Data & identity · also called Similar audience, lookalike modelling
Lookalike audience
A lookalike audience is a group of new people who statistically resemble an advertiser's existing customers or best users, found by a model to extend reach beyond known contacts.
Start with a seed list, such as your best customers, and a lookalike model finds people who share their characteristics and behaviours but are not yet customers. Platforms like Meta, Google, TikTok and LinkedIn popularised the approach, and DSPs, data providers and retail media networks offer it too.
The model compares the seed group's attributes, such as interests, demographics, content consumed, apps used and purchase patterns, against the wider population, then ranks everyone by similarity. Advertisers usually choose a trade-off between closer similarity with smaller reach and broader reach with looser similarity.
Lookalikes are only as good as the seed and the data behind the model. A seed polluted by fraudulent sign-ups will teach the model to find more fraud-like users, a real risk in app install campaigns. Uploading customer lists to create seeds involves personal data, so consent and hashing requirements apply, and signal loss has reduced accuracy on some platforms.
Think of it like this
A lookalike audience is like a matchmaker saying: your best friends are like this, so here are strangers you would probably get along with too.
An example
An Indonesian e-wallet uploads a hashed list of 50,000 high-value users; the platform builds a 2 million-person lookalike, which converts at twice the rate of broad targeting.
Related terms
Audience segment
An audience segment is a group of people or devices defined by shared characteristics, such as demographics, interests, purchase behaviour or intent, used to target or measure advertising.
First-party data
First-party data is information a company collects directly from its own customers or audience, such as purchases, sign-ups, app use and site behaviour, through its own channels.
Behavioral targeting
Behavioral targeting shows ads to people based on their past online behaviour, such as sites visited, searches, purchases or app use, collected over time and often across different services.
Hashed email
A hashed email is an email address transformed by a one-way cryptographic function, usually SHA-256, into a fixed string used to match users across systems without sharing the plain address.
Paid social
Paid social is advertising on social media platforms, such as Facebook, Instagram, TikTok, YouTube, LinkedIn, Snapchat, Pinterest and X, where ads appear in feeds, stories and short videos.