What Is a Meta Brand Lift Test and How Does It Work?

A brand lift test asks whether your ads changed what people remember or know about your brand. Meta withholds your ads from a randomly chosen holdout group, polls that group and a test group with survey questions, and compares the answers. It has rules about how you run your campaigns while it is live.

How does a brand lift test work?

Meta says brand lift selects a representative sample of people eligible to see your advertising and randomly divides it into test and holdout groups. It then uses causal inference techniques to measure the impact of your advertising within the sample. You can run a test on one Meta ad campaign or on all your campaigns across Meta technologies.

Both groups are polled about topics such as ad recall, brand awareness and message association. The difference between the groups is the brand lift of your ads.

Meta says a lift test is not an A/B test, because A/B tests have no randomized holdout group; see how Meta A/B tests work. Meta pairs brand lift with awareness objectives and conversion lift with sales, leads and engagement objectives.

When should you use a brand lift test?

Meta says to use brand lift when you:

• Run one campaign for one brand or product
• Plan to keep the creative and budget the same throughout the campaign

Ad sets you create for the test should fall completely inside or completely outside the test dates. Meta says not to use brand lift if other campaigns for the same brand or product are active across Meta technologies, because people in the holdout group could still see your ad from another campaign and contaminate the results.

What are the two ways to run one?

Point	Meta-managed test	Brand survey test in Experiments
Who sets it up	You work with your Meta account representative	You, in the Experiments tool
Availability	Eligibility depends on multiple factors and availability is not guaranteed	Only in certain regions, with minimum budgets that depend on your audience’s location
Results	Ready to view within 48 hours of the poll ending; your representative shares the report	Collected and analyzed automatically, then reviewed by you

Meta says a Meta-managed test is free to create. It also says results and performance aren’t guaranteed, and that you should hold off on changes or optimizations while the test campaigns run, except to solve problems.

How do you set up a brand survey test in Experiments?

You need a campaign with a sufficient budget for your audience and country. Then:

1. Go to Experiments, click Create test at the top, find Brand lift and click Get started.
2. Choose the purpose of your test below Learning objectives.
3. Choose whether the test covers all your Meta ads on an ad account or one campaign.
4. Select the region and the vertical (industry) Meta will use as comparisons for your results.
5. Schedule the start and end dates, which can differ from the dates of the campaign.
6. Name the test and select the survey language, which must match your audience’s language.
7. Create up to 3 brand survey poll questions.
8. Choose whether to also measure conversion lift; you must already be measuring conversions.
9. Click Review test details, then Create test to submit it for setup.

Meta says to run the test long enough to get more than 250 survey responses, recommends at least 4 weeks, and says it can’t overlap any other ongoing test. Meta also notes some pages can’t run brand surveys in Experiments depending on the kind of brand, such as brands promoting alcohol sales in certain locations. Figures are as checked on 2 Oct 2026.

What questions can the survey ask?

Every brand lift test must include a standard ad recall question, and the survey can have up to 3 questions in total.

Question type	Meta’s example wording
Standard ad recall (required)	“Do you recall seeing an ad for [page name] online or on a mobile device in the last 2 days?”
Standard brand awareness	“Have you heard of [page name]?”
Abstract favorability	“How would you describe your overall opinion of [page name]?”
Familiarity	“How familiar are you with [page name]?”
Recommendation	“Will you recommend [page name] to a friend?”

You can run a brand survey in only one language, even if you reach several countries.

FAQ

Is a brand lift test the same as a conversion lift study?

No. Both use test and holdout groups, but Meta lists them as separate lift tests: conversion lift for sales, leads and engagement objectives, and brand lift for awareness objectives. A brand lift test polls people with survey questions; see what a conversion lift study is.

Does a brand lift test cost anything?

Meta says a Meta-managed test is free to create, and that brand lift polls are available without additional cost for qualifying campaigns. Brand survey tests in Experiments need a minimum budget that depends on your audience’s location.

Can I edit a test after I set it up?

For a Meta-managed test, Meta says your account representative can edit or cancel it until shortly before the campaign starts delivering, and not after. Once it runs, Meta recommends holding off on campaign changes, because they make meaningful results less likely.