Can a B2B advertiser run a Google Ads Conversion Lift study on Search or Performance Max now, and do we have enough conversions?

TofuAnswer by the Tofu team · checked

A paved park path splits into two directions around a patch of grass among green trees

Yes, if your account qualifies. Google's setup page now lists Search and Performance Max for Conversion Lift, with a bar of 1,000 observed conversions and a $5,000 budget. PPC News Feed reported on October 7, 2026 that both used to need a Google rep, but that page and others still say Conversion Lift isn't available for all accounts, so check Lift studies in your account. An account that tracks only demo requests may fall short, and Google doesn't say over what period it counts the 1,000. Below the bar, a campaign experiment is the fallback.

Ask Tofu next

Pick a question and Tofu Ads Agent answers it in chat.

As of November 2026.

What changed for Search and Performance Max?

Google rewrote its help page "Set up Conversion Lift based on users." We found no announcement from Google, and Google doesn't say when the page changed. PPC News Feed reported it on October 7, 2026 and Search Engine Roundtable on October 8. Relevant Audience, an agency, wrote on October 9 that the April 2026 version told anyone interested in "Display, Search, Shopping or Performance Max" to "reach out to their Google account representative."

The page now lists supported campaign types as "Display, Search, Video, Demand Gen, App Campaigns and Performance Max." Its eligibility section asks you to contact a rep only "For new campaign types in alpha or beta phase." App campaigns targeting iOS and Travel Ads aren't supported. A campaign can be in only one lift study at a time.

Google's own pages don't agree yet:

  • The setup page itself, "About Conversion Lift" and the geography setup page all still open with "Conversion Lift isn't available for all Google Ads accounts. To use Conversion Lift, contact your Google account representative."
  • "Comparing lift types" still says user-based Conversion Lift is available for Video, Discovery and Demand Gen, and to reach out to your rep for Search and Performance Max.

So the real test is your own account. When you pick campaigns for a study, Google marks each one Eligible, Needs attention (inactive, in another study, or not compatible) or Not found.

Don't confuse it with Search Lift. Search Lift measures whether YouTube and Demand Gen ads raise searches for your brand, and Google says you can't use it without a Google rep.

Do we have enough conversions?

Check four numbers from Google's help:

RequirementWhat Google says
Conversions"at least 1,000 observed conversions." Modeled ("supplementary") conversions don't count
Budget"a minimum campaign budget of $5,000 USD." Google also lists spend below $5,000 as a reason results don't show
TrackingAt least one conversion action compatible with Conversion Lift
ResultsShown only after 150 conversions in the test group and 65 in the holdback group, from the fourth reporting day (it can take up to 10 days)

Google doesn't say over what period it counts the 1,000, or whether secondary conversion actions count. Relevant Audience noted the first gap too. For scale: at 80 demo requests a month, 1,000 is about a year of leads.

Ways to get closer, from Google's help:

  • Measure a more frequent action too. Google says conversions that happen more often raise certainty, and that page views or lead forms "show higher lift." It suggests measuring a bottom-funnel action plus an upper- or mid-funnel one as a secondary KPI.
  • Send CRM leads with enhanced conversions for leads. Google lists these under "Directly measured conversions," while it models other offline leads ("supplementary" conversions). If your CRM uploads aren't working yet, see how to upload offline conversions to Google Ads.
  • Include every campaign of the same type. Google says testing only a subset "reduces your chances of detecting lift."
  • Use a bigger holdback or a longer study. You can hold back 1% to 50% of users. Smaller holdbacks need longer studies, and Google suggests up to 56 days to raise certainty.

During setup, Google shows "study power," your expected certainty of lift, from 50% to 95%. It suggests aiming for 90% and offers budget guidance below that. Results between 50% and 90% are directional.

Conversion Lift by users vs by geography vs a campaign experiment

Conversion Lift based on usersConversion Lift based on geographyCampaign experiments
What it measuresConversions your ads caused, against users held back from seeing themConversions your ads caused, against regions where they don't runWhich of two campaign versions did better (custom), or what adding Performance Max adds (uplift)
Search and Performance MaxListed as supported (first reported October 7, 2026)SupportedCustom experiments cover Search, Display, Video and Hotel. Performance Max has its own experiments
Who sets it upYou, under Campaigns > Experiments > Lift studies, if your account qualifiesYou, under Goals > Lift measurement, once your Google account representative gives access (a beta)You, under Experiments in the Campaigns menu
Minimums1,000 observed conversions, $5,000 budget, one compatible conversion actionA compatible conversion action, or offline data sent to your account team. Campaigns must target one countryCustom: an active campaign with Smart Bidding and exact or phrase match keywords. No conversion minimum published
CRM leadsEnhanced conversions for leads are measured directlyOffline conversions summed to ZIP code or cityThe conversion metrics in your account
LengthAt least 7 days. Google recommends more than 14Not stated. A feasibility status (High, Medium, Low) shows at setupGoogle suggests 2 to 3 weeks
CostNo fee stated. You give up conversions from the held-back usersNo fee stated. Google says the budget needed "tends to be higher" than user-based studiesNo fee stated. The experiment shares the campaign's traffic and budget
Main limitThe 1,000-conversion bar, and Google's pages disagree on accessNeeds a rep, is "noisier," and has no demographic breakdownsShows which version won, not what your ads caused against no ads

When Conversion Lift based on users is the right pick

Pick it when you clear the bar and want to know whether Search or Performance Max brings leads you would have got anyway. A brand campaign is a common question; see whether to bid on your own brand name. Before you launch, Google suggests enhanced conversions for web and for leads, consent mode and Google tag gateway. During the study, keep to everyday changes such as bids and budgets. Google says new creatives or audiences "might impact" the result. Once a study ends, you can't restart it, and Google says most advertisers run one or two a year.

When Conversion Lift based on geography is the right pick

Pick it when you have a Google rep, sell mainly in one country, and want outcomes from outside Google Ads, such as opportunities from your CRM, summed by city or ZIP code. Google offers it as a beta through your rep, and it supports only holdback and go-dark studies.

When a campaign experiment is the right pick

Pick it when you're below the 1,000-conversion bar, or when you want to test one change (a bid strategy, broad match, adding Performance Max) rather than the whole channel. It tells you which version did better. It doesn't tell you what the campaign adds against showing no ads. Meta's lift test has its own budget and conversion bar; see what Meta requires for a Conversion Lift study.

Sources (13)

Related answers

More on