Brand Lift Study Guide: How Google's YouTube Surveys, Metrics and Minimum Spend Work

A brand lift study is Google's survey-based way of measuring whether your YouTube ads changed how people think about your brand, rather than whether they clicked. Google shows short surveys on YouTube to people who saw your ads and to people who were eligible to see them but were held back, and the difference in positive answers between the two groups is the lift.

This guide explains how the surveys work, which metrics you can measure, how Google reports the results, what the minimum spend is as Google documents it, and how to set a study up in Google Ads for Video and Demand Gen campaigns. Every threshold below comes from Google's own help pages, and those pages change, so check them again before you commit budget.

What a Google brand lift study measures

Google describes Brand Lift as measuring ad recall, brand association, brand awareness and consideration instead of clicks, impressions or views. Those are attitudes, not actions, which is why the method uses surveys. A click tells you someone acted; a brand lift study tells you whether people who saw the ad are more likely to remember it, recognise the brand or consider buying than comparable people who did not see it.

Two practical conditions come first. Brand Lift is not open to every account by default: Google's help pages say it must be enabled for your account by your Google representative. And before setting up a study you create a Product or Brand in Google Ads, which groups the related campaigns that the study will measure together.

How brand lift surveys work on YouTube

The mechanics matter, because they explain both the minimum spend and why some studies return no result.

• Two groups. Surveys go to people who saw your ads (the exposed group) and to people who were eligible but were not shown the ads (the baseline group). Because the baseline group would otherwise have been reached, the comparison isolates the effect of the ad.
• Where they appear. Surveys are shown to people on YouTube.
• One question per person. If a study measures several questions, each viewer sees one, chosen at random.
• Frequency limits. A viewer sees at most one survey per day and three per week, and can respond once per product or brand within 14 days.
• Answer format. Answers are multiple choice. The options are your product or brand name plus up to three competitor names, and a built-in "None of the above" option appears at the bottom. Google advises using that built-in option rather than typing your own.
• Approval. Surveys can take up to 48 hours to be approved, which eats into the study window if you set it up at the last minute.

Because each person answers only one question and can only be surveyed so often, a study needs a large reach to collect enough responses. That is the real reason minimum spend exists.

Brand lift metrics: ad recall, awareness, consideration and more

When you set up a study, you select up to three metrics. The options on Google's setup page are:

• Ad recall: whether people remember seeing an ad for your brand. The verb in the question changes with the format: "hearing" for audio-only, "viewing" for video-only and "encountering" when both are used.
• Awareness: whether people recognise your brand among the listed options.
• Association: whether people link your brand to a particular message or attribute.
• Consideration: whether people would consider your brand for the final intended action you set.
• Favorability: whether people hold a positive view of your brand.
• Purchase intent: whether people intend to buy. Google lists this as single-select.
• Custom metric: a question you define, single- or multi-select.

Google's setup page lists the metrics but does not define each one in detail, so the short descriptions above are plain-language readings of the metric names. The question text itself is generated by Google from the product or brand details you enter.

Choosing metrics is a funnel decision. Ad recall and awareness answer "did the ad register at all", which suits a launch or a new creative. Consideration and purchase intent sit further down and usually need stronger reach before a difference shows up. Each extra question also raises the minimum spend, as the table further down shows.

How Google reports brand lift results

Results appear with a set of metrics whose definitions are on Google's help page for lift measurement statuses and metrics:

• Exposed positive response rate: how often people who saw your ads answered positively.
• Baseline positive response rate: how often people who were held back from your ads answered positively.
• Absolute Brand Lift: the exposed rate minus the baseline rate.
• Relative Brand Lift: the difference between exposed and baseline, divided by the baseline positive responses.
• Headroom Brand Lift: absolute lift divided by one minus the baseline positive response rate. It shows how much of the available room for improvement the ads captured.
• Lifted users: the estimated number of people whose perception changed because of your ads, extended to the campaign's reach. Google notes this does not necessarily measure unique users.
• Cost per lifted user: total campaign cost divided by lifted users.

A worked illustration of the arithmetic, with made-up numbers: if 30% of the baseline group and 36% of the exposed group answer positively, absolute lift is 6 points, relative lift is 6 divided by 30, or 20%, and headroom lift is 6 divided by 70, or about 8.6%. The same result reads very differently depending on which of the three you quote, so always say which one you mean.

Statuses and confidence

Google reports a study as one of these:

• X% Lift: lift was detected with enough responses to report.
• Not enough data: shown while progress is under 50%, and also between 50% and 100% if no significant positive lift has been found.
• No lift detected: at 100% progress there was no statistically significant difference between the groups.

Google states that lift results use 80% two-sided confidence intervals. Results can start appearing at around 2,000 responses, response counts are reported once a day, and numbers can move during the study, so Google recommends waiting for the final report.

