What Is a Lookalike Audience on Meta? DTC Guide

A lookalike audience on Meta is an algorithmically generated audience of people who share similar behavioral and demographic characteristics with a source audience you provide, such as your customer list or website visitors, used by DTC brands to find new potential buyers who resemble their best existing customers.

Last updated: February 2026

Table of Contents

How Lookalike Audiences Work on Meta

When you upload a customer list (email addresses, phone numbers) or connect a pixel-based custom audience to Meta, Meta compares the people in that source audience against its broader user base. It analyzes hundreds of behavioral signals: what pages people like, what content they engage with, what purchases they've made, what demographics they share.

Meta then finds people in the broader Facebook and Instagram user population who match those patterns most closely. These are your lookalike audience.

The "percentage" you choose determines how similar the lookalike is to your source. A 1% lookalike is the top 1% of people most similar to your source audience in your target country. A 10% lookalike includes the top 10%, which is a much larger but less precisely similar audience.

The business logic: If your best customers share certain characteristics, finding more people with those characteristics is the fastest path to new customer acquisition at scale.

Source Audiences for DTC Lookalikes

The quality of a lookalike audience is entirely dependent on the quality of the source audience. Better sources produce better lookalikes.

Highest quality sources for DTC:
    • High-LTV customers (best): Upload your top 20% of customers by lifetime value. A lookalike based on your best customers produces better acquisition outcomes than a lookalike based on all customers.
    • Recent purchasers: Customers who bought in the last 90 to 180 days. Recent purchase behavior is a strong signal of current buyer intent.
    • Subscription customers: If you have a subscription product, a lookalike based on active subscribers finds audiences most likely to become recurring buyers.
    • Pixel purchase events: A lookalike based on your website purchase pixel events. Requires at least 100 purchaser events for Meta to build a quality lookalike (500+ is better).
Lower quality sources:

Lookalike Percentage: 1% vs 5% vs 10%

1% Lookalike: Smallest audience. Most similar to your source. Highest accuracy, lowest reach. Best for: High-budget precision targeting when you want the most similar audience possible. 2% to 3% Lookalike: Good balance of accuracy and scale. Most DTC brands get the best efficiency here. Best for: Mid-funnel prospecting where you want strong similarity but enough audience for efficient delivery. 5% to 10% Lookalike: Larger audience. Less similar to source. More reach but lower conversion rates. Best for: Building brand awareness at scale or when 1% audiences are exhausted. The practical recommendation: Test 1% and 3% to 5% simultaneously for new DTC campaigns. Most brands find the 1% converts better per impression but exhausts faster. The 3% to 5% is more sustainable for ongoing prospecting at scale.

With Advantage+ audience becoming Meta's preferred approach, standalone lookalike targeting is being somewhat supplanted, but customer list-based LALs still perform well as audience seeds within Advantage+ campaigns.

Best Practices for DTC Lookalike Audiences

Use multiple source audiences: Create separate lookalikes from your high-LTV customers, recent purchasers, and pixel purchasers. Test all three. The best-performing source often varies by product and brand. Keep source lists updated: Your customer base changes over time. Update your custom audience uploads monthly to keep lookalike data current. Minimum source audience size: Meta needs at least 100 people in your source audience to build a lookalike. For best results, use 1,000+ people. Below 100, Meta can't identify reliable similarity patterns. Stack percentage variations: Run 1%, 3%, and 5% lookalikes in separate ad sets within a CBO campaign. Meta will allocate budget toward the most efficient performer. Exclude existing customers: Add your existing customers as an exclusion to your lookalike campaigns. You don't need to spend acquisition budget on people who already bought from you.

Lookalike Audiences After iOS 14

iOS 14 affected lookalike quality in two ways:

    • Smaller pixel-based source audiences: Fewer purchase events are captured due to iOS tracking restrictions. A source audience that previously had 1,000 purchasers may now have 650, reducing lookalike quality.
    • Less behavioral signal for matching: Meta has less data on iOS users' off-platform behavior (since they opted out of tracking), which reduces the accuracy of lookalike matching for iOS users.
What still works well: Customer list-based lookalikes (email/phone uploads) are largely unaffected by iOS restrictions because you're providing the matching data directly. Email-based customer lists remain the highest-quality lookalike source.

Pixel-based lookalikes have been most affected. If you rely on pixel events, implement CAPI to recover more purchase events for your source audiences.

Lookalike vs Advantage+ Audience

Meta's Advantage+ audience gives the algorithm maximum flexibility to find the best-performing users, essentially a very broad targeting approach with Meta's algorithm determining who to show ads to.

When to use each: Lookalike audiences: Best when you have a high-quality source audience and want to constrain targeting to similar users. Better for niche products or highly specific customer profiles. Advantage+ audience: Best when Meta's algorithm has significant purchase history data to work from (established accounts with 500+ monthly conversions). Consistently outperforms manual targeting for accounts with strong pixel history. The trend: Most experienced DTC advertisers have shifted toward Advantage+ audience for prospecting and reserved lookalike targeting for specific use cases (high-LTV lookalikes as Advantage+ audience "suggestions").

How to Create a Lookalike Audience for DTC

Step 1: Create your source custom audience Facebook Ads Manager > Audiences > Create Audience > Custom Audience

For customer list: Upload a CSV with email, phone, name columns. Meta hashes the data for privacy matching.

For pixel-based: Select "Website" as source, choose "Purchase" event, set lookback window (180 days recommended).

Step 2: Create the Lookalike Audiences > Create Audience > Lookalike Audience

Select your custom audience as the source. Choose your target country. Select the lookalike percentage (1%, 3%, 5%).

Step 3: Wait for audience creation Meta takes 6 to 24 hours to build a lookalike audience. It will show as "Populating" until ready. Step 4: Use in your campaign Add the lookalike audience to your ad set targeting. Combine with age/gender restrictions if relevant. Add existing customer exclusions.

FAQ

How many customers do I need for a good lookalike audience? Minimum 100 for Meta to create a lookalike. For reliable quality, aim for 1,000+. For high-LTV-based lookalikes, even 300 to 500 high-value customers creates a good quality lookalike because the quality of the source matters more than quantity. Does lookalike audience performance decline over time? Yes. As Meta's algorithm serves your lookalike ads, it reaches the people most likely to convert first. Over weeks to months, the remaining audience in the lookalike pool is increasingly less similar to your best source customers. Refreshing your source audiences monthly helps maintain quality. Should I create lookalikes based on purchasers or newsletter subscribers? Purchasers every time. Subscribers have lower purchase intent than buyers. The best lookalike source is always people who have demonstrated purchase behavior, not just email interest. What's the minimum budget to use lookalike audiences effectively? Lookalike audiences need enough budget to exit Meta's learning phase. As a rule of thumb, your daily budget should be enough to generate 5 to 10 purchases per day. Below $30 to $50/day, lookalike campaigns often underperform because delivery can't optimize effectively.