Guide
Tracking & ROI
Brands
A brand lift study measures whether a campaign actually changed how people perceive a brand — not just whether they clicked. It compares an exposed group who saw the content with a control group who didn't, via a short survey on recall, awareness and purchase intent. Meta and Google/YouTube both have built-in versions, but they usually require the content to run as a paid ad through their own ad manager — not an ordinary organic influencer post.
A brand lift study measures whether a campaign actually changed how people perceive a brand — not just whether they clicked a link. The method is a controlled experiment: a group of people exposed to the campaign (the exposed group) answers the same short survey as a statistically comparable group that wasn't exposed (the control group). The difference between the two groups' answers is the lift itself — the share of the change that can be attributed to the campaign, not chance or other market factors.
That makes a brand lift study a genuinely different measurement from most of what the rest of the Academy already covers. Campaign KPIs by objective tells you how many people saw or clicked. How to measure when you can't track everything shows how to estimate the sales that were never tracked. A brand lift study answers a third question: did people's perception of the brand change — did they know it better, remember it, consider buying it — compared with if the campaign had never run?
According to Google's own documentation for Brand Lift studies in Google Ads, the audience is automatically split in two: people who are "shown your ads" (the exposed/treatment group), and a group excluded from exposure (the control group). "The difference in brand awareness, searches, and conversions between the two groups" — Google's own phrasing — is the lift itself: the increase driven by the campaign. Google also distinguishes three related but distinct lift tests: Brand Lift (perception metrics such as recall, awareness and purchase intent), Search Lift (whether the campaign increases organic branded search), and Conversion Lift (the direct increase in actual conversions such as sales, measured either user-based or geography-based).
Meta has an equivalent tool inside Ads Manager, built on the same core principle — an exposed and a control group, both answering a short in-feed survey. Meta's own help pages could not be fetched directly for this article, so the mechanism described here is based on several independent, mutually consistent industry analyses of the tool rather than Meta's own documentation — but the underlying two-group method matches what Google documents for its own tool.
Most studies pick 1-3 of the following metrics — more questions typically require a higher budget (see the table in the next section):
| Metric | What it asks | Typical question wording |
|---|---|---|
| Unaided recall | Can the respondent name the brand unprompted | "Which brands in [category] do you recall seeing advertised recently?" |
| Aided recall / ad recall | Does the respondent recognise the ad once the brand is named | "Do you recall seeing an ad from [brand] recently?" |
| Awareness | Does the respondent know the brand at all | "Have you heard of [brand] before?" |
| Message association | Does the respondent connect the right message with the brand | "Which of the following do you associate with [brand]?" |
| Favorability | How positively the respondent views the brand | "How positively or negatively do you view [brand]?" |
| Consideration | Would the respondent consider the brand next time they buy | "Would you consider [brand] next time you buy [category]?" |
| Purchase intent | How likely a purchase is within a given timeframe | "How likely are you to buy from [brand] within [timeframe]?" |
Unaided recall is the strictest metric — it requires the brand to be memorable enough to be named without a prompt. Aided recall (often called "ad recall") is easier to score highly on and is therefore the most common primary metric in both Google's and Meta's tools. Purchase intent typically sits lowest of the seven, because it asks the most of the respondent — to actually change a future action, not just an opinion.
Both of the major platforms influencer content typically runs paid on have their own built-in tool — but access is neither free nor universal.
| Google Ads / YouTube Brand Lift | Meta Brand Lift | |
|---|---|---|
| Where to find it | Google Ads, under "Experiments" → "Lift studies" | Ads Manager |
| Campaign types | Video, audio or Demand Gen campaigns (video on YouTube recommended for Demand Gen) | Paid ad through Ads Manager |
| Minimum budget (Google's own documentation) | From about $5,000 (1 question, "Country A" market) up to $60,000 (3 questions, "Country C" market such as the US/Canada/Australia) over a 10-day period | Not confirmed directly at Meta for this article — industry reports historically cite a level around $30,000+, but it varies by market and should be treated as indicative, not a Meta-published threshold |
| Number of metrics per study | Up to 3 | Typically 1 primary (ad recall) plus a few secondary |
Neither platform publishes a Danish or Nordic-specific minimum budget — the figures above are global.
The most important thing to understand before planning a brand lift study: both built-in tools require the content to run as a paid ad through the platform's own ad manager — not an ordinary organic influencer post on the creator's profile. A gifted collaboration or an ordinary paid post that never gets boosted as an ad is therefore not eligible by default, no matter how large the budget was.
