January 20266 min read

Audience Targeting

Master Custom Audiences, Lookalikes, and Advantage+ targeting for maximum reach and relevance.

Targeting Overview

Meta offers four main audience types:

| Audience Type | Description | Best For | |---------------|-------------|----------| | Core Audiences | Demographics, interests, behaviors | Cold prospecting | | Custom Audiences | Your existing data | Retargeting, exclusions | | Lookalike Audiences | Similar to your customers | Scaled acquisition | | Advantage+ Audience | AI-expanded targeting | Automated optimization |

The 2025 Targeting Shift

Key Insight: In 2025, broad targeting often outperforms detailed interest targeting. Meta's algorithm has become sophisticated enough to find your ideal customers without manual segmentation.

| Broad Targeting | Lookalike Targeting | |-----------------|---------------------| | Higher reach | More precise | | Lower CPM (+45% cheaper) | Higher CPM | | 113% avg ROAS | 76% avg ROAS | | Better for established products | Better for niche markets |

Core Audiences

Core audiences use Meta's data on user demographics, interests, and behaviors.

Demographics

code

Location → Country, State, City, Radius, Zip Codes
Age → 18-65+ (or specific ranges)
Gender → All, Men, Women
Languages → Primary languages

Detailed Targeting

Interests: Based on pages liked, content engaged with

Business & Industry
Entertainment
Family & Relationships
Fitness & Wellness
Food & Drink
Hobbies & Activities
Shopping & Fashion
Sports & Outdoors
Technology

Behaviors: Based on purchase history, device usage

Purchase Behavior
Travel
Mobile Device User
Digital Activities
Seasonal & Events

Demographics: Based on profile information

Education
Financial
Life Events
Parents
Relationship
Work

Targeting Expansion Options

| Option | Description | |--------|-------------| | Advantage Detailed Targeting | Expands beyond selected interests | | Advantage Lookalike | Expands beyond Lookalike percentage | | Advantage Custom Audience | Finds similar users to custom audiences |

Custom Audiences

Custom Audiences are built from your first-party data—the highest-value targeting available.

Website Custom Audiences

Based on Meta Pixel/Conversions API data:

code

All Website Visitors (last 180 days)
├── Visitors by Time Spent (top 25%)
├── Visitors by Page (specific URLs)
├── Visitors by Event (Add to Cart, Purchase)
└── Visitors by Frequency (visited 2+ times)

Best Practices:

Create audiences at different time windows (7, 30, 90, 180 days)
Segment by action (viewers vs. add-to-cart vs. purchasers)
Exclude purchasers from acquisition campaigns

Customer List Audiences

Upload your CRM data:

code

Accepted Fields:
- Email (required or phone)
- Phone Number
- First Name, Last Name
- City, State, Country, Zip
- Date of Birth
- Gender
- Mobile Advertiser ID

Match Rates:

Email only: 40-60% match rate
Email + Phone: 60-75% match rate
All fields: 70-85% match rate

Privacy Note: Meta hashes all data before matching. Original data is not stored.

Engagement Custom Audiences

Based on interactions with your Meta content:

| Source | Engagement Types | |--------|-----------------| | Video | Viewers (3s, 10s, 25%, 50%, 75%, 95%) | | Lead Form | Opened, Submitted | | Instant Experience | Opened, Clicked | | Shopping | Viewed products, Added to cart | | Instagram | Profile visits, Saved posts, Engaged | | Facebook Page | Engaged, Messaged, Clicked CTA | | Events | Responded, Purchased tickets |

App Activity Audiences

For mobile app advertisers:

App installers
Active users (by frequency)
In-app purchasers
Specific event completers
Lapsed users

Lookalike Audiences

Lookalike Audiences find new users similar to your best customers.

How Lookalikes Work

code

Source Audience (your data)
        ↓
Meta analyzes patterns (demographics, interests, behaviors)
        ↓
Finds similar users in target country
        ↓
Returns Lookalike Audience (1-10% of population)

Lookalike Sizes

| Size | Description | Use Case | |------|-------------|----------| | 1% | Most similar (top 1%) | Best performance, limited reach | | 1-3% | Very similar | Balanced approach | | 3-5% | Similar | Scaling campaigns | | 5-10% | Broader similarity | Maximum reach |

US Market Example (330M population):

1% LAL = ~3.3M people
3% LAL = ~10M people
5% LAL = ~16.5M people
10% LAL = ~33M people

Best Source Audiences for Lookalikes

Ranked by Quality:

High-LTV Purchasers - Your best customers
All Purchasers - Proven converters
Add-to-Cart - High intent
Email Subscribers - Opted-in prospects
Website Visitors (Top 25% by time) - Engaged visitors
Video Viewers (75%+) - Interested prospects

Lookalike Best Practices

code

## Source Audience Requirements

□ Minimum 100 users (1,000-5,000 recommended)
□ Same country as target Lookalike
□ Quality over quantity (use best customers)
□ Update source monthly for freshness

## Testing Strategy

□ Start with 1% Lookalike
□ Test 1% vs 3% vs 5% in separate ad sets
□ Compare performance over 7-14 days
□ Scale winning Lookalike sizes

Advantage+ Audience

Advantage+ Audience is Meta's AI-powered targeting that goes beyond your defined parameters.

How It Works

code

Your Targeting Inputs (suggestions to AI)
        ↓
Meta AI finds additional high-performers
        ↓
Ads delivered to expanded audience
        ↓
Continuous optimization based on results

Advantage+ Audience Settings

| Setting | Description | |---------|-------------| | Audience Controls | Hard limits (age, location) the AI cannot exceed | | Audience Suggestions | Interests/behaviors AI can expand beyond | | Custom Audiences | Can be used as suggestions or controls |

When to Use Advantage+ Audience

Use Advantage+ When:

Running Sales or Leads campaigns
Have 50+ weekly conversions
Strong creative variety
Broad market appeal

Use Manual Targeting When:

Niche B2B products
Geographic restrictions required
Strict age requirements (alcohol, etc.)
Testing specific audience hypotheses

Targeting Strategy 2025

The New Best Practices

1. Start Broader, Refine Later

code

Week 1-2: Broad targeting + Advantage+ Audience
Week 3-4: Analyze which segments perform
Week 5+: Create Custom/Lookalike audiences from converters

2. Stack Audiences Strategically

code

Campaign: Prospecting
├── Ad Set 1: Broad (US, 25-54)
├── Ad Set 2: LAL 1% Purchasers
└── Ad Set 3: Interest Stack (exclude LAL)

Campaign: Retargeting
├── Ad Set 1: Add-to-Cart 7 days (exclude purchasers)
├── Ad Set 2: Website Visitors 30 days (exclude ATC)
└── Ad Set 3: Email Subscribers (exclude visitors)

3. Use Exclusions Effectively

Always exclude:

Recent purchasers (from acquisition campaigns)
Current customers (from new customer campaigns)
Existing leads (from lead gen campaigns)
Irrelevant demographics (if applicable)

4. Refresh Audiences Regularly

| Audience Type | Refresh Frequency | |---------------|-------------------| | Customer Lists | Monthly | | Lookalikes | Every 2-3 months | | Website Audiences | Auto-updating | | Engagement Audiences | Auto-updating |

2025 Insight: The biggest targeting mistake is over-segmentation. Meta's algorithm needs sufficient audience size (1M+) and conversion volume (50+/week) to optimize effectively.

January 20266 min read

Audience Targeting

Master Custom Audiences, Lookalikes, and Advantage+ targeting for maximum reach and relevance.

Targeting Overview

Meta offers four main audience types:

The 2025 Targeting Shift

Key Insight: In 2025, broad targeting often outperforms detailed interest targeting. Meta's algorithm has become sophisticated enough to find your ideal customers without manual segmentation.

Core Audiences

Core audiences use Meta's data on user demographics, interests, and behaviors.

Demographics

code

Location → Country, State, City, Radius, Zip Codes
Age → 18-65+ (or specific ranges)
Gender → All, Men, Women
Languages → Primary languages

Detailed Targeting

Interests: Based on pages liked, content engaged with

Business & Industry
Entertainment
Family & Relationships
Fitness & Wellness
Food & Drink
Hobbies & Activities
Shopping & Fashion
Sports & Outdoors
Technology

Behaviors: Based on purchase history, device usage

Purchase Behavior
Travel
Mobile Device User
Digital Activities
Seasonal & Events

Demographics: Based on profile information

Education
Financial
Life Events
Parents
Relationship
Work

Targeting Expansion Options

Custom Audiences

Custom Audiences are built from your first-party data—the highest-value targeting available.

