Scalewell / Blog

Brand lift study for B2B: how to plan one on LinkedIn, YouTube and Meta and read it against pipeline

A planning checklist and readout method for B2B brand lift studies on LinkedIn, YouTube and Meta, with the November 2026 minimums and a way to line results up against pipeline.

Two groups of people, one facing a bright screen and one turned away, with a measuring scale between them, picturing a brand lift test and control group.

A brand lift study surveys two random groups, one that saw your ads and one that didn't, and reports the gap in answers such as ad recall. For B2B teams, LinkedIn's version is usually the affordable one that reads by job function, and the result earns budget only once checked against CRM pipeline over two to three quarters.

You probably know why you're reading this. The awareness campaign ran for a quarter, sales says nobody mentioned it, and the CFO wants to know whether to renew it. You have impressions, a video completion rate and a feeling, and none of those will survive the budget meeting.

This guide is for the head of paid media or demand gen who has to plan that study and defend its result. It covers what each platform offers as of November 2026, with the minimums from their own help pages, a planning checklist you can copy, prompts for having AI check feasibility and read the export, and a way to line the result up against pipeline without pretending a survey is revenue.

A brand lift study measures what people remember and think, which is a different outcome from a conversion lift test

A brand lift study is a randomized survey test: the platform splits people who are eligible to see your campaign into a test group and a control group, asks both a short question in the feed or before a video, and reports the difference in favorable answers as lift. Google describes Brand Lift as measuring "ad recall, brand association, brand awareness, and consideration" rather than clicks or views (Google Ads Help).

The sums behind it are short. Absolute lift is the test group's favorable answer rate minus the control group's, in percentage points, and relative lift divides that gap by the control rate. If 21% of the test group recalls your ad and 14% of the control group says it does, absolute lift is 7 points and relative lift is 50%.

A conversion lift study uses the same random split but counts conversions the platform can track instead of survey answers. Our guide to incrementality testing for B2B covers conversion lift studies and account holdouts, so this article stays on the survey side and on how to connect it to pipeline afterward.

Scalewell helps paid and demand gen teams write brand measurement into the plan before the media is booked, so each study answers a budget question agreed in advance and AI does the readout against the CRM.

The three platforms differ most on budget, access and what the breakdowns show

Google, LinkedIn and Meta all run brand lift studies for free, and all three gate them behind a minimum media budget. The table sets out what each official page says as of November 2026, using US figures where the minimum varies by country.

LinkedIn Brand Lift TestingGoogle Brand Lift (YouTube)Meta Brand Lift
Where it livesCampaign Manager: Measure, then Test, then Create test, then Brand Lift testCampaigns menu: Experiments, then the Lift studies tabExperiments: Create test, then Brand Lift, then Get started
AccessCampaign manager access or higher on the ad accountOnly through a Google account representativeSelf-serve in Experiments, or through a Meta representative
Minimum budget (US)$60,000 for one question, $90,000 for two, plus $45,000 per extra question, budgeted across the test$10,000 for one question, $20,000 for two, $60,000 for three, within 10 days$175,000 (minimums dated September 15, 2026)
Questions1 to 6Up to 3Up to 3, ad recall required
Duration14 to 90 daysStandard Lift usually 3 to 14 daysAt least four weeks recommended
Responses before resultsAt least 30 per group, with 4 positive and 4 negative, for a lift statusAbout 2,000 responses per metric to detect lift in a strong campaignAt least 250 responses
Breakdowns that matter in B2BJob function, job seniority, company industry, company sizeAge, gender, device, campaign, videoAge and gender

Sources: LinkedIn's budget guidelines, test setup page, status and methodology page and results page; Google's Set up Brand Lift and statuses and metrics pages; Meta's minimum requirements, test setup and results pages.

The breakdown row decides more than the budget row. A recall lift among "adults aged 35 to 44" tells you very little about whether IT directors at 500-person companies now know your name, and only LinkedIn reports lift by the professional attributes your ICP is defined in.

LinkedIn brand lift fits B2B best, but only on a broad persona audience

LinkedIn brand lift is the study most B2B teams can run without a sales rep, and the one whose results speak your language. You create it under Measure, then Test, in Campaign Manager, at account level or for chosen campaigns, with one to six questions and a schedule of 14 to 90 days (LinkedIn Help). The metrics include ad recall, aided awareness, top-of-mind awareness, product consideration, recommendation, brand familiarity and favorability, plus four employer versions for recruitment campaigns (LinkedIn Help).

