Meta Ads Targeting

What you’ll learn

• How Meta Ads targeting works with Advantage+ AI and manual options.
• Detailed manual targeting options for Facebook and Instagram ads.
• Best practices for setting up targeting to maximize conversions.
• Why Advantage+ is recommended, with an advanced strategy for manual testing.

Introduction

Google Search Ads get triggered based on an action someone takes.

Namely, searching for a particular search term, which signals intent.

Meta Ads get triggered based purely on what Meta knows about you and what you're likely to be interested in and considering buying. This is based on your behavior both on Facebook/Instagram and on sites that have the Meta ad pixel installed.

You can either target manually or by getting Meta to 100% handle it for you.

This lesson explores both approaches, ultimately recommending Advantage+ for startups with limited budgets (< $10K/month), while offering an advanced manual strategy to test creatives. We’ll focus on targeting for an ecomm startup selling fitness gear, ensuring you can apply these principles to your business.

A quick example:

Based on the content someone views or interacts with both on and off Facebook, the algorithm might determine that they’re interested in environmentalism, dogs, and silly humor. So Facebook shows them dog videos, animal comics, and ads for reforestation nonprofits.

How Meta ads targeting works

As mentioned, two main approaches exist, and oddly they're also kind of mixed:

• Manual targeting: You define specific audiences using detailed filters (e.g., demographics, interests, behaviors), offering control for testing hypotheses or refining underperforming campaigns. Manual targeting is less efficient overall but valuable for advanced creative testing.
• Advantage+ AI targeting: Meta’s Advantage+ system uses machine learning to analyze billions of data points, automatically selecting audiences based on your campaign objective (e.g., Sales). Confusingly, when you do Advantage+, you can also choose specific Demographics, Audiences, Interests, Behaviors, etc, but just do the Advantage+ option, which will be more liberal and experimental with its targeting outside that core set when it feels that it will perform better to do so.

For startups, Advantage+ targeting is the default recommendation due to its efficiency and scalability, but learning manual options ensures you understand the full toolkit—and it's useful for advanced strategies.

Manual targeting options

Manual targeting lets you refine down to the specific audience characteristics you think is most likely to convert. The success of this heavily depends on whether you're right about those specific chatacteristics.

Here are the key options to explore, using a fitness gear startup as an example:

1. Demographics

• Age: Set a range (e.g., 18–45 for fitness gear targeting active adults).
• Gender: Include “All,” “Men,” or “Women” based on your product (e.g., unisex fitness gear).
• Location: Target “People living in” specific regions (e.g., U.S., California) or exclude areas (e.g., rural zones with low demand).
• Language: Match your audience’s language (e.g., English for U.S.-based users).
• Best practice: Exclude the aps
• that cannot or will not buy your product. But leave the rest for Meta to determine which is best.

2. Interests

• Categories: Here, you target users interested in related topics (e.g., “Fitness,” “Health & Wellness,” “Gym Equipment” for fitness gear). These are predefined by Meta.
• Detailed targeting: Add specific interests (e.g., “Yoga,” “Running,” “Home Workouts”) to reach niche segments.
• Comparison to Google: Interests are similar to Google’s affinity audiences (e.g., “Health & Wellness Enthusiasts”), but Meta’s data comes from user interactions on Facebook/Instagram (e.g., liking fitness pages) rather than search or browsing behavior.
• Best practice: Use 3–5 interests initially, avoiding overlap (e.g., “Fitness” and “Gym” might duplicate), and refine based on engagement after 14 days.

3. Behaviors

These are pre-defined audiences

• Overview: Behaviors target users based on actions they’ve taken on Meta’s platforms, similar to Google’s predefined audiences (e.g., in-market or affinity), but focused on social engagement rather than search intent.
• Categories:
  • Purchase behavior: Target users with habits like “Frequent Shoppers” or “E-commerce Shoppers” (akin to Google’s in-market “Online Shoppers”).
  • Engagement behavior: Reach users who engage with content (e.g., “Video Views: 25%+ Completed,” similar to Google’s YouTube engagement audiences).
  • Device usage: Target based on device (e.g., “Mobile Device Users”).
• Comparison to Google: While Google’s in-market audiences (e.g., “Fitness Products”) focus on search intent (e.g., users researching fitness gear), Meta’s Behaviors emphasize actions on-platform (e.g., buying fitness products via Marketplace, watching fitness videos).
• Best practice: Combine with interests (e.g., “Frequent Shoppers” + “Fitness”) to boost intent, monitoring CPA after 30 days.

