What is a lookalike audience?
A lookalike audience is a targeting option in social media advertising. You give the ad platform a group of people who already matter to you, such as buyers or newsletter subscribers. The platform studies what they have in common and finds other users who look similar. Then it shows your ads to those new people, who are more likely to care than a random crowd.
You never see who is in a lookalike audience. You only see its rough size and the results your ads get. The platform does the matching.
How lookalike audiences work
Everything starts with a source audience, sometimes called a seed. Common sources include:
- A customer list you upload, such as emails from past buyers.
- People who visited your website, tracked with a pixel or tag.
- People who engaged with your page, videos or lead forms.
- Your existing followers.
The platform matches that list to its users and looks for patterns in their activity and interests. Then it builds a new audience of people who share those patterns. Most platforms let you choose how close the match should be. A tight match is smaller and more alike. A loose match is bigger but less precise.
Choosing a source and a match size
The source decides what the platform looks for. If you upload everyone who ever bought from you, it looks for people like your average buyer. If you upload only customers who bought three times or more, it looks for people like your best buyers. A smaller, sharper source often finds better people, as long as it meets the platform's minimum size.
Here is an example. A pet food shop has a list of past buyers. It makes one lookalike from all buyers and a second one from buyers on a monthly plan. The second audience is smaller, but the people in it are more likely to sign up for a plan too.
Match size works the same way. A close match stays near your source and reaches fewer people. A wide match reaches many more people who share fewer traits. Start close, see what it costs to get a result, then try wider matches once the close one runs out of new people.
How to create a lookalike audience
- 1Choose a source
Your best customers or most engaged fans.
- 2Check permission
Use only data you collected with consent.
- 3Upload and pick a region
Add the source in the ad manager and choose where to reach.
- 4Set the match size
Start close, then test wider matches.
- 5Test against others
Same ad, same budget, different audiences.
- Pick your best source. Your highest-value customers teach the platform more than everyone who ever clicked.
- Check permission. Make sure you collected the data with consent and follow privacy laws.
- Upload or select the source. Do this in the platform's ad manager.
- Choose a location. Pick the country or region you want to reach.
- Set the match size. Start with a close match, then test wider ones.
- Exclude current customers. There's no point paying to reach people who already bought.
- Run a test. Compare results against other audiences with the same ad and budget.
How to test and measure a lookalike
The fair way to judge a lookalike is a side-by-side test. Run the same ad with the same budget to a few audiences at once, such as a close lookalike, a wider lookalike, an interest audience and broad targeting. Then compare the rate at which each one turns viewers into sign-ups or sales.
One ad shown to four audiences with the same budget
| Sign-up rate | |
|---|---|
| Close lookalike | 3.2% |
| Wide lookalike | 1.9% |
| Interests | 1.4% |
| Broad | 2.1% |
In the test above, the close lookalike won. In real life, the result can go any way. Platforms have grown better at finding buyers with broad targeting, so lookalikes don't always win. The only way to know for your ads is to test.
Watch these numbers for each audience:
- Click-through rate: how many people who saw the ad clicked it.
- Conversion rate: how many clickers signed up or bought.
- Cost per result: what you paid for each sign-up or sale.
- Audience size: whether it is big enough to keep spending on.
Lookalike audience vs retargeting
Retargeting shows ads to people who already know you, like someone who left items in their cart. Lookalike audiences reach new people who have never heard of you but resemble your fans. Retargeting helps close sales. Lookalikes help find new customers. Many ad plans use both, with lookalikes filling the top of the funnel and retargeting bringing people back.
| Lookalike audience | Retargeting | |
|---|---|---|
| Who sees the ad | New people like your customers | People who already know you |
| Main job | Find new customers | Bring people back to buy |
| Needs a source list or pixel | ||
| Audience size | Usually large | Usually small |
| Place in the funnel | Top | Bottom |
Common mistakes
- A source list so small that the platform has little to learn from.
- A source full of low-quality leads, which teaches the platform to find more of the same.
- Mixing very different customers in one source, which blurs the match.
- Forgetting to exclude current customers.
- Judging a lookalike after a day or two, before it has enough results to compare.
Tools that help
Meta Ads ManagerMeta's tool for building ad audiences and running ads on Facebook and Instagram.Visit site
LinkedIn Campaign ManagerLinkedIn's tool for creating ad audiences and running ads to professionals.Visit siteUTM builderOur toolTag your ad links so you can see which audience brought visitors to your site.Open
Google AnalyticsWebsite analytics for seeing what ad visitors do after they land.Visit site
Related terms
- RetargetingRetargeting is showing ads to people who already visited your website, used your app or engaged with your content, to bring them back.
- Target audienceA target audience is the specific group of people you want your posts and ads to reach, described by who they are and what they need.
- Paid socialPaid social is advertising on social media apps, where you pay to show posts or ads to people you choose by interest, location, age or behavior.
- PsychographicsPsychographics are the values, interests, attitudes and lifestyle choices that explain why people act the way they do, used to understand an audience.
- Boosted postA boosted post is an organic post you pay to show to more people, turning it into a simple ad without building a full ad campaign.