PPC

Google Ads gives Demand Gen advertisers a way to keep Lookalike targeting restrictions

Google Ads is preparing to change how Lookalike audience segments work in Demand Gen campaigns, moving their default behavior toward audience suggestions rather than strict targeting restrictions.

Why it matters. Lookalike segments have traditionally helped advertisers reach new people who share characteristics with their existing customers. Moving toward suggestion-based targeting could give Google’s AI more flexibility to find conversions, but it also means advertisers may have less control over who sees their ads.

What’s changing. Google says Lookalike segments will evolve toward a suggestion mode by default. Advertisers who want to preserve the original behavior can opt out and continue using Lookalike segments as a targeting constraint.

The key difference. With targeting constraints, advertisers can restrict delivery to their selected Lookalike audience. Suggestion-based targeting gives Google’s systems more flexibility to use that audience as a starting point rather than a strict boundary.

How it works. Lookalike segments use first-party data, including customer lists, website visitors, app users and YouTube audiences, to identify potential customers with similar characteristics.

The targeting options: Advertisers can choose between three Lookalike audience sizes:

  • Narrow: Targets the most similar 2.5% of users in the selected location.
  • Balanced: Expands the audience to 5%.
  • Broad: Reaches up to 10%, trading some similarity for additional scale.

The requirements. Lookalike segments need more than 100 active matched users across their seed lists. Google refreshes these audiences every one to two days, helping keep targeting aligned with recent customer activity.

Why we care. The change could significantly affect how Demand Gen campaigns reach potential customers. Advertisers relying on tightly defined Lookalike audiences may want to preserve targeting constraints, while those prioritizing reach and conversion volume may prefer Google’s more flexible approach.

Yes, but. Expanding beyond a Lookalike audience doesn’t automatically mean better performance. Advertisers should monitor conversion quality, acquisition costs and customer value to determine whether broader targeting delivers genuinely useful growth.

The bottom line. Demand Gen advertisers will have a choice between maintaining strict Lookalike targeting and allowing Google’s AI more flexibility. Those who want to preserve the existing behavior should review Google’s opt-out option before the default changes.

Dig deeper. Google Ads — Use Lookalike segments in Demand Gen campaigns as a Targeting Constraint

Explore more on

Artificial intelligence Demand generation Google Ads Targeted advertising Google YouTube
Add Search Engine Land as a preferred source on Google

Editorial disclosure

Search Engine Land is owned by Semrush. We remain committed to providing high-quality coverage of marketing topics. Unless otherwise noted, this page’s content was written by either an employee or a paid contractor of Semrush Inc.

About the author

Anu Adegbola

Paid Media News Editor, Search Engine Land

Anu Adegbola has been Paid Media Editor of Search Engine Land since 2024. She covers paid search, paid social, retail media, video and more.

In 2008, Anu started her career delivering digital marketing campaigns (mostly but not exclusively Paid Search) by building strategies, maximising ROI, automating repetitive processes and bringing efficiency from every part of marketing departments through inspiring leadership both on agency, client and marketing tech side. Outside editing Search Engine Land article she is the founder of PPC networking event - PPC Live and host of weekly podcast PPC Live The Podcast.

She is also an international speaker with some of the stages she has presented on being SMX (US, UK, Munich, Berlin), Friends of Search (Amsterdam, NL), brightonSEO, The Marketing Meetup, HeroConf (PPC Hero), SearchLove, AdWorld Experience (Bologna, IT) and more.