Lookalike prospecting is finding new companies and contacts that resemble the customers you already win. You describe your best accounts by traits, then search for others with the same traits.
Key takeaways
- A lookalike list starts from your customers, not from a guess about who might buy. The seed is real wins.
- The match runs on shared traits: industry, size, location, tech stack and keywords. More matching traits means a closer lookalike.
- It differs from an ICP list. An ICP is a written profile. A lookalike list is built by measuring customers against it.
- LeadOcean has no "find similar" button. You build the lookalike with its company filters, then pull the people.
What it is
Lookalike prospecting builds a target list by taking the traits your best customers share and searching for other companies with the same traits.
The idea comes from ad platforms, where a "lookalike audience" is a group of people who resemble a seed list. In B2B the seed is a set of accounts, and the traits are company fields.
Two things make a good seed. It is made of customers who stayed and paid well, not everyone who signed. And it has at least ten or so accounts, so a pattern is visible and not an accident.
How it works
Lookalike prospecting turns a customer list into a search. You find what the winners share, then ask a database for more companies with those traits.
- Export your best customers. Pick the ones with high retention or revenue, not just recent signups.
- Tag each with company traits: industry, headcount bracket, HQ country, funding stage, detected tools.
- Count how often each trait repeats. A trait shared by most of the seed is a hard filter. A trait shared by a third is a soft one.
- Turn the hard traits into filters and run a company search.
- Remove accounts you already sell to. Then pull the right buyer at each remaining company.
Worked example, with placeholders:
Seed (12 customers):
Acme, Globex, Initech ...
Shared traits:
industry Software Development (12 of 12)
headcount 51-200 (10 of 12)
HQ country US (9 of 12)
tech category Marketing automation (7 of 12)
Filters used:
industry + employeeRange + hqCountry (hard)
technologyCategories (soft, tested separately)
Buyer to pull:
VP, Sales & Business DevelopmentThe soft trait is where lookalikes go wrong. If you add it as a hard filter, you cut away companies that would have bought. Test the list with and without it.
Check the result before you scale it. Take a sample of 50 companies from the new list and read them against your seed. If a third look wrong, a filter is too loose. If the list is tiny, a filter is too tight.
The seed also decays. A list built from customers who joined three years ago describes the market as it was. Rebuild the seed every quarter from recent wins, and watch for traits that start to drift.
Lookalike prospecting vs ICP targeting
Teams treat these as the same thing. They are two steps of one process, and the difference is where the list comes from.
| ICP targeting | Lookalike prospecting | |
|---|---|---|
| Starting point | A written profile of the ideal buyer | A list of customers you already have |
| Who writes the rules | You, from judgment | The shared traits in your data |
| Risk | Rules reflect opinion, not wins | Rules copy past customers and miss new segments |
| Needs customers first | No | Yes |
| Best for | A new product with no wins | A product with 10 or more good customers |
| On LeadOcean | industry, employeeRange, jobLevel | The same filters, set from your seed |
Start with an ICP when you have no customers. Switch to lookalikes once you have enough wins to see a pattern. The two should agree. If they do not, trust the customers.
When it matters
Lookalikes are a poor fit when you have under five customers or a single huge one. The seed is too thin, and you copy an accident. Use an ICP list instead. Four situations reward a lookalike list.
Your first outbound push after product-market fit
Once you have a dozen paying customers, a lookalike list is the fastest way to a second dozen. You skip the debate about who the buyer is and copy what already worked.
A segment that converts better than the rest
If one industry or size band closes at twice the rate, build a list of only that. A lookalike search lets you go wider inside the segment, such as a new country, without leaving it.
Account-based plays on a named seed
When ABM targets are picked by hand, lookalikes find the accounts you would have missed. Seed with your best five accounts and search for companies of the same type and size.
Entering a new market
Before you open a new country, run the same traits with a different hqCountry and count the result. A free count tells you how many lookalikes exist there before you hire anyone or write copy.
How LeadOcean handles it
LeadOcean has no similarity score and no "companies like this one" lookup. It does have the company fields a lookalike is built from, as filters on the company search. You supply the traits, and the search returns the matching companies.
These filters map to a lookalike seed:
industry(534 values) andemployeeRange(8 brackets).hqCountry,minFoundedandmaxFounded.technologiesandtechnologyCategories(105 categories, from about 7.8M crawled sites).keywords, which searches company descriptions, specialties and categories.
The full list is in the filter reference (LeadOcean docs, September 2026).
On 2026-10-01, leadocean_count_leads returned 2,170 mailable people (verified, catch_all_valid and catch_all) with industry Software Development, employeeRange 51-200, hqCountry US, jobLevel VP and jobFunction Sales & Business Development. Adding technologyCategories Marketing automation cut that to 211. The second filter is the soft trait from the example, and it shrinks the list by about 90%.
Size a company list for free with count=true and limit=1. Send count as a query parameter and limit in the body, and meta.total comes back capped at 100,000 (LeadOcean OpenAPI, September 2026).
curl -X POST "https://api.leadocean.io/v1/companies/search?count=true" \
-H "x-api-key: $LEADOCEAN_API_KEY" \
-H "Content-Type: application/json" \
-d '{"industry": ["Software Development"], "employeeRange": ["51-200"], "hqCountry": ["US"], "limit": 1}'Drop count=true and raise limit to return companies. Each company returned counts one record. To reach the buyers, run GET /v1/companies/{domain}/people for each domain, or use POST /v1/people/search with the same filters plus jobLevel and jobFunction. To exclude current customers, pass their domains to your own suppression list before you pull people.
The same filters work in the MCP server and on the Exports page of the app (app.leadocean.io), which shows the record price before you start. Pricing is two plans: Free (1,000 records, one-off, no card) and Pro at $499 a month. See pricing.
For the rest of the workflow, see rep prospecting, and compare tools in software with the best integrations and software with the best customization.
FAQ
What is an example of lookalike prospecting?
A company sells to 51-200 person software firms in the US. It searches for more firms with the same industry, size and country, then contacts their VP of Sales. The lookalike traits come from its best customers.
How is lookalike prospecting different from an ideal customer profile?
An ideal customer profile is a description you write. Lookalike prospecting builds that description from the traits your real customers share, then searches for more of them. The first is a hypothesis, the second is measured.
How many customers do I need for a lookalike list?
About ten good customers is a workable floor. Below five, the traits you find may be an accident of who signed first. Use an ICP list until you have more wins.
Does LeadOcean have a lookalike feature?
No. LeadOcean has no "find similar companies" tool. You build the lookalike yourself with company filters such as industry, employeeRange, hqCountry and technologyCategories, and size it free with count=true.
Which traits matter most for a lookalike list?
Industry and company size come first, because they are the most stable and the easiest to match. Location and detected technology come next. Treat funding and keyword traits as soft filters, and test them one at a time.
Build your first lookalike list from your best customers
Free to start. No credit card. 1,000 records to spend whenever you like.
Get your free API key →