Lookalike Audiences
A targeting method where ad platforms find new users who share behavioral and demographic characteristics with your existing customers or high-value prospects.
Lookalikes Let You Scale Beyond Your Known Audience
Every SaaS company hits a ceiling with direct targeting. You have retargeted your website visitors, emailed your list, and exhausted your LinkedIn audience of “VP Marketing at 50-200 employee SaaS companies.” Lookalike audiences break through that ceiling by finding people who look like your best customers but have never heard of you.
The quality depends entirely on your seed audience. Feed the algorithm your top 100 customers by LTV, and it finds people with similar job functions, company profiles, and online behaviors. Feed it every lead who ever downloaded an ebook, and you get a lookalike of tire-kickers.
How Lookalikes Work Under the Hood
The platform analyzes your seed audience across hundreds of signals — demographics, job titles, company size, browsing behavior, content engagement, purchase patterns. It then scores the broader platform population against those signals and returns the closest matches.
| Platform | Min Seed Size | Lookalike Options | B2B Effectiveness |
|---|---|---|---|
| Meta | 1,000 | 1-10% similarity | Good for SMB SaaS |
| 300 | Similar Audiences (fixed) | Strong for mid-market/enterprise | |
| 1,000 | Similar Segments | Moderate for B2B |
Getting the Most Out of Lookalikes
Layer demographic or firmographic filters on top of your lookalike. A 1% lookalike of your customers on Meta still includes consumers, students, and people at 5-person companies. Add filters for company size, seniority, or industry to clean up the audience. Test multiple seed lists — your highest LTV customers might produce a different lookalike than your fastest-closing customers. Both could work, but they target different buyer profiles. Refresh your seed list quarterly as your customer base evolves.
Frequently Asked Questions
How do you create a lookalike audience for B2B SaaS?
Upload a source list of your best customers — not all customers, just the highest LTV, fastest-closing, or highest NPS accounts. You need at least 300 records for LinkedIn and 1,000 for Meta. The platform then finds users with similar attributes. For B2B, customer email lists and website visitor lists work best as seed audiences.
What lookalike percentage should you use?
Start with 1% (most similar to your seed list) and test up to 3%. Beyond 3%, the audience becomes too broad and loses the precision that makes lookalikes work. On LinkedIn, use 'similar audiences' based on your matched segments or website visitors. On Meta, 1% lookalikes in the US give you roughly 2.5 million people — narrow further with interest or behavior overlays.
