What is Lookalike Audience?
An ad platform audience built from your existing customers. Meta, TikTok, and Google find people who share characteristics with your best customers but have not discovered you yet.
How lookalike audiences work
You upload a source audience (your customer list, website visitors, or engagement audience) to the ad platform. The platform analyzes shared characteristics (demographics, interests, behaviors) and finds new people who resemble your source audience but have not interacted with your brand. On Meta, you choose a percentage (1-10%) that determines how closely the lookalike matches your source. 1% is the closest match (smallest, highest quality), 10% is the broadest (largest, lower quality).
Source audience quality matters most
The quality of your lookalike depends entirely on the quality of your source. Best sources: top 25% customers by LTV (highest quality), all purchasers (good quality), add-to-cart visitors (decent quality). Weak sources: all website visitors (too broad), email subscribers who never purchased (unclear intent). Use your best customers, not your largest list, as the source. A 1% lookalike from 500 top customers outperforms a 1% lookalike from 50,000 website visitors.
Lookalike testing strategy
Test three lookalike percentages: 1% (precision), 3% (balance), 5% (reach). Run the same creative across all three ad sets at equal budget for 7 days. The 1% will typically have higher conversion rate but higher CPM. The 5% will have lower CPM but lower conversion rate. Find your sweet spot where CPA is lowest. For most DTC brands spending $5K-20K/month, the 2-3% lookalike is the sweet spot. Scale to 5% when you need more volume.