How to Create Facebook Lookalike Audiences: Full Guide

Here’s the honest, plain-English version: to create a Facebook lookalike audience, you first build a source custom audience made of people you already value — ideally your best customers or purchasers — then in Meta’s Audiences tool you tell it to find a lookalike of that source, choose the country you want to reach, and pick a percentage size that trades reach against similarity. Meta studies the shared traits of your source and finds new people who resemble them. The whole trick is this: a lookalike is only ever as good as the source you feed it. Get the source right and everything downstream gets easier.

If you’ve been staring at the Audiences screen wondering whether you’re about to waste money, take a breath — I’ve got you. Learning how to create Facebook lookalike audiences is really about understanding one relationship (source in, similar people out) and then testing your way to what works for your specific business. Let’s walk it together, calmly, like I’m sitting next to you talking you through each click.

What is a Facebook lookalike audience, really?

Okay, let’s be honest about what’s actually happening under the hood, because it demystifies the whole thing. A lookalike audience is Meta’s way of saying, “Show me a group of people you love, and I’ll go find more people who resemble them.” You hand Meta a source audience — a list of people who matter to your business — and Meta analyzes the traits and behaviors those people share, then builds a fresh audience of others across Facebook and Instagram who look similar in those ways. You never see the individual traits it uses; it’s pattern-matching at a scale no human could do by hand.

That’s the beauty and the catch of lookalikes in one breath. The beauty: you get to reach strangers who statistically resemble your happiest customers, instead of guessing at interests and demographics from scratch. The catch: Meta faithfully mirrors whatever you give it. Feed it a messy, low-intent source and it will diligently find you more low-intent people. So when we talk about how to create Facebook lookalike audiences, ninety percent of the real work is choosing a source worth cloning. Hold that thought — we’ll come back to it because it matters that much.

One quick note before we go further: Meta’s tools, menu names, and available options change over time, and they sometimes differ by country and account. Everything here is the durable method and the honest trade-offs — but always confirm the current wording and choices in your own Ads Manager rather than trusting any screenshot, including mine in spirit.

How do lookalike audiences fit into your Facebook Ads setup?

Before we build one, it helps to see where this sits in the bigger picture. A lookalike is a targeting tool — it answers the question “who should see this ad?” It doesn’t design your campaign, write your creative, or set your budget. So if you haven’t yet got the bones of a campaign in place, start with the foundation: my full walkthrough on how to set up a Facebook Ads campaign covers objectives, structure, and budgets, and lookalikes slot in at the ad-set targeting stage once that’s ready.

Lookalikes are one of three big targeting approaches you’ll weave together. There’s broad and interest-based targeting (covered in how to target audiences on Facebook Ads), there’s warm re-engagement of people who already know you (that’s how to retarget with Facebook Ads), and then there’s the lookalike — the bridge that takes what you’ve learned from your warmest people and uses it to find brand-new cold prospects who resemble them. Think of it as prospecting with a really good reference photo.

Why does the source audience matter more than anything else?

Here’s the part nobody tells you loudly enough: the quality of your source audience determines the quality of your lookalike. Everything else — the size, the placements, the creative — is a rounding error next to this one decision. If you remember nothing else from this guide, remember to be picky about your source.

Why does it matter so much? Because Meta builds the lookalike by finding the common threads among the people in your source. If your source is “everyone who ever visited my website,” you’ve handed Meta a blurry, mixed bag — bargain-hunters, accidental clicks, competitors snooping, people who bounced in two seconds. The resulting lookalike will be equally blurry. But if your source is “people who actually purchased in the last 180 days,” you’ve handed Meta a crisp, high-intent portrait, and it can find you more people who resemble genuine buyers.

Sources ranked from strongest to softest

Not all sources are created equal. Here’s how I mentally rank them, strongest signal at the top:

Source audience	Signal strength	When to reach for it
High-value purchasers (your best customers, repeat buyers, highest lifetime value)	Strongest	Your default first choice whenever you have enough of them.
All purchasers / customer list	Very strong	When you have a solid list of people who’ve actually bought.
Add-to-cart or lead form completions	Strong	Great when purchase numbers are still small; high intent short of buying.
Engaged social followers (video viewers, page/profile engagers)	Moderate	Useful early on before you have purchase data to work with.
All website visitors	Softer	A fallback when you have little else; expect a broader, fuzzier match.

See the pattern? The closer your source is to “money changed hands,” the sharper the lookalike. So the honest first question isn’t “how do I make a lookalike?” — it’s “what’s the best-intent group of people I can point Meta at?” Start there every single time.

You need enough people in the source

One practical constraint: a source can’t be tiny. Meta needs a minimum number of matched people in your source to build a lookalike at all, and generally the more (good-quality) source records you have, the better the pattern it can learn. If your purchaser list is too small to qualify, that’s your cue to step one rung down the ladder — use add-to-carts or leads instead — until your purchase data grows. Don’t pad a great source with low-quality records just to hit a number; a smaller, cleaner source beats a bigger, muddier one.

How do you build the source custom audience first?

You can’t make a lookalike out of thin air — you make it from a Custom Audience. So step one is always creating that source. In Meta’s Audiences tool (inside Ads Manager, under the Audiences section) you’ll create a Custom Audience from one of a few origins:

• Your customer list: A file you upload of customers — typically emails and/or phone numbers. This is often the highest-quality source because you decide exactly who goes on it (say, only buyers, or only your top spenders). Crucial honesty note: that list must be data you collected properly and have permission to use for advertising. More on that below, because it genuinely matters.
• Website activity (via the Meta Pixel): If you’ve installed the Meta Pixel on your site, you can build audiences from real behavior — all visitors, viewers of specific pages, add-to-carts, or purchasers. This is powerful because it captures intent as it happens. If your Pixel isn’t set up yet, that’s worth doing first; it unlocks the best website-based sources.
• App activity or offline events: If you have an app or feed offline conversions, those can seed audiences too.
• Meta-based engagement: People who engaged with your Facebook Page or Instagram profile, watched your videos, opened a lead form, and so on — built right from your Meta assets, no website required.

Pick the origin that gets you closest to real intent given what you have. If you sell online and have the Pixel firing, a purchaser-based website audience is a beautiful source. If you’re service-based with a customer spreadsheet, a clean uploaded list of actual clients is gold. If you’re brand-new with neither, your engaged video viewers or profile engagers are a perfectly reasonable starting point while you build up better data.

How do you actually create the lookalike, step by step?

Once your source Custom Audience exists, making the lookalike itself is genuinely quick. Here’s the flow, and remember the exact labels may shift, so match the intent rather than the literal words:

• 1. Open the Audiences tool in Ads Manager and choose to create a new Lookalike Audience.
• 2. Choose your source. Select the Custom Audience you just built (your purchasers, your customer list, your engagers). This is the decision that matters most — you already did the hard thinking above.
• 3. Choose the location. Pick the country or region where you want to find new people. The lookalike will be drawn from that place, so choose where you actually want customers.
• 4. Choose the audience size. You’ll set a percentage of the chosen location’s population — smaller means a tighter match, larger means broader reach. We’ll unpack this trade-off next.
• 5. Create it. Meta builds the audience (it takes a little time to populate) and it becomes available to select at the ad-set level in your campaigns.

That’s the mechanical part — honestly less intimidating than it looks from the outside. The lookalike now lives in your Audiences library, and when you’re building an ad set, you’ll pick it as your targeting. If you want a refresher on where that ad-set targeting step lives within the campaign build, that’s exactly what how to target audiences on Facebook Ads walks through.

What percentage size should you choose?

This is the question everyone gets stuck on, so let me make it simple. When you create a lookalike, you choose a size expressed as a percentage of the population in your selected country — commonly ranging from the smallest, closest match up to a much larger, broader one. The trade-off is beautifully intuitive once you see it:

• A smaller percentage (like 1%) gives you the people who most closely resemble your source. It’s the tightest, most “similar” match — but it’s a smaller pool, so reach is limited.
• A larger percentage widens the net to include people who resemble your source less closely. You reach far more people, but the average similarity drops, so the audience behaves more like broad targeting.

Here’s the honest guidance: there is no universal “best” size, and anyone who tells you “always use 1%” is oversimplifying. The right size depends on your source quality, your budget, how big your target country is, and what you’re optimizing for. A tiny, ultra-precise audience can be wonderful for a small premium offer, but starve it with too big a budget and performance can wobble. A broader lookalike gives Meta’s delivery system more room to optimize, which sometimes works better than you’d expect.

So rather than me inventing a number, do the thing that actually answers it: test. Set up the same creative against a couple of different lookalike sizes as separate ad sets, give each a fair shot, and let your own cost-per-result tell you the truth. Your data beats every generic rule of thumb, including mine.

Should you use a lookalike or just go broad?

I want to be straight with you here, because the paid-media world has genuinely shifted. For years, tightly crafted lookalikes were the crown jewel of Facebook targeting. But Meta’s delivery system has become dramatically more automated, and its machine learning increasingly favors broad audiences that give the algorithm maximum room to find the right people on its own. In many accounts today, a broad audience with strong creative and clean conversion tracking performs as well as — or better than — a carefully built lookalike.

That doesn’t mean lookalikes are dead. It means the honest answer is “it depends, so test both.” Lookalikes still shine when you have a genuinely excellent source (those high-value purchasers), when you’re in a smaller or more niche market where broad is too broad, or when you want a warm-ish prospecting layer that’s a notch more qualified than wide-open targeting. The smart move isn’t to pick a side in the lookalike-versus-broad debate — it’s to run them against each other in your own account and believe your results.

My practical take: if you’re just starting and your source is thin, don’t agonize — a broad audience with great creative is a totally legitimate first move, and you can layer in lookalikes as your purchase data grows. If you’ve got a rich source of real buyers, absolutely build the lookalike and test it head-to-head against broad. Either way, the creative and the offer usually move the needle more than the exact audience choice, so don’t let audience-tweaking become procrastination.

How do you keep lookalikes working over time?

A lookalike isn’t quite “set it and forget it,” but it’s close, with a few habits that keep it healthy:

• Refresh the source as it grows. As you accumulate more purchasers, your customer-based source gets richer. Many advertisers let Meta keep the lookalike updated from a dynamic source (like a Pixel purchase audience) so it evolves as new buyers come in, rather than freezing a snapshot.
• Exclude people you don’t want to prospect. When you run a lookalike for cold prospecting, exclude your existing customers and recent site visitors so you’re not paying to reach people you already have. Pair that exclusion thinking with a proper retargeting layer — here’s how to retarget with Facebook Ads so warm and cold audiences don’t step on each other.
• Watch for overlap and fatigue. If several ad sets target overlapping lookalikes, they can compete against each other. Keep an eye on audience overlap and creative fatigue, and rotate fresh creative regularly.
• Judge by cost per result, not vanity metrics. A lookalike with a lovely low cost-per-click but no sales is not winning. Anchor every decision to the outcome that actually makes you money.

What are the privacy and policy rules you can’t ignore?

I’d be doing you a disservice if I skipped this, because it’s not optional — it’s the responsible, and required, part of the job.

Your source data must be properly collected and consented. When you upload a customer list, you’re telling Meta these are people you have a legitimate relationship with and permission to use for advertising. Don’t upload lists you scraped, bought, or otherwise gathered without people’s knowledge. Beyond it being against the rules, it’s simply not okay to treat people’s contact information as a free-for-all. Collect data transparently, honor your privacy policy, and respect the consent and data-protection laws that apply where your customers live. Website- and Pixel-based audiences carry their own consent responsibilities too — make sure your site’s cookie and tracking notices reflect what you’re actually doing.

Special Ad Categories restrict lookalikes. This one catches a lot of people off guard, so hear me clearly: if your ads relate to housing, employment, or credit (and in some regions, social issues, elections, or politics), Meta classifies them under Special Ad Categories, which exist to prevent discrimination. In those categories, certain targeting is limited, and the lookalike tools are restricted — Meta offers a differently-built “special ad audience” concept in their place, and the exact rules evolve. If you advertise in any of these areas, do not assume standard lookalikes are available to you. Check Meta’s current policies and follow them to the letter. This protects real people from harm, and it protects your account from being shut down.

None of this is meant to scare you off — the vast majority of businesses use lookalikes perfectly appropriately. It’s just the grown-up part of paid media: reach people ethically, with data you’re allowed to use, and you’ll sleep fine and build something that lasts.

Where does organic social fit alongside your lookalike ads?

Here’s something the ads-only crowd tends to skip. When your lookalike audience does its job and sends a curious stranger to check you out, a lot of them will glance at your Facebook and Instagram profiles before they trust you with a click, a lead, or a purchase. And if they land on a profile that last posted seven months ago, some of that hard-won interest quietly cools. A living, consistent social presence is what turns “huh, interesting ad” into “oh, this is a real, active brand I can trust.”

Let me be totally clear about where SocialBlaze fits, because I never want to overpromise: SocialBlaze does not run, bid on, or manage Facebook ads or lookalike audiences. It isn’t an ad platform, and I’d never pretend it were. What it does is make your organic presence effortless — scheduling and auto-publishing your posts, and unifying your analytics and inbox, across Facebook, Instagram, and every other network from one calm place. So while your lookalike ads bring new people to the door, your organic feed is what reassures them they’ve found the real deal. Paid and organic aren’t rivals; they’re a team.

Your simple lookalike workflow, start to finish

Let me hand you the whole thing as a checklist you can start today:

• 1. Identify your best-intent group — ideally high-value purchasers or a clean customer list.
• 2. Build that group as a Custom Audience from your customer list, Pixel/website activity, or Meta engagement.
• 3. Confirm your data is consented and that you’re not in a restricted Special Ad Category.
• 4. Create a Lookalike Audience from that source, choosing your target country.
• 5. Pick a size to start — then build a second ad set with a different size so you can compare.
• 6. Also test a broad audience head-to-head, because Meta’s automation often loves it.
• 7. Exclude existing customers and recent visitors so you’re truly prospecting.
• 8. Run, then judge by cost per result — not clicks, not reach, the outcome that pays you.
• 9. Refresh the source as your buyer data grows, and rotate creative to fight fatigue.

That’s it — that’s genuinely how to create Facebook lookalike audiences from a blank Audiences screen to a tested, working part of your prospecting. It won’t be flawless on day one; nobody’s is. But you’ll have something real that gets smarter every time you feed it better data and read your own results honestly. You’ve got this. Start with a great source, stay curious, and let the numbers lead.

Frequently asked questions

What is the best source for a Facebook lookalike audience?

The best source is the one closest to real money and real intent — usually your high-value purchasers or a clean customer list of actual buyers. Meta builds the lookalike by mirroring whoever is in your source, so a sharp, high-intent source produces a sharper lookalike. If your purchase data is still small, step down to add-to-carts, lead completions, or engaged followers, and upgrade the source as your buyer data grows.

What lookalike audience size should I choose?

There’s no universal best size. A smaller percentage like 1% gives the closest match to your source but reaches fewer people, while a larger percentage reaches more people who resemble your source less closely. The right choice depends on your source quality, budget, and market size, so the honest approach is to test a couple of sizes as separate ad sets and let your own cost-per-result decide.

Are lookalike audiences still worth it, or should I go broad?

Both can work, and the smart move is to test them against each other. Meta’s delivery has become far more automated and often performs well with broad audiences and strong creative, sometimes matching or beating a lookalike. Lookalikes still shine when you have an excellent source, a niche market, or want a slightly more qualified prospecting layer, so run them head-to-head in your own account rather than assuming one always wins.

Can I create lookalikes for any type of ad?

No — if your ads fall under a Special Ad Category like housing, employment, or credit (and in some regions social, electoral, or political issues), Meta restricts standard lookalike targeting to prevent discrimination. In those cases Meta offers a differently-built audience option in place of regular lookalikes, and the rules change over time. Always check Meta’s current policies before advertising in these areas, and follow them exactly.

Does SocialBlaze create or manage Facebook lookalike audiences?

No. SocialBlaze is an organic social media tool — it schedules, auto-publishes, and analyzes your posts and unifies your inbox across networks, and it does not build, run, bid on, or manage Facebook ads or lookalike audiences. It complements your paid campaigns by keeping your organic presence active and credible, so the new people your lookalike ads reach find a brand that looks alive and trustworthy when they check you out.