Glossary /
Lookalike Audience
Efficiency & cost
Lookalike Audience
Find new prospects who resemble your best existing customers — by handing a platform a seed and letting it model the rest.
Attribution & Measurement
Unit Economics
Data Governance & Nomenclature
Creative & Delivery
Audiences & Targeting
Mobile & Privacy
A lookalike audience is a group of new users a platform identifies as similar to a seed or source audience you provide, used to scale prospecting beyond people you already reach.
?
?
What is Lookalike Audience?
A lookalike audience starts with a seed: a list of customers, converters, or high-value users you supply. The platform analyzes the traits and behaviors that group shares, then finds other users who resemble them. The promise is reach with relevance — new people who look like the ones you already know convert. The quality of a lookalike is mostly the quality of its seed. A small, noisy, or low-intent seed produces a weak model; a clean seed of genuine high-value customers produces a much sharper one. Seed size, recency, and signal all matter more than the size of the resulting audience. Where it gets confusing is across platforms. The concept is shared, but the product names and the controls aren't. Some platforms let you dial similarity tightly; others have folded lookalike modeling into automated targeting where you have far less direct say over how broadly the model casts.
How Lookalike Audience differs across ad platforms
Meta
Meta's Lookalike Audiences let you set a similarity range from 1% to 10% of a chosen country's population — 1% is tightest and most similar to the seed, 10% is broadest. You build them from a source like a Custom Audience, pixel data, or a customer list.
Google deprecated Similar Audiences in 2023 and removed them from reporting in 2023. The capability is now absorbed into optimized targeting and audience expansion, where Google's automation finds similar users on your behalf — you no longer manage a discrete similarity percentage the way you once did.
TikTok
TikTok offers Lookalike Audiences built from a source Custom Audience, with breadth settings (narrow, balanced, broad) rather than a precise percentage. The modeling is TikTok-specific and tuned to its engagement signals.
Common Lookalike Audience misconceptions
A bigger lookalike audience reaches better prospects.
Broadening a lookalike (e.g., 1% to 10% on Meta) trades similarity for scale. The wider the audience, the less it actually resembles your seed. Bigger reaches more people but dilutes the very similarity that makes a lookalike useful.
Google still has Similar Audiences, I just need to find the setting.
Google deprecated Similar Audiences in 2023. The functionality didn't disappear but it moved into automated systems — optimized targeting and audience expansion — so there's no longer a standalone Similar Audience you select and control directly.
Related Terms
Frequently Asked Questions
What is Lookalike Audience in simple terms?
A lookalike audience is a way to reach new people who resemble your existing customers. You give a platform a list of people you value, and it goes and finds others who behave similarly so you can target them.
How does Lookalike Audience work?
Why does Lookalike Audience differ across ad platforms?
How does Clarisights report on Lookalike Audience?
Attribution & Measurement
Unit Economics
Data Governance & Nomenclature
Creative & Delivery
Audiences & Targeting
Mobile & Privacy
A lookalike audience is a group of new users a platform identifies as similar to a seed or source audience you provide, used to scale prospecting beyond people you already reach.
?
?
What is Lookalike Audience?
A lookalike audience starts with a seed: a list of customers, converters, or high-value users you supply. The platform analyzes the traits and behaviors that group shares, then finds other users who resemble them. The promise is reach with relevance — new people who look like the ones you already know convert. The quality of a lookalike is mostly the quality of its seed. A small, noisy, or low-intent seed produces a weak model; a clean seed of genuine high-value customers produces a much sharper one. Seed size, recency, and signal all matter more than the size of the resulting audience. Where it gets confusing is across platforms. The concept is shared, but the product names and the controls aren't. Some platforms let you dial similarity tightly; others have folded lookalike modeling into automated targeting where you have far less direct say over how broadly the model casts.
How Lookalike Audience differs across ad platforms
Meta
Meta's Lookalike Audiences let you set a similarity range from 1% to 10% of a chosen country's population — 1% is tightest and most similar to the seed, 10% is broadest. You build them from a source like a Custom Audience, pixel data, or a customer list.
Google deprecated Similar Audiences in 2023 and removed them from reporting in 2023. The capability is now absorbed into optimized targeting and audience expansion, where Google's automation finds similar users on your behalf — you no longer manage a discrete similarity percentage the way you once did.
TikTok
TikTok offers Lookalike Audiences built from a source Custom Audience, with breadth settings (narrow, balanced, broad) rather than a precise percentage. The modeling is TikTok-specific and tuned to its engagement signals.
Common Lookalike Audience misconceptions
A bigger lookalike audience reaches better prospects.
Broadening a lookalike (e.g., 1% to 10% on Meta) trades similarity for scale. The wider the audience, the less it actually resembles your seed. Bigger reaches more people but dilutes the very similarity that makes a lookalike useful.
Google still has Similar Audiences, I just need to find the setting.
Google deprecated Similar Audiences in 2023. The functionality didn't disappear but it moved into automated systems — optimized targeting and audience expansion — so there's no longer a standalone Similar Audience you select and control directly.
Frequently Asked Questions
What is Lookalike Audience in simple terms?
A lookalike audience is a way to reach new people who resemble your existing customers. You give a platform a list of people you value, and it goes and finds others who behave similarly so you can target them.
How does Lookalike Audience work?
Why does Lookalike Audience differ across ad platforms?
How does Clarisights report on Lookalike Audience?
See Viewable Impression across every channel in one report
See Viewable Impression across every channel in one report
See viewability and vCPM across every platform—display, video, programmatic— in one normalized report, instead of reconciling vendor numbers by hand.
Book a demo