The budget counts what your active campaigns are set to spend during the test, not what they have spent before. A $1,500 daily budget across a 60-day test counts as $90,000, enough for two questions, while paused and draft campaigns count for nothing (LinkedIn Help).

Audience size is where B2B plans usually break. LinkedIn recommends "an audience size of at least 1 million" and strongly encourages at least 300,000 members (LinkedIn Help). A named list of 3,000 target accounts rarely gets there, so run the study on a persona audience built from job function, seniority, industry and company size, and read the account side later through the pipeline readout below.

Two scheduling rules catch people out. An account-level Conversion Lift test and an account-level Brand Lift test can't run at the same time in one account, and neither can an A/B test and a Brand Lift test (LinkedIn Help). Results can also move for up to 15 days after the end date, so book the readout meeting for 16 days after the end date.

YouTube brand lift needs a Google rep and spends its minimum in 10 days

YouTube brand lift runs on Video, audio and Demand Gen campaigns, and Google states plainly that "Without a representative, you won't be able to use this feature" (Google Ads Help). If your account has no rep, this option is closed whatever your budget.

The minimum is lower than LinkedIn's but arrives faster. The US sits in Google's Country B bucket, which needs $10,000 for one question, $20,000 for two and $60,000 for three, all within the first 10 days of the study, and targeting several countries applies the highest bucket among them (Google Ads Help). Enhanced Lift collects more responses for three times the minimum and runs 9 to 14 days.

Google also offers Search Lift, which measures whether the ads made people more likely to search for your brand on Google and YouTube, at no extra budget when you already qualify for one Brand Lift question (Google Ads Help). For B2B, that's the most useful add-on of the lot, because branded search is the first place an awareness campaign tends to show up in your own data.

The catch is the audience. Google picks respondents from everyone eligible to see your ads on YouTube, and its breakdowns stop at age, gender, device, campaign and video, so a YouTube lift is only as B2B as your targeting was. Our guide to B2B video ads with AI covers how to build those campaigns.

Meta brand lift is priced for consumer brands in the US

Meta brand lift is self-serve in Experiments, but the US minimum of $175,000, as of September 15, 2026, puts it beyond most B2B awareness budgets (Meta Business Help Center). Other markets cost less, with the United Kingdom at $64,000 and Germany at $110,000.

Meta also asks for conditions that suit a single consumer launch better than an always-on B2B program. It advises using Brand Lift when you "Run one campaign for one brand or product" with the creative and budget held steady, and against running it while other campaigns for the same brand are live (Meta Business Help Center). Every Meta study must include its standard ad recall question, and results break down by age and gender only.

If Meta carries part of your B2B awareness spend outside the US, a study can still make sense. In the US, most B2B teams will get a more useful answer from LinkedIn for half the money.

Plan the study with this checklist before you book the media

A brand lift study that wasn't planned against a decision ends as a slide nobody acts on. Fill in every row before launch, and have AI check the numbers against the minimums above.

FieldWhat to writeExample (hypothetical)
Decision it informsThe budget call the result will changeRenew the persona video program for H2 at $96,000 or cut it
HypothesisWhich answer should move, for whomAd recall and consideration rise among IT directors at 200 to 2,000 employee companies
Platform and levelAccount or named campaignsLinkedIn, campaign level, two video campaigns
Audience and sizeTargeting and LinkedIn's forecast size450,000 members, above the 300,000 floor
QuestionsTwo at most for most B2B budgetsAd recall, product consideration
Budget against minimumBudgeted spend in the window against the platform minimum$96,000 over 8 weeks against $90,000 for two questions
Dates and readoutStart, end, and the final readout date8 weeks, readout 15 days after the end date
Conflicts clearedOther lift or A/B tests paused, creative frozenNo account-level conversion lift, no A/B test in the window
Success barStatus, size and cost you'll acceptLift Detected on recall, at least 3 points absolute, under $15 per lifted member
Pipeline readoutsDates and the CRM comparisonBranded search at 6 weeks, opportunities at 90, 180 and 270 days
OwnerWho calls the resultHead of paid media, with the demand gen lead

Write the success bar before launch. Once results arrive, any number can be argued into a win, and a written threshold is what stops the team from doing it. If the study sits inside a wider plan, our media plan template has a measurement row to hold it.

