What is a lookalike audience?

Also called lookalike, similar audience.

A lookalike audience is a group of new people an ad platform finds because they share characteristics with a source audience you provide, such as your customers, website visitors or followers.

Why it matters

Finding new customers is the hardest part of advertising. A lookalike audience uses what the platform knows about your best existing customers to find others like them, which is often more effective than guessing interests yourself, especially for a small business without a marketing team. It also keeps working as your customer base grows, because a better source gives the platform more to learn from.

Example

A subscription coffee roaster uploads a list of its 600 repeat customers (with their consent to be used this way) as a custom audience, creates a lookalike from it, and runs a first-bag discount ad to that lookalike in its home country.

How lookalike audiences work on Meta

From Meta's Business Help Centre, checked on 14 October 2026. Meta uses information such as demographics, interests and behaviours from your source audience to find people who share similar qualities. When you create the lookalike, you choose a percentage range: smaller percentages match the source more closely, larger ones make a bigger, broader audience.

Meta says the source needs at least 100 people and generally recommends a source of 1,000 to 5,000 people. People in your source audience are excluded from the lookalike, so the ad reaches new people. Saved audiences can't be used as a source. Quality matters as well as size: Meta notes you may get better results from a source of your best customers than from all of them.

How to create one

The basic steps in Meta's tools:

  1. Create or choose a source: a custom audience such as website visitors, a customer list, video viewers or Page engagers.
  2. In Audiences (or in Ads Manager when building an ad set), select Create audience, then Lookalike audience.
  3. Pick your source and choose the audience size as a percentage.
  4. Select Create audience. It can take some time to populate, but you can start using it straight away.
  5. Use it in an ad set; the ad set's location settings decide where Meta looks for similar people.

When not to use a lookalike audience

A lookalike is only as good as its source. If your list is small, out of date or full of one-off bargain hunters, the lookalike will copy that. Meta also limits or removes lookalike audiences for some ads about financial products and services, employment or housing. And for a very local business, such as a hairdresser serving one town, simple location targeting may work just as well, because the pool of nearby people is small anyway.

Other ad platforms offer similar audience-expansion tools under their own names. The idea is the same, but the rules and minimum sizes differ, so check each platform's help pages before you build one.

Common questions

What size lookalike audience should I choose?

Start with the smallest percentage, which matches your source most closely, then test larger ones if you need more reach. Compare results on cost per result rather than audience size.

Can I make a lookalike from my Instagram followers?

You can create a custom audience of people who interacted with your Instagram business account and use that as the source for a lookalike, as long as it is large enough.

Is a lookalike audience better than interest targeting?

Not always. Lookalikes often work well when you have a good source; interest targeting can be better for a new business with no customer data yet. Test both with the same ad.

Does a lookalike audience update by itself?

On Meta, the audience shows a Last updated date for when it was last refreshed, and you don't need to wait for an update to use it. If your source changes a lot, such as a much bigger customer list, creating a new lookalike from it is worth testing.

Sources: Meta Business Help Centre: about lookalike audiences, checked 14 October 2026; Meta Business Help Centre: create a lookalike audience, checked 14 October 2026; Meta Business Help Centre: about custom audiences, checked 14 October 2026

Go further