Lookalike audience

A lookalike audience is a set of new users an ad platform generates by matching the characteristics of an advertiser's existing customer list, a "seed" audience, against its own user base, then targeting people who share similar attributes or behavior patterns. The seed is typically built from first-party data, such as a customer list or a site-visitor audience, uploaded to the platform.

Lookalike targeting is essentially platform-run customer segmentation applied to prospecting: instead of an advertiser manually defining interest or demographic targeting, the platform's own propensity model finds statistical similarity across signals the advertiser cannot see directly. This is the opposite goal of retargeting, which re-engages people who already interacted with the brand rather than finding new prospects who merely resemble existing customers.

Lookalike audiences matter because they scale prospecting beyond an advertiser's own list while still targeting people more likely to convert than a broad, untargeted audience, and quality generally improves with a larger, cleaner seed list. The main pitfalls are a poor seed list, for example one mixing high-value repeat customers with one-time discount shoppers, which produces lookalikes that resemble the wrong customer; and audiences delivered via programmatic advertising that shrink or degrade in quality once narrowed by geography or other filters.