Website Custom Audiences

Based on Meta Pixel/Conversions API data:

code

All Website Visitors (last 180 days)
├── Visitors by Time Spent (top 25%)
├── Visitors by Page (specific URLs)
├── Visitors by Event (Add to Cart, Purchase)
└── Visitors by Frequency (visited 2+ times)

Best Practices:

Create audiences at different time windows (7, 30, 90, 180 days)
Segment by action (viewers vs. add-to-cart vs. purchasers)
Exclude purchasers from acquisition campaigns

Customer List Audiences

Upload your CRM data:

code

Accepted Fields:
- Email (required or phone)
- Phone Number
- First Name, Last Name
- City, State, Country, Zip
- Date of Birth
- Gender
- Mobile Advertiser ID

Match Rates:

Email only: 40-60% match rate
Email + Phone: 60-75% match rate
All fields: 70-85% match rate

Privacy Note: Meta hashes all data before matching. Original data is not stored.

Engagement Custom Audiences

Based on interactions with your Meta content:

App Activity Audiences

For mobile app advertisers:

App installers
Active users (by frequency)
In-app purchasers
Specific event completers
Lapsed users

Lookalike Audiences

Lookalike Audiences find new users similar to your best customers.

How Lookalikes Work

code

Source Audience (your data)
        ↓
Meta analyzes patterns (demographics, interests, behaviors)
        ↓
Finds similar users in target country
        ↓
Returns Lookalike Audience (1-10% of population)

Lookalike Sizes

US Market Example (330M population):

1% LAL = ~3.3M people
3% LAL = ~10M people
5% LAL = ~16.5M people
10% LAL = ~33M people

Best Source Audiences for Lookalikes

Ranked by Quality:

High-LTV Purchasers - Your best customers
All Purchasers - Proven converters
Add-to-Cart - High intent
Email Subscribers - Opted-in prospects
Website Visitors (Top 25% by time) - Engaged visitors
Video Viewers (75%+) - Interested prospects

Lookalike Best Practices

code

## Source Audience Requirements

□ Minimum 100 users (1,000-5,000 recommended)
□ Same country as target Lookalike
□ Quality over quantity (use best customers)
□ Update source monthly for freshness

## Testing Strategy

□ Start with 1% Lookalike
□ Test 1% vs 3% vs 5% in separate ad sets
□ Compare performance over 7-14 days
□ Scale winning Lookalike sizes

Advantage+ Audience

Advantage+ Audience is Meta's AI-powered targeting that goes beyond your defined parameters.

How It Works

code

Your Targeting Inputs (suggestions to AI)
        ↓
Meta AI finds additional high-performers
        ↓
Ads delivered to expanded audience
        ↓
Continuous optimization based on results

Advantage+ Audience Settings

When to Use Advantage+ Audience

Use Advantage+ When:

Running Sales or Leads campaigns
Have 50+ weekly conversions
Strong creative variety
Broad market appeal

Use Manual Targeting When:

Niche B2B products
Geographic restrictions required
Strict age requirements (alcohol, etc.)
Testing specific audience hypotheses

Targeting Strategy 2025

The New Best Practices

1. Start Broader, Refine Later

code

Week 1-2: Broad targeting + Advantage+ Audience
Week 3-4: Analyze which segments perform
Week 5+: Create Custom/Lookalike audiences from converters

2. Stack Audiences Strategically

code

Campaign: Prospecting
├── Ad Set 1: Broad (US, 25-54)
├── Ad Set 2: LAL 1% Purchasers
└── Ad Set 3: Interest Stack (exclude LAL)

Campaign: Retargeting
├── Ad Set 1: Add-to-Cart 7 days (exclude purchasers)
├── Ad Set 2: Website Visitors 30 days (exclude ATC)
└── Ad Set 3: Email Subscribers (exclude visitors)

3. Use Exclusions Effectively

Always exclude:

Recent purchasers (from acquisition campaigns)
Current customers (from new customer campaigns)
Existing leads (from lead gen campaigns)
Irrelevant demographics (if applicable)

4. Refresh Audiences Regularly

2025 Insight: The biggest targeting mistake is over-segmentation. Meta's algorithm needs sufficient audience size (1M+) and conversion volume (50+/week) to optimize effectively.