What Is a Lookalike Audience? — Free Tool
A lookalike audience is one the platform builds by finding users who resemble a source list you supply. Learn why the source list decides the quality, what size does to precision, and the mistakes that produce an audience that resembles nothing useful.
An ad audience the platform builds by finding users who resemble a source list you supply — customers, site visitors or engagers. Quality depends almost entirely on how good and how specific the source list is.
What it is
A lookalike audience is an advertising audience the platform assembles for you: you supply a source list of people, and it finds users who resemble them on whatever signals it holds. You never see what those signals are, which is the honest description of the tool. You are not choosing an audience, you are choosing an example and asking for more like it.
How it is measured
Everything depends on the source. The platform can only extend the pattern it is given, so the question that decides the result is what your list actually describes. A list of purchasers describes buyers. A list of all site visitors describes wherever your traffic comes from, which is usually your existing marketing. Size is the second dial: a narrower lookalike sits closer to the source and costs more per result, a broader one reaches further and resembles the source less.
Commonly misunderstood
People treat it as a targeting feature and tune the percentage, when the percentage is the small decision and the source list is the large one. The related error is feeding in the biggest list available on the theory that more data is better. More data is better only when it is data about the same thing; a list that mixes buyers, bounced visitors and a competition entry has no single pattern to extend, and the audience it produces resembles the mixture rather than your customers.
When it matters
It matters when you have run out of people who already know you and want to grow beyond retargeting, which is the point most accounts reach after their first few months of paid work. It also matters as a diagnostic: if a lookalike of your best customers performs no better than a broad audience, that is worth knowing about your targeting assumptions.
Related terms
Features
- Plain definition, with the source list as the whole story
- Why a small list of your best customers beats a large list of everyone
- What the percentage size actually trades away
- The common failure: a source list that describes your existing marketing, not your customers
- How it differs from retargeting, which is a different audience entirely
Frequently asked questions
What makes a good source list?
Specificity, not size. A few hundred of your highest-value customers usually produces a better audience than everyone who ever visited the site, because the second list mostly describes who your existing marketing already reached.
How large should the lookalike be?
Smaller means closer to the source and usually more expensive per result; larger means broader reach and more dilution. Starting narrow and widening once it stops delivering is the ordinary approach, because you can see the trade rather than guess it.
Is a lookalike the same as retargeting?
No, and they are close to opposites. Retargeting shows ads to people who already interacted with you. A lookalike finds people who never have, chosen because they resemble those who did.
Why did my lookalike perform badly?
Usually the source list. A list built from all site visitors is a portrait of your traffic sources, so the lookalike will resemble your existing marketing rather than your customers, and it will find more of the same people you were already reaching.
Does it work without a large customer list?
It needs a minimum, and below that the platform cannot find a pattern worth extending. Small businesses often get more from a well-defined interest or retargeting audience than from a lookalike built on too few people.
