Lookalike Company Search API

Give it the domains of your best customers and get back companies that look like them, scored, filtered and flattened into rows you can export.

Most company search APIs make you describe your ideal customer in filters and hope the filters are the thing that made those customers good. This one starts from the customers themselves: it reads what a company's website actually says it does and finds others that say something similar. Structured filters for industry, technology stack, headcount, revenue, funding and location sit on top of that, so you can narrow a semantic result set the ordinary way.

Four modes

Mode	What it does	What you give it
company-search	Returns full company records, cursor-paged	Lookalike domains and/or filters
company-preview	Counts how many companies match and returns a small sample	The same filters
company-enrich	Resolves domains you already hold into full records	A list of domains
people-search	Returns individual contacts at matching companies	People filters, company filters

Run company-preview first. It reports the total match count for a filter set and returns a handful of example companies, so you can size an audience and sanity check the filters before committing to a full paged search.

What comes back

Company rows carry name, legal name, domain, root URL, description and logo; size as a headcount band, an employee count and a LinkedIn employee count; revenue band and founding year; headquarters country, city, region, postcode and street, plus every country the company operates in and a location count; industries, industry categories, the LinkedIn industry and free-text keywords; the detected technology stack and its categories; emails and phone numbers; LinkedIn, Twitter, Facebook, Instagram and YouTube profiles; estimated monthly visits, page views and pages per visit; last funding date, type, amount in USD and Crunchbase link; headcount growth over three, six and twelve months; department headcounts; and the provider's last update date.

Person rows carry name, first and last name, job title and its English translation, computed seniorities and departments, current job description, headline, summary, skills, country, state, location, LinkedIn profile, photo, email and phone where held, connection and follower counts, and the domain of the company they were found at.

Lookalike results also carry similarityScore between 0 and 1, and relevance as an A, B or C band.

Every row carries raw, the untouched provider record, so nothing is lost by the flattening.

Example output

{
  "recordType": "company",
  "matched": true,
  "domain": "example.com",
  "name": "Example Labs",
  "legalName": "Example Labs, Inc.",
  "description": "Battery recycling software for grid storage operators.",
  "companySize": "51-200",
  "employeeCountLinkedin": 148,
  "revenue": "10-50M",
  "yearFounded": 2018,
  "primaryCountry": "us",
  "hqCity": "Oakland",
  "hqRegion": "California",
  "industries": ["Climate Tech", "Industrial Software"],
  "technologies": ["Segment", "Snowflake", "HubSpot"],
  "technologyCategories": ["Analytics", "CRM"],
  "primaryEmail": "hello@example.com",
  "linkedinUrl": "https://www.linkedin.com/company/example-labs",
  "webTrafficVisits": 41200,
  "lastFundingType": "Series B",
  "lastFundingAmountUsd": 32000000,
  "lastFundingDate": "2026-02-11",
  "headcountGrowthTwelveMonths": 39,
  "similarityScore": 0.94,
  "updatedAt": "2026-07-28T09:14:02Z"
}

Pricing

Pay per result. You are charged once per record saved to the dataset. Duplicates are removed before charging, and a domain that resolves to nothing is stored with matched: false free of charge, so a list of dead domains costs you nothing. Apify platform usage is included in the per-result price. Your own provider plan is billed separately by the provider, and each run logs how many provider credits it consumed.

Bring your own API key

This actor does not include data access. You use your own API token from Ocean.io, which is the provider whose API this actor calls. The token is sent as the x-