AI with code execution, such as Claude or ChatGPT, can run the feasibility check in a minute:

You are checking whether a brand lift study is feasible before we book media.
Platform minimums (US, November 2026):
- LinkedIn: $60,000 for 1 question, $90,000 for 2, +$45,000 per extra question,
  counted as budgeted spend of active campaigns during the test; test 14-90 days;
  audience at least 300,000 members (1 million recommended).
- Google Brand Lift: $10,000 / $20,000 / $60,000 for 1 / 2 / 3 questions within
  10 days; needs a Google account representative.
- Meta: $175,000 in the US; no other campaigns for the same brand during the test.
Our plan: platform [X], daily budgets [list], test length [N] days,
audience size [N], questions [list], other tests running [list].
1. Calculate budgeted spend inside the test window and compare it with the minimum.
2. Flag every rule the plan breaks, quoting the rule above.
3. If the plan fails, give the cheapest change that fixes it (fewer questions,
   longer window, broader audience) and say what that change costs us.
Do not invent requirements that are not listed above.

Read the lift result on its status first, then its size, then its cost

The status tells you whether the lift is real, and each platform draws that line differently. Read it before you look at any percentage.

  • LinkedIn moved to Bayesian methodology in June 2026. "Lift Detected" means a lift probability of 90% or more, "Directional Lift Detected" 80% to 90%, and LinkedIn says it doesn't recommend acting on "No Lift Detected" (LinkedIn Help).
  • Google reports lift with 80% two-sided confidence intervals, meaning a 90% chance the true lift sits above the lower bound, and says you can get "directional insights" at 50% certainty of lift or more (Google Ads Help). A 50% certainty is a coin toss, so hold YouTube results to the same 80% bar as LinkedIn's directional line before they move money.
  • Meta says a confidence percentage above 90% means it "reliably" believes the ads caused lift (Meta Business Help Center).

Then read size and cost together. LinkedIn's estimated total lift multiplies the members you reached by the absolute lift, and dividing spend by that figure gives a cost per lifted member that you can compare across campaigns. Google's own guidance makes the same point, calling cost per lifted user a better primary metric than absolute lift because "it factors in both reach and cost."

Take the hypothetical study from the checklist. Over 8 weeks and $96,000, 231 of 1,100 test respondents recall the ad (21%) against 126 of 900 in control (14%), so absolute lift is 7 points. With 160,000 members reached, estimated total lift is about 11,200 members, at roughly $8.57 each. On consideration, 10.0% of the test group against 9.1% of control is a gap of about 1 point; a simple frequentist check puts the chance it's real at roughly 76%, below the directional line, though LinkedIn's Bayesian figure will differ. The honest summary is that people now remember the ads and nobody has yet changed their shortlist.

Here is a brand lift export: for each question, test and control respondents,
desired responses in each group, platform lift status, and reach. Spend: $[S].
Our success bar, written before launch: [paste].
1. For each question, give absolute lift, relative lift and cost per lifted member.
2. Report the platform's status as given. Do not upgrade "directional" or
   "no lift detected" into a finding.
3. Compare each question with our success bar and say pass or fail.
4. For job function, seniority, industry and company size breakdowns, list only
   those the platform marks significant or directional, and say how many
   respondents each one rests on.
5. Write a three-sentence summary a CFO could read, with no adjectives.

Reading brand lift against pipeline takes three readouts over three quarters

A brand lift result is an early signal, and B2B pipeline arrives late, so plan three readouts rather than one. The timings below are this guide's suggestion, set against published journey lengths, not a platform rule. Dreamdata's 2026 LinkedIn Ads Benchmarks Report put the full B2B buyer journey at 272 days (PR Newswire, March 10, 2026), and 6sense's 2025 Buyer Experience Report found that "94% of buying groups ranked preferred vendors before first contact" and bought from that favorite 77% of the time (6sense). Those two figures explain why a recall lift can matter a great deal and still take most of a year to reach the CRM.

Readout 1, at the study's final date. Record the status, size and cost from the previous section, plus the breakdowns by job function and company size. This tells you whether the right people noticed.

Readout 2, about six weeks in. Check branded search. On YouTube, Search Lift measures it directly. For LinkedIn and Meta studies, compare weekly impressions on your branded search campaign in Google Ads, and branded queries in Search Console, with the same weeks of the previous year. A recall lift with flat branded search means people remember the ad but aren't looking you up yet.

Readout 3, at 90, 180 and 270 days. Export LinkedIn's professional demographics report, which lists company names for members who interacted with your ads (LinkedIn Help), match those companies to your target account list, and compare the opportunity rate of matched accounts with target accounts in the same tier that never appear. Our guide to LinkedIn thought leader ads walks through that account comparison in detail.

Be clear in the readout about what it proves. Accounts that engage with ads are often accounts already warming up, so the comparison shows association, not cause. If the CFO needs cause, run an account holdout as described in the incrementality guide.