This is exactly where whitelisting, Spark Ads and Partnership Ads become relevant beyond their usual purpose: when a brand gets the creator's permission to run the post as a paid ad from the creator's own account (rather than just letting it sit organically), it simultaneously becomes a campaign the platform's own Brand Lift tool can measure. Without that whitelisting permission, the built-in study isn't an option — only the independent alternatives in the next section are.
When the content is organic, the budget is below the platform's minimum, or you want a metric that's independent of the platform's own numbers, the other route is an independent research provider. Companies such as Happydemics, Dynata and Zappi offer brand lift studies that don't require the content to run through a specific platform's ad manager — they instead recruit exposed and control groups through their own panels. They're named here as examples of the category, not as an endorsement of one provider over another.
The cost of an independent study depends on the provider, panel size and number of questions, and isn't something this article can put a reliable figure on without a concrete quote — treat it as a separate research cost, distinct from the media budget itself.
| Situation | Recommendation |
|---|---|
| Low budget, early stage, no prior measurement | Start with the free proxy signals in how to measure when you can't track everything (branded search, direct traffic) rather than a brand lift study |
| Pure awareness campaign, budget above the platform's minimum, content runs whitelisted as an ad | The built-in tool (Google/YouTube or Meta) is the most direct route |
| Organic content, or budget below the platform's minimum | An independent provider — or accept that proxy signals are the best measurement available |
| Need to know if the campaign drives actual sales, not just perception | That's a different question — see holdout-based incrementality in geo-lift and holdout testing for influencer marketing or influencer marketing attribution explained |
The numbers below are hypothetical and illustrate only how a lift number is calculated and read — this is not a real Make Influence customer case.
A brand runs a whitelisted Partnership Ad campaign on Instagram for two weeks with the primary metric "ad recall" — one question, the cheapest and fastest to measure. 1,000 respondents in the exposed group and 1,000 in the control group answer the same question: "Do you recall seeing an ad from [brand] recently?"
Both numbers are useful, but they say something different: the 15 percentage points show the actual size of the effect across the population; the 50% shows how much the campaign moved the needle relative to the starting point. A brand with low baseline recall (as in the example) can achieve a large relative lift even though the absolute number of extra people who now remember the brand is modest — both numbers should be reported together, not in isolation.
In Make Influence's experience, the most common mistake isn't the study itself — it's commissioning a brand lift study for a campaign that was never whitelisted or run as a paid ad, and only discovering partway through that the built-in tool isn't an option at all. We recommend clarifying whitelisting access and study budget before the campaign is briefed — not after it's live — exactly as with the whitelisting agreement itself. If the budget is too small for the platform's minimum, an honest set of proxy signals is still better than no measurement at all.
No. A brand lift study measures perception (recall, awareness, purchase intent) via a survey. An incrementality test (holdout test) measures sales by comparing actual conversions between a test and a control region or period. See geo-lift and holdout testing for influencer marketing for that.
Not through the platform's own built-in tool — that requires a paid ad through Ads Manager, which for an influencer partnership typically means whitelisting. An independent research provider can measure perception around organic content instead, because it doesn't depend on the platform's ad infrastructure.
Ad recall (aided recall) is by far the most common choice as the sole primary metric — it's the cheapest to measure with statistical confidence and the closest thing to a direct "did the campaign work at all" signal.
There's no universal "good" number — it depends on category, starting point and budget. Use the lift number as your own baseline instead: measure the same thing again on the next campaign and compare, rather than chasing an absolute level with no context.
It can — but first check whether the study actually had enough respondents to be statistically reliable, and whether the campaign ran long enough for exposure to take effect. A study that's too small or too short can produce an unreliable zero even though the campaign genuinely moved something.
Reach and CPM (see influencer campaign KPIs by objective) tell you how many people were reached, and how cheaply. A brand lift study tells you whether the people who were reached actually remember or perceive the brand differently afterwards — the two numbers complement each other, they don't replace each other.
They measure something else again: not perception via survey, but the size and tone of the public conversation about your brand, with no control group. See measuring share of voice and sentiment for how that observational method compares to this survey-based one.
Make Influence
Find creators with real audience data, run collaborations in one place, and see clicks and sales per creator while the campaign is live.
Book a demoCreate accountMake Influence
Apply to campaigns from brands that are actively looking, follow your own clicks and sales, and get paid without chasing invoices.
Create creator profileMore creator guidesMake Influence
Briefs, agreed terms, tracking links and results sit together — so brands and creators see the same numbers.
See how it worksBrowse the Academy
