Mastering Facebook Lookalike Audiences for Lead Generation

What Are Lookalike Audiences?

Lookalike audiences let you reach new people who share characteristics with your existing customers. Meta's algorithm analyses your source audience and finds similar users across Facebook and Instagram.

Why Lookalikes Work

Instead of guessing who might buy from you, lookalikes use real customer data:

Purchase behaviour patterns

Engagement signals

Demographic similarities

Interest overlaps

The result? Higher-quality leads at lower costs than interest-based targeting.

Building Effective Source Audiences

Best Source Audiences

Quality of your source determines quality of your lookalike:

Purchasers - Best for e-commerce

High-value customers - Top 25% by spend

Qualified leads - Not just form fills, actual qualified prospects

Engaged users - Video viewers, page engagers

Minimum Size

**Minimum**: 100 people (1,000+ recommended)

**Optimal**: 1,000-50,000 people

Larger sources = more data for the algorithm

Avoid These Sources

All website visitors (too broad)

Email lists with poor quality

Old data (12+ months)

Mixed intent audiences

Lookalike Percentages Explained

1% Lookalike

Closest match to source

Smallest audience size

Highest quality, lowest scale

Best for: Initial testing, limited budgets

2-5% Lookalike

Good balance of quality and scale

Our recommended starting point

Best for: Most lead gen campaigns

6-10% Lookalike

Broader reach

Lower relevance

Best for: Brand awareness, retargeting pool building

Setup Best Practices

Layer Your Lookalikes

Don't rely on one lookalike. Create multiple:

1% purchasers

2% high-value customers

3% qualified leads

5% engaged video viewers

Test each and scale winners.

Exclude Existing Customers

Always exclude:

Current customers

Recent purchasers

Existing leads in pipeline

Refresh Regularly

Source audiences should update:

Monthly for active businesses

Quarterly minimum

After major campaigns

Scaling Lookalike Campaigns

Horizontal Scaling

Add new lookalike variations:

Different source audiences

Different percentages

Different countries

Vertical Scaling

Increase budget on winners:

20% budget increases

Monitor CPA closely

Don't scale losing campaigns

Common Mistakes

Wrong source audience - Using all visitors instead of buyers

Too small source - Under 1,000 people limits algorithm

Stacking lookalikes - Creates overlap, wastes budget

Set and forget - Need to refresh sources regularly

Ignoring creative - Great audience + poor ads = poor results

Our Approach

We typically start with:

2% lookalike of best customers

Strong creative testing

Clear conversion tracking

Weekly optimisation

Then expand to multiple lookalikes based on performance data.