Audience Targeting

Interest Targeting vs Lookalike Audiences

In short: Both extend reach beyond people you already know, but they start from opposite ends of the evidence spectrum. Interest Targeting starts with zero data of your own and asks Meta to guess based on generic topic affinity. Lookalike Audiences start with a seed list of your real customers and ask Meta to find strangers who resemble that seed on the signals it holds, sized as a percentage of a country. A lookalike is only as good as its seed - a small precise seed with real value data beats a huge unfiltered one - while interest targeting has no seed dependency at all, just a taxonomy pick. Use interest targeting when you have no customer data to seed from, and switch to a lookalike the moment you have a decent seed list, since it almost always outperforms a guessed interest.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Interest Targeting

The subset of Meta's detailed targeting built from topics and pages people engage with, used to approximate affinity when no first-party signal exists yet.

Interests are inferred from Facebook and Instagram activity, so they describe engagement patterns rather than verified intent, and Meta shows you nothing about how a user qualified. They earn their place at the top of the funnel when you have no pixel history to work from. The usual error is layering dozens of interests, producing an audience indistinguishable from broad targeting anyway.

Full definition

Lookalike Audiences

Meta audiences modelled on a seed list, matching users who resemble your source on the signals Meta holds, sized as a percentage of a chosen country.

You pick a source - purchasers, high-value customers, engaged visitors - and a percentage; smaller percentages mean closer resemblance and less reach, larger ones the reverse. A value-based source lets Meta weight by revenue. Their edge has narrowed as broad targeting and Advantage+ audience improved. The persistent error is seeding from a large low-quality list, which models the wrong person very precisely.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Interest Targeting	Lookalike Audiences
Starting point	A topic or behavior category picked from Meta's taxonomy.	A seed list of real people - customers, purchasers, high-value buyers.
Requires your own data	No.	Yes - needs a seed audience of at least a workable size to model well.
Sizing control	Broad or narrow by stacking or removing interests.	A percentage slider - smaller means closer resemblance, larger means more reach.
Weighting by value	Not possible - every person matching the interest is treated equally.	Possible - a value-based seed lets Meta weight toward your highest-revenue customers.
Visibility into match logic	None - Meta doesn't show how a person qualified for the interest.	None - Meta doesn't show which specific seed traits drove a match either.
Dependency on source quality	None, since there's no seed.	High - a large, low-quality seed models the wrong person with confidence.
Competitive edge today	Narrow but still useful for very specific, well-defined niches.	Narrowed as broad targeting and Advantage+ audience have improved on their own.

What actually separates them.

Interest targeting needs no data from you at all, while a lookalike audience cannot exist without a seed list you provide first.

A lookalike's quality is a direct function of seed quality and size, while an interest's quality is fixed by Meta's taxonomy and unaffected by anything in your account.

Lookalikes let you weight toward high-value customers using a value-based seed; interest targeting has no equivalent, every matched person counts the same.

Lookalike audience size is controlled by a percentage-of-country slider with a clear tradeoff between resemblance and reach, while interest audience size is controlled by how many interests you stack, with no equivalent precision dial.

Both keep matching logic hidden from you, but a lookalike's opacity centers on which seed traits mattered while an interest's opacity centers on how a stranger qualified in the first place.

Which one should you use?

Use Interest Targeting when

You have no customer list or pixel data to build a seed from yet.
You're testing a brand-new product category with no existing buyer pattern.
You want very specific niche reach that a broad lookalike percentage would dilute.
You're prospecting on a fresh account before enough conversions exist to seed anything.

Use Lookalike Audiences when

You have at least a modest list of real purchasers or high-value customers to seed from.
You want Meta to weight toward your best customers using a value-based seed.
Broad targeting or Advantage+ audience isn't finding enough of your best-fit buyers on its own.
You want a tunable dial between tight resemblance and larger reach rather than a fixed interest category.

Common questions.

How small can a lookalike seed be and still work?

Meta recommends a seed in at least the low thousands of people to model well, though it will technically build one from a smaller list. Below that, the model has too little signal to generalize, and the resulting lookalike tends to perform inconsistently.

Are lookalike audiences still worth using now that Advantage+ audience exists?

For many accounts, less than before, since Advantage+ audience's automated expansion covers some of the same ground a lookalike used to. They still earn their place when you have a genuinely high-value seed, like top-decile purchasers, that you want the system weighting toward specifically, rather than trusting fully automated expansion to find.

Can I build a lookalike from an interest-targeted audience?

Not directly - a lookalike needs a seed of real people, which is a custom audience, and an interest selection isn't a list of people, it's a targeting rule. You'd first need to run a campaign to the interest, capture the resulting engagers or converters as a custom audience, and seed the lookalike from that.

Does a 1% lookalike mean the top 1% best customers?

No, the percentage refers to audience size relative to the chosen country's population, not a ranking of customer quality. A 1% lookalike is simply the group Meta judges to most closely resemble your seed among everyone in that country; the seed's quality, not the percentage, determines how good those matches actually are.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons