What is a lookalike audience and how do I use my customer list to find more buyers on Meta?

Interest-based targeting — homeowners, home improvement interest, age 28–55 — is a starting point. Lookalike audiences are the upgrade. Instead of telling Meta "find people who are interested in home improvement," you're telling it "find people who look exactly like these 400 customers who already paid us." Meta has behavioral and demographic data on over 3 billion users. The lookalike algorithm is its most powerful targeting tool, and local service businesses rarely use it.

How lookalike audiences work

You upload a CSV file with your customers' names, email addresses, and/or phone numbers. Meta hashes this data (anonymizes it cryptographically) and matches it against its user database. It finds the Facebook and Instagram accounts associated with those contacts — typically matching 40–70% of your list to real accounts. Then it analyzes the behavioral patterns, interests, purchase behaviors, and demographic traits shared by those matched accounts. Finally, it finds millions of other users who share those same patterns — these are your lookalike audience members.

The 1% lookalike setting means Meta will target the top 1% of users in your target country who most closely resemble your seed list. For the United States, 1% is approximately 2.1 million people. Narrow enough to be meaningful; broad enough to deliver ads.

Best input produces best output

This is the most important principle of lookalike audiences: the quality of your seed list determines the quality of your lookalike. If you upload your entire contact database — including tire-kickers, low-value jobs, one-time customers who complained — Meta finds more people like all of them. If you upload only your best customers — the ones who hired you for high-value jobs, came back for repeat work, or referred others — Meta finds more people like those customers.

Before exporting your customer list, filter it. Remove people who were difficult, low-value, or unlikely to refer. Upload the 100–300 customers you'd most like to have 1,000 more of. This is the single most impactful optimization you can make to a lookalike audience.

How to create a lookalike audience — step by step

1. Export your best customer list from your CRM, invoicing software, or contact database as a CSV. Include columns for: first name, last name, email address, phone number. More data points = better match rate.
2. Go to Meta Business Suite → Audiences → Create Audience → Custom Audience → Customer List.
3. Upload the CSV. Map your columns to Meta's fields (first name, last name, email, phone). Meta hashes all data client-side before it's sent.
4. Wait 24–48 hours for Meta to process the match. You'll see a match percentage — 40–70% is typical and acceptable.
5. Once the Custom Audience is ready, click "Create Lookalike" from that audience. Select 1% for your first lookalike.
6. Select your country and/or add a geographic restriction at the ad set level — critical for local businesses (see below).

Seed list size and lookalike quality

When to use 3–5% vs. 1% lookalikes

A 1% lookalike is the tightest statistical match — highest quality, smallest audience, usually lowest CPL. A 3–5% lookalike is a broader match — lower quality similarity, much larger audience, useful when you need more volume and your 1% audience is saturating. For most local service businesses targeting a specific metro area, 1% is the right starting point. You'll typically reach the audience ceiling before you need to expand to 3–5%.

Update your seed list every 90 days

Your customer base evolves, and so should your seed list. Customers who hired you 4 years ago may represent different demographics or use cases than customers from the last 12 months. Re-export and re-upload your filtered customer list every 90 days to keep your lookalike audience current with your best recent customers.