How lookalike audiences work
- You provide a source audience, for example a customer list or the people who bought on your website.
- The platform analyzes what those people have in common.
- It builds a new audience of similar users, usually in a country or region you choose, and you target your ads at them.
Where you can use them
- Meta (Facebook and Instagram): lookalike audiences from 1% (most similar) to 10% of a country’s users.
- LinkedIn: predictive audiences based on your data.
- TikTok and other platforms: similar audience tools.
- Google Ads: similar audiences were retired in 2023 and replaced by optimized targeting and audience expansion.
Tips
The better the source, the better the lookalike: use your best customers, not everyone who ever visited. Many platforms now find similar people automatically with broad targeting, so test lookalikes against broad audiences and keep what performs.
Lookalike audience example
An online store uploads a list of its 2,000 best customers, those who bought at least twice, to Meta. It creates a 1% lookalike audience for its country, and runs ads for a new product to them. Compared with a broad interest audience, the lookalike group has a lower cost per sale, so the store keeps testing 1% against 3% and refreshes the source list every few months.
How to build a good lookalike audience
- Choose a source that reflects what you want more of: best customers, not all visitors.
- Use a source of at least a few hundred good people, with clean data.
- Get consent for using customer data, and follow the platform’s data rules.
- Start with a small percentage and test larger ones.
- Exclude people who are already customers.
- Refresh the source as the customer base changes.
Lookalike audience mistakes
- Using a weak source, such as everyone who visited the site once.
- Uploading personal data without a legal basis.
- Making the audience too broad too soon.
- Leaving the same audience running for a year with no refresh.
Lookalike audience FAQs
Do lookalike audiences still work with privacy limits?
They can, but tracking limits mean the quality of source data matters even more. See remarketing and customer data platforms.
Where are they available?
On Meta, and similar tools on other platforms. See Instagram advertising and LinkedIn advertising.