A lookalike audience is a group of new people or companies chosen because they share traits with a seed set you already know, usually your best customers.
You start with a seed: a list of accounts that closed, stayed and paid well. You pull out what they have in common, such as industry, headcount, country and tools used. Then you search for other companies and people that match the same traits.
Why it matters for outbound
Outbound works best on accounts that look like the ones you already win. A lookalike list gives you a target set grounded in real closed deals, not a guess at your ideal customer.
The method also scales: a small seed can produce a long list of similar accounts. The risk is a thin seed. Three customers in three industries give you a noisy pattern, so start with ten or more that share the traits you care about.
Be clear about what "similar" means. In ad platforms it is a model score you cannot see. In outbound data it is usually a set of filters you can read, test and change.
Example
Say your seed is ten customers. All ten are software companies with 51 to 200 employees, headquartered in the United States.
You turn that pattern into filters and search for companies that match, excluding the ten you already have. Then you find the right buyer at each one.
| Step | Value |
|---|---|
| Seed | Acme and nine other customers |
| Shared traits | Software, 51 to 200 employees, US headquarters |
| Search | Companies matching those three traits, minus the seed |
| Next | Heads of Sales at each match, with a work email |
The names and values are placeholders. Your own seed decides the filters.
In LeadOcean data
LeadOcean has no lookalike model and no one-click lookalike button. You build the audience from filters, which keeps it transparent.
Read your seed with POST /v1/companies/enrich (or the get_company tool) to see industry, employee range and headquarters country. Then run /v1/companies/search with the matching industry, employeeRange and hqCountry filters, and use excludeName to drop the seed. To find people at those matches, call GET /v1/companies/{domain}/people, or size them first with count_leads, which is free. Enum values are exact strings, so check list_enum_values before you filter.
Each company or person returned counts as one record. The free plan gives 1,000 records, one-off, with no card, and Pro is $499 a month. See /pricing.
The lookalike-companies Claude skill runs these steps for you from one seed domain: Lookalike Companies. To size the result before you spend records, use Audience Sizer. For the method behind it, read what is lookalike prospecting.
Related terms
Build your first lookalike list with a free API key
Free to start. No credit card. 1,000 records to spend whenever you like.
Get your free API key →