What is Lookalike / Similar Audiences?
A lookalike audience is a targeting segment built by an ad platform's machine learning to find new users who resemble an existing source audience (such as your customer list or website converters). Meta calls them Lookalike Audiences; Google deprecated standalone 'Similar audiences' in 2023, folding the capability into Smart Bidding's automatic optimized targeting.
What to know in practice
- Source-audience quality determines lookalike quality. A seed list of your best customers produces a far stronger lookalike than a seed of all leads including unqualified ones.
- Google sunset manually-managed Similar Audiences in August 2023 — the modeling now happens automatically inside Smart Bidding and Performance Max when you supply Customer Match or conversion data.
- On Meta, smaller, tighter lookalike percentages (1%) resemble the seed most closely; larger percentages (5-10%) widen reach but dilute similarity.
- Lookalikes are a prospecting tool, not a retargeting tool — they find new people, so they should be measured on cost-per-new-customer, not blended ROAS.
Common misconception
Bigger source audiences don't make better lookalikes. A focused seed of 1,000 high-value customers usually outperforms a seed of 50,000 mixed-quality leads, because the model learns from signal, not volume.
Related terms
- Customer Match → Paid Media
- Smart Bidding → Paid Media
- Performance Max → Paid Media