Lookalike audience

Why it matters

A lookalike audience lets you ask Meta, "find me more people like these." You provide a source, such as a customer list, people who purchased, or high-value leads, pick a country and a size, and Meta finds people who resemble it. A smaller percentage means a closer match and a smaller reach; a larger one widens the pool and loosens the match. Meta's help centre describes the source and setup in its lookalike audiences guide.

The quality of the source is the whole game. A lookalike of everyone who ever filled in a form finds more form fillers. A lookalike of customers who paid and returned finds more of those.

How to use it well

1. Start from value, not volume. Use paying customers, repeat buyers or qualified leads rather than all website visitors.
2. Keep the source clean and large enough. Meta sets requirements for the source audience (see its help page for the current ones); in practice, larger, fresher lists give the system more to learn from.
3. Test sizes. Run 1 percent against a broader range, and against Advantage+ audience, on equal budgets and compare the cost per customer.
4. Refresh the source every few months as your customer base changes.
5. Exclude existing customers so you are paying to find new ones.

Common mistakes

• Using the wrong source, such as page likes, which describes fans but not buyers.
• Never comparing against broad targeting. Meta's automation may now do as well or better.
• Forgetting privacy rules when you upload customer data; you need permission to use it for ads.
• Expecting a fixed result. Lookalike quality varies by source and market.

BridgeE tests lookalikes against broad and automated audiences inside Meta Ads and performance marketing so the choice rests on your numbers.