What is the best seed audience for B2B lookalikes?

Your best seed audience is a list of closed-won customer email addresses. This teaches Facebook's algorithm what a buyer looks like based on the behavioral patterns of people who actually purchased your product. If your customer list is too small (under 1,000), use qualified opportunity contacts as an alternative seed. Avoid using all leads or all website visitors as seeds — these contain too much noise to produce effective lookalikes.

What lookalike percentage should I use for B2B?

Start with 1% lookalike for conversion campaigns (highest quality, smallest reach). Expand to 1-3% for lead generation campaigns once you validate performance. 3-5% is acceptable for awareness campaigns. Avoid 5-10% for B2B — the audience is too diluted. If your 1% audience is too small for your budget, add interest-based targeting layers rather than expanding the percentage.

How often should I update lookalike audiences?

Refresh your seed audiences quarterly by uploading updated customer and opportunity lists. Facebook automatically updates the lookalike model when you update the source custom audience. More frequent updates (monthly) are beneficial if your customer base is growing rapidly or if your ICP has shifted.

Can Facebook lookalikes replace LinkedIn targeting for B2B?

Lookalikes complement LinkedIn targeting but do not fully replace it. Facebook lookalikes excel at finding people similar to your customers at lower cost, but they lack the professional-data precision of LinkedIn. The best approach is using both: LinkedIn for precision decision-maker targeting and conversion campaigns, Facebook lookalikes for expanding reach, retargeting, and cost-effective lead generation.

Facebook Lookalike Audiences for B2B: Advanced Guide

Lookalike Audiences for B2B: Facebook's Secret Weapon

Facebook's lookalike audience modeling is arguably the most powerful B2B targeting capability on the platform. While Facebook's native B2B targeting (job titles, interests) is inferior to LinkedIn's, its lookalike algorithm is superior — it processes thousands of behavioral signals to find users who share characteristics with your seed audience, reaching potential buyers that LinkedIn's more rigid targeting criteria would miss.

For B2B, the key to effective lookalikes is seed audience quality. A lookalike built from your 500 best customers will dramatically outperform one built from all leads or all website visitors, because the algorithm mirrors the characteristics of whatever seed you provide.

Building B2B Lookalike Audiences

Choosing Your Seed Audience

The seed audience determines lookalike quality. Prioritize seeds by downstream value, not volume:

Best seed: Closed-won customers. Upload your customer list (email addresses). This teaches the algorithm what a buyer looks like.
Good seed: Qualified opportunities. Contacts from opportunities, including those that did not close. These represent people who fit your ICP and showed buying intent.
Adequate seed: Marketing-qualified leads. Contacts that met your qualification criteria but may not have progressed to opportunity. Higher volume but lower signal.
Avoid: All leads or all website visitors. These seeds contain too much noise — including unqualified contacts and irrelevant traffic — to produce useful lookalikes.

Lookalike Size Selection

Facebook offers lookalike sizes from 1% to 10% of a country's population. For B2B:

1% lookalike: Most similar to your seed. Highest quality, smallest reach. Start here for conversion campaigns.
1-3% lookalike: Good balance of quality and reach. Use for lead generation campaigns once you have validated the 1% performance.
3-5% lookalike: Broader reach with diluted quality. Use for awareness campaigns or when 1-3% audiences are too small for your budget.
5-10% lookalike: Generally too broad for B2B. The audience is so large that the B2B signal is diluted by consumer users.

Build Precision Lookalikes With AI

MetadataONE builds and optimizes lookalike audiences using your CRM data, automatically refreshing seeds and testing audience sizes.

Book a Demo

Layering Lookalikes With Targeting

Improve lookalike quality by adding targeting layers on top:

Interest layer: Add business-relevant interests (enterprise software, marketing technology, etc.) to filter out consumers in the lookalike audience.
Age layer: Exclude ages 18-24 to remove students and early-career users who are unlikely B2B buyers.
Exclusions: Exclude existing customers, existing leads, and competitors.

These layers narrow the lookalike audience, reducing reach but increasing relevance. Test layered vs pure lookalike campaigns to find the right balance for your program.

Optimizing Lookalike Campaigns

Refresh seeds quarterly. Update your customer and opportunity lists to keep the lookalike model current with your evolving customer base.
Test multiple seeds. Run separate campaigns with different seed audiences (customers vs opportunities vs MQLs) and compare cost per qualified lead.
Combine with conversion optimization. Set campaign optimization to a conversion event (lead form submission, demo request) rather than link clicks. This lets the algorithm further refine delivery within the lookalike to users most likely to convert.
Connect to CRM. Import offline conversion data so Facebook's algorithm learns which lookalike users become actual pipeline and revenue, not just form fills.

Expected Performance

B2B lookalike campaigns on Facebook typically achieve:

CTR: 0.8-1.5% (higher than interest-based targeting)
CPL: $25-$80 (30-50% lower than LinkedIn)
Lead quality: 60-80% of LinkedIn quality (measured by lead-to-opportunity conversion rate)
Cost per qualified lead: often comparable to LinkedIn despite lower per-lead quality, because volume is higher