4. Custom audiences

‍

• Website visitors: Upload a pixel-tracked list of users who visited your site (e.g., fitness gear product pages) for remarketing.
• Customer list: Upload a CSV of existing customers (e.g., email addresses) to target or exclude.
• Engagement: Target users who interacted with your page, video, or lead form.
• Best practice: Start with a 30-day website visitor list (minimum 100 users) and test a 1% lookalike audience (see below) for scale.

5. Lookalike audiences

• Overview: Here, you get Meta to create a list of people who "look like" people in a source audience (like your top 500 customers)
• Similarity: For lookalikes you choose a percentile of similarity between 1% and 10%. 1% means it's the top 1% of people who are similar to your audience (i.e. the most similar).
• Best practice: Use a high-value customer list (e.g., $100+ LTV buyers) for a 1% lookalike, expanding to 3% after 30 days of positive ROI.

6. Exclusions

• Exclude custom audiences (e.g., recent converters), locations (e.g., low-performing states), or interests (e.g., “Women's clothes” when you sell men's clothes) to refine reach.
• Best practice: Exclude converters after purchase to avoid overspending, adding exclusions after 14 days of data.

Tip: Again, we generally recommend automatic targeting, but for advanced creative testing strategies we will use some manual targeting.

Advantage+ targeting

Advantage+ AI automates targeting, replacing manual efforts with machine learning for efficiency:

• How it works: Upload your ad creatives and set a goal (e.g., Sales). If you have a product that is not super broad, it might be a good idea to use an Advantage+ audience.
• Automatic: It uses Meta’s data (interests, behaviors, lookalikes) without specific input, optimizing dynamically. It will also go outside your specified audiences if it feels it can achieve better performance.
• Placements: Optimizes across Feed, Stories, Reels, Explore, and Marketplace, selecting the best options.
• Learning phase: Requires 7–14 days or 50+ conversions for stabilization—ensure a budget to support this (which depends on the acceptable CPA)
• Best for: Everyone as it does yield the best results.

Recommendation: Use Advantage+ as your primary strategy for its automation and cost efficiency. Let Meta’s AI handle targeting, focusing on high-quality creatives (e.g., 3–5 images, 1–2 videos).

Next we'll quickly introduce a more advanced campaign strategy that uses both Advantage+ and Manual.

Advanced strategy: Manual targeting campaign for creative testing

Advantage+ can sometimes be overly opinionated with its creative and not give them enough opportunity to be adequately proven.

For startups with some experience or data (e.g., 50+ conversions), an advanced manual targeting campaign can test creatives effectively before scaling with Advantage+:

• Setup:
  • Create a separate ad set with a Sales objective.
  • Use manual targeting (e.g., “Fitness” interest, 18–45 age, U.S. location, “Frequent Shoppers” behavior, website visitors).
  • Upload 3–5 ad variations (e.g., different video hooks, image styles) within one ad set to compare performance.
• Execution:
  • Allocate at least ~$1,000/month per ad set, monitoring for 14–30 days.
  • Use Lowest Cost bidding to test which creative drives the lowest CPA.
  • Analyze the Breakdowns > By Delivery report to identify top-performing creatives.
• Outcome:
  • Pause underperforming ads and cycle in new creatives to be tested.
  • If you find winners, pause them as well and move them over to the Advantage+ campaign to be scaled.
  • Use insights (e.g., video outperforming images) to refine future creatives.
• Best for: Startups with enough budget and data to isolate creative impact, complementing Advantage+ campaigns for long-term optimization.

Tip: Reserve manual targeting for creative testing after 30 days of Advantage+ data. Keep it simple (1–2 ad sets) to avoid budget dilution, transitioning to Advantage+ for scale.

Quick tips

• Start with Advantage+: Launch an Advantage+ Sales campaign first. Let Meta’s AI optimize targeting across Facebook and Instagram for conversions.
• Learn manual targeting: Use demographics, interests, behaviors, and custom/lookalikes to point Advantage+ in the right direction given you're starting from scratch, but rely on Advantage+ for efficiency unless testing creatives.
• Test creatives manually (advanced): After 30+ days of Advantage+ data, run a manual campaign with a smaller budget to test 3–5 ad variations, scaling winners to Advantage+.
• Monitor learning phase: Allow 14–30 days (50+ conversions) for Advantage+ to stabilize, ensuring a healthy budget to avoid skewed results.
• Combine with Google Ads: Use Meta Ads for demand creation (awareness, consideration), complementing Google Search/Shopping Ads for capturing high-intent users.