A lift study can show that a campaign changed recall, awareness or attitude among the people it reached, compared with similar people it did not reach. It is good at comparing creative and at confirming that a campaign registered at all, which click-based metrics cannot do for advertising that rarely gets clicked.
It cannot show sales. Lift in recall is necessary for most brand effects and not sufficient for any of them. It also cannot be added up across platforms: each platform measures its own campaign against its own holdout, with its own survey, so the sum of several platforms' lifts is not the brand's total effect. And a platform that grades its own inventory has an interest in a generous reading, so the study is best treated as that platform's estimate rather than an independent audit. The brand measurement section places lift studies beside tracking surveys and other measures the brand controls itself.
Brand lift versus conversion lift and incrementality
All three use the same logic of exposed against unexposed. Brand lift reads the difference on survey answers. Conversion lift reads it on actions such as purchases or sign-ups recorded by the platform. A broader incrementality test, designed by the advertiser across regions, audiences or time, reads it on the business's own sales data and can cover several channels at once; the incrementality testing guide explains those designs. A campaign can show strong brand lift and no measurable sales effect in the short term, which is not a contradiction but a reminder that the measures answer different questions over different time scales.
For video-led awareness work, the YouTube set-up for brand awareness explains how flights are planned so that recall and lift, rather than views, are the measures of success.
Related terms
IncrementalityConversions caused by marketing. is the general principle; brand lift applies it to attitudes. The Holdout GroupA randomly chosen part of an audience kept from exposure to measure what exposure caused. is what makes the comparison fair. ReachUnique people who saw an ad. and frequency decide how much of the audience the campaign had a chance to change. Share of VoiceYour share of market visibility. supplies the competitive context a single lift study lacks.
Common confusions
- Relative lift is the headline. Quote the absolute lift and the control group's base.
- Ad recall lift is brand growth. It shows the ad was noticed; consideration and intent are closer to a decision.
- Lift is a sales result. Sales effects need conversion lift, an incrementality test or a model.
- Lifts from different platforms add up. Each measures its own effect against its own control group.
- A platform's estimated recall metric is a study. Some platforms report a modelled recall estimate without surveys; it is not a measured lift.
