A lookalike audience is a group of new people an ad platform finds because they resemble a list you give it, such as your best customers. You supply the source list, the platform's models look for people with similar traits, and your ads reach them as a new cold audience. Meta offers Lookalike Audiences, Google Ads offers Lookalike segments in Demand Gen campaigns, and LinkedIn replaced its lookalikes with predictive audiences in 2024. For B2B, the source list matters most: build it from customers you want more of, not from every lead.
If you run a B2B company, you probably have a list of customers you would happily clone: the right size, the right industry, good margins, easy to work with. A lookalike audience is the closest thing advertising has to that. You hand an ad platform a list of people who already matter to you, and it finds new people who resemble them.
This guide explains how lookalike audiences work, how each major platform handles them today (two of the three have changed recently), and where they fit in the four audience layers from our manufacturing marketing strategy guide.
What is a lookalike audience?
A lookalike audience is a new audience an ad platform builds by finding people who share traits with a source audience you provide. The source is usually a customer list, website visitors, or people who converted. The platform studies what those people have in common and looks for others who match. You never see who is in the lookalike; you choose the source, and often how close a match you want.
In audience-layer terms, a lookalike turns your warm first-party data into a new cold audience. The people in it have never heard of you, but they were chosen because they resemble people who already buy from you.
How the major platforms handle lookalikes today
Swipe sideways to see the full table.
| Platform | Feature today | How it works | Worth knowing |
|---|---|---|---|
| Meta (Facebook and Instagram) | Lookalike Audiences | Built from a source audience such as a customer list, site visitors, or page engagers; you pick a percentage range | Smaller percentages match the source more closely; larger ones are broader |
| Google Ads | Lookalike segments in Demand Gen campaigns | Built from a seed list, with narrow, balanced, or broad reach | Through 2026 Google is turning lookalikes into a signal: the system may reach people outside the similarity thresholds to hit your goal |
| Predictive audiences (lookalikes were discontinued on February 29, 2024) | LinkedIn's AI uses your first-party or third-party data to build an audience predicted to perform | Conversion, Lead Gen Form, or retargeting sources need more than 300 member accounts |
How to choose a source list for B2B
The source list decides everything. A lookalike built from the wrong people finds more of the wrong people, efficiently. For B2B and manufacturing:
- Start from customers you want more of. Your best accounts by fit and value beat a list of every lead or newsletter subscriber.
- Use people who did the thing you want. Contacts who requested a quote, became opportunities, or bought are stronger seeds than people who downloaded a guide.
- Keep it clean and current. Remove students, vendors, competitors, and old contacts. In our view, a smaller, accurate list beats a large, mixed one.
- Respect permission. Upload only data you are allowed to use for advertising, and follow each platform's rules for customer lists. Our guide to first-party data covers how to organize it.
Where lookalikes fit in a manufacturing campaign
| Audience layer | Role of lookalikes |
|---|---|
| Warm first-party data | The source: customers, past quotes, and CRM opportunities |
| Cold audiences | The output: new people who resemble your buyers, alongside LinkedIn targeting by role, industry, and company size |
| Retargeting | The next step: people from the lookalike who visit your site join your retargeting audiences. See our B2B retargeting guide |
| Third-party data | A check: intent data shows which companies are in market, so you can focus on accounts that fit. See our B2B intent data guide |
For your most valuable target accounts, we prefer a named list and an account-based approach to a lookalike. Our guide to account-based marketing for manufacturers covers that side.
How to test a lookalike audience
- Pick one source list built from your best customers or opportunities, and one goal such as quote requests.
- Launch the lookalike as its own ad set or campaign, so its results are not mixed with other audiences.
- Send the traffic to a page that matches the ad, with a clear next step.
- Exclude existing customers and open opportunities, so you pay only to reach new people.
- Judge it over a full sales cycle by qualified leads, opportunities, and cost per qualified lead, not by clicks.
Common mistakes
- Building the source from every lead instead of your best customers.
- Expecting a consumer platform's lookalike to understand company size or industry.
- Running an old LinkedIn lookalike strategy that no longer exists instead of predictive audiences or Matched Audiences.
- Forgetting to exclude current customers.
- Judging a lookalike in a week when your sales cycle takes months.
Next step
If you want help building lookalike and predictive audiences from your own customer data, our media team plans and runs them across Meta, Google, and LinkedIn, and our industrial marketing team fits them into a full campaign for manufacturers.
Sources
- About Lookalike Audiences, Meta Business Help Center
- Use Lookalike segments to grow your audience, Google Ads Help
- LinkedIn lookalike audiences have been discontinued, LinkedIn Marketing Solutions Help
- LinkedIn predictive audiences, LinkedIn Marketing Solutions Help
- Company list targeting in Campaign Manager, LinkedIn Marketing Solutions Help
- Contact list targeting in Campaign Manager, LinkedIn Marketing Solutions Help
- Targeting options for LinkedIn Ads, LinkedIn Marketing Solutions Help
Turn your data into audiences that convert
Tell us what you sell and what data you already have. We plan how your CRM, website visitors, retargeting, and outside data fit into one campaign.
Get a campaign plan