Audience Targeting and Retargeting

Core audiences are built using Meta's own data about its users: demographics such as age, gender, and location; interests inferred from page likes and content interactions; and behaviors like purchase history and device usage. Core audiences are the starting point for reaching cold prospects who do not yet know your brand.

Uploading a CSV file of existing customer email addresses or phone numbers to Meta creates a custom audience matched against Meta profiles. This allows you to run win-back campaigns for lapsed customers, cross-sell campaigns to specific segments, or exclude existing customers from prospecting campaigns.

The Meta Pixel tracks visitor behavior on your website and lets you build audiences based on specific pages visited, time spent on site, or actions taken. A website custom audience of people who viewed your pricing page but did not purchase is a high-intent retargeting segment worth prioritizing.

Meta lets you build audiences from people who have interacted with your content on-platform: video viewers, people who engaged with your Instagram or Facebook page, lead form openers, and event responders. These warm audiences are often cheaper to convert than cold audiences because they already have some brand familiarity.

People who watched a significant portion of your video — commonly 25%, 50%, or 75% completion thresholds — represent warm prospects with demonstrated interest. Building a retargeting audience from 75% video viewers and following up with a direct offer often produces strong conversion rates at lower cost-per-result.

Lookalike audiences instruct Meta to find new users who share characteristics with a source audience you define — typically your best customers or highest-value website visitors. A 1% lookalike is the most similar to your source, while a 10% lookalike is broader. Start with 1 to 2% for precision prospecting.

When building website custom audiences for retargeting, you choose how recently the user visited: 7 days captures high-intent recent visitors, 14 days casts a wider net, and 30 days captures a larger pool at the cost of lower recency. Match the window to your product's typical consideration period.

Stacking involves adding interest or behavior layers on top of a custom or lookalike audience to narrow it further. For example, targeting lookalike customers who also match the interest "entrepreneurship" filters your cold audience to those most likely to respond to your specific offer.

Exclusion audiences tell Meta who NOT to show your ads to. Excluding recent purchasers from a prospecting campaign prevents wasted spend; excluding your full customer list from a first-purchase offer maintains profitability. Well-configured exclusions are often the highest-return optimization in a Meta account.

A highly specific audience delivers better message relevance but limits Meta's ability to find enough users to spend your budget efficiently. A very broad audience gives Meta more optimization data but dilutes relevance. The sweet spot for most campaigns is an audience of 500,000 to three million users.

When multiple ad sets in the same campaign target overlapping audiences, they compete against each other in auction, driving up your own costs. Use Meta's Audience Overlap tool to check how much your key audiences share before launching. Heavy overlap warrants either merging ad sets or applying exclusions between them.

Test your understanding of custom audience types.

Effective Meta targeting combines core audiences for cold prospecting, custom audiences from customer lists, Pixel-driven website data, engagement-based warm audiences, lookalikes for scalable expansion, and strategic retargeting windows. Stacking audiences adds precision, exclusions prevent waste, the Audience Overlap tool eliminates internal competition, and balancing size against specificity ensures the algorithm can optimize efficiently.