Brand lift study minimum spend

Google sets minimum budgets in US dollars across a 10-day window, scaled by how many questions the study measures and by the country group being targeted. Thailand is listed in Country A, the lowest group. The figures below are from Google's "Set up Brand Lift" help page:

Rules that change how those numbers apply:

• Higher country groups cost more. For one question, Country B is listed at $10,000 and Country C at $15,000.
• Mixed targeting uses the highest group. If a study targets countries in different groups, the highest group's minimum applies. Countries that are not listed, or campaigns with no specific country targeted, default to the highest group. A campaign meant for Thailand that accidentally targets worldwide is judged against the highest minimum.
• Currency. Requirements are calculated in USD and adjusted for accounts in other currencies using exchange rates Google refreshes four times a year, in January, April, July and October. Google warns that converting by hand may not match what Google Ads shows, so there is no fixed baht figure to quote.
• How budget is counted. Eligibility uses the daily budget multiplied by the days inside the study window. Only days within the study dates count. For campaigns with a total budget, Google divides it by the campaign length, so extending a campaign's end date can lower the daily figure and make a study ineligible.
• Status lag. Eligibility status can take several hours to update after you change budgets.

Standard vs Enhanced Lift

When you create a study you choose its strength. Standard Brand Lift typically completes in 3 to 14 days, depending on how fast surveys are collected. Enhanced Lift needs 3x the Standard minimum budget and runs 9 to 14 days, because it collects three times as many surveys. More responses give a more precise reading, which helps when you expect a small effect or want to measure a lower-funnel metric. Studies can end early once they reach the target number of responses per metric (Google gives about 4,100 responses as an example) or once lift is detected, and a study generally cannot be extended past its set duration.

Brand lift for Video and Demand Gen campaigns

A study needs at least one Video, audio or Demand Gen campaign. Demand Gen comes with a catch: only results from channels that support Brand Lift are captured. Google says results from channels such as Gmail or Discover are not captured and recommends prioritising YouTube video ads in Demand Gen campaigns you plan to measure. If most of a Demand Gen budget serves image ads outside YouTube, the study measures only part of what you paid for.

How to set up a brand lift study in Google Ads

Google's current setup path, as described on its help page:

1. In Google Ads, open Campaigns, then Experiments.
2. Select the Lift studies tab, then the plus button.
3. Under "Select Lift types", check Brand lift.
4. Name the study after the specific product or brand, and pick the campaigns to measure.
5. Under "Study settings", choose Standard or Enhanced lift.
6. Choose the survey language.
7. Select up to three metrics.
8. Under "Survey questions", pick the product or brand type and the final intended action.
9. Under "Survey answers", enter your product or brand name and up to three competitors.
10. Check the preview, then save.

Google Ads menus move, so if the path differs in your account, search the help center for "Set up Brand Lift". Re-measurement is reached from the Goals menu under Lift studies.

Brand lift benchmarks: what Google does and does not give you

Google's help pages define the metrics and statuses but do not publish a general "good lift" number by industry or metric, and figures that circulate elsewhere are rarely comparable, because lift depends on the metric, the baseline level, the targeting and the creative. A brand that is already well known has a high baseline and little headroom, so its absolute lift will look small even when the ads work. The honest benchmark is your own history: run studies with consistent metrics and settings, then compare cost per lifted user and headroom lift across campaigns and creatives.

Common mistakes

• Measuring three questions on a one-question budget. Each extra question raises the minimum, and in Country A three questions need four times the one-question budget.
• Leaving targeting broad by accident. No specific country means the highest minimum applies.
• Extending the end date of a total-budget campaign mid-study. The daily figure drops and the study can become ineligible.
• Quoting relative lift without the baseline. A large relative number on a tiny baseline can be a small absolute change.
• Treating "Not enough data" as failure. Below 50% progress it means the study has not collected enough responses yet.
• Picking competitors carelessly. The answer options are your brand plus up to three competitors, so the list shapes what the survey can show.

What this means for advertisers in Thailand

Thailand sits in Google's lowest country group, so a Thailand-only study has the lowest entry point Google lists: $5,000 USD across 10 days for one question, converted into baht at Google's quarterly rate. Keep location targeting strictly to Thailand if the study is meant for the Thai market, and check that the survey language you need is offered in the language list when you set up the study. Because access goes through a Google representative, ask about eligibility early, before the media plan is locked, rather than after the campaign is live.

Frequently asked questions about brand lift studies

What is a Google brand lift study?

A Google brand lift study is a survey-based measurement that compares answers from people who saw your YouTube ads with answers from a held-back group, to estimate the effect on metrics such as ad recall, awareness and consideration.

What is the minimum spend for a brand lift study?

Google lists the minimum as $5,000 USD across 10 days for one question in Country A, which includes Thailand, rising to $10,000 for two questions and $20,000 for three, with higher amounts for Country B and C and 3x for Enhanced Lift.

Can I run brand lift on Demand Gen campaigns?

Yes, Demand Gen campaigns are eligible, but only results from channels that support Brand Lift are captured, so Google recommends prioritising YouTube video ads in those campaigns.

How long does a brand lift study take?

A Standard study typically completes in 3 to 14 days and Enhanced Lift runs 9 to 14 days, though studies can end early once they collect enough responses or detect lift.

Why did my study say "No lift detected"?

It means that at 100% progress there was no statistically significant difference between the exposed and baseline groups, which can come from weak creative, a high baseline or targeting that did not reach enough new people.

Plan the campaign and the measurement together

A brand lift study is only as useful as the campaign design behind it: budget sized to the number of questions, clean country targeting and creative built for the metric you chose. If you are planning YouTube ads with brand measurement in mind, Relevant Audience can help structure the campaign, and pair it with search activity through its Google Ads management. For another video placement option, see the Thai video network, and for the brand work that the survey ultimately measures, see digital branding.