A long path with a glowing early stretch, a flag, a parallel trail of footprints and three lantern markers leading to a distant final post, picturing a brand lift study and its later pipeline readouts.
This is a hypothetical LinkedIn brand lift study. The study runs in weeks 1 to 8, and the final lift result (readout 1) arrives at week 8 plus 15 days. The branded search check (readout 2) starts from week 6. At days 90, 180 and 270 after launch, readout 3 compares the opportunity rate of engaged target accounts with unreached accounts in the same tier. Dreamdata's 272-day average B2B journey is marked at the end.

Combine the readouts into one decision with this table. It is a suggested starting point for your team to adapt, not a benchmark:

Lift resultPipeline at 180 to 270 daysA likely readingWhat to do
Lift detectedEngaged accounts open more opportunitiesThe program is working on the right peopleRenew, and test whether a holdout confirms cause
Lift detectedNo differenceRecall moved among people who don't buy, or the offer downstream is weakCheck the job function breakdown, then the sales follow-up
No liftEngaged accounts open more opportunitiesPipeline came from accounts already in marketDon't credit the awareness spend; retarget those accounts directly
No liftNo differenceThe creative or audience isn't workingChange one of them before spending again

AI can do the matching and the arithmetic each quarter:

Files: (1) LinkedIn professional demographics export with company names,
(2) our target account list with tier and domain, (3) CRM opportunities with
account domain, created date and stage, (4) the brand lift summary.
1. Fuzzy-match company names in file 1 to domains in file 2. List matches
   below 90% confidence for me to check by hand.
2. For engaged and unreached target accounts in each tier, report the share
   that opened an opportunity within 90, 180 and 270 days of the study start.
3. Place the result in the decision table I paste below and say which row applies.
4. State that engagement is self-selected and the comparison is not causal.

When the result goes upward, the board version is in our guide to reporting marketing ROI to the board.

Brand lift studies mislead B2B teams in four predictable ways

Most bad brand lift calls in B2B come from asking the wrong people, measuring the wrong thing, or reading too early.

The respondents aren't your buyers. Google and Meta survey whoever was eligible to see the ads, so a lift there may come from people who will never sit on a buying committee. On LinkedIn, check the job function and seniority breakdowns before you quote the headline.

Product creative measured against the brand. Google warns that a creative focused on one product, measured on the overall brand, will likely show "No lift detected" (Google Ads Help). Match the question to what the ad actually says.

A contaminated control group. Ads that ran before the study, or other campaigns for the same brand during it, reach the control group and shrink the lift. Google says creative seen before launch can contaminate the control.

Reading early or rounding up. A result read before LinkedIn's 15 days of settling is provisional, and a directional result is a hint. Recall lifts are also the easiest metric to move, so a program that lifts recall and nothing else is still waiting for its real verdict in the CRM.

Where to start with a brand lift study

Fill in the planning checklist for your next awareness campaign this week and run the feasibility prompt on it; if no platform passes, a branded search and CRM readout alone is still a better answer for the CFO than impressions. Scalewell works with B2B paid teams on where studies like this sit in the measurement roadmap and on making the readout something AI runs every quarter.

Frequently asked questions

How much does a brand lift study cost?

The study itself is free on LinkedIn, Google and Meta; you pay through the media budget needed to qualify. As of November 2026, the US minimums are $60,000 for one LinkedIn question, $10,000 within 10 days for one Google question, and $175,000 for Meta.

What is a good brand lift result?

No credible public benchmark exists for B2B brand lift. LinkedIn and Meta show benchmarks for your selected region and industry inside the results, and LinkedIn's are directional, built only from completed tests that showed a detected, negative or directional lift. Judge your result against the success bar you wrote before launch, especially cost per lifted member.

What is the difference between brand lift and brand awareness?

Brand awareness is one thing a study can measure: whether people have heard of you. Brand lift is the measured difference between people who saw your ads and people who didn't, on any of several questions, such as awareness, ad recall, consideration or favorability. A lift study can show awareness rising while consideration stays flat.

Can a brand lift study run on a LinkedIn matched audience?

It can if the audience is big enough, but LinkedIn strongly encourages at least 300,000 members and recommends 1 million, which most named account lists won't reach. Run the study on a broader persona audience instead, and compare your named accounts separately through the CRM.

How many questions should a B2B brand lift study ask?

Two is the practical limit for most B2B budgets. Each extra LinkedIn question adds $45,000 to the minimum, and Google's three-question tier costs three times the two-question one in the US. Pair ad recall, which moves first, with consideration, which matters more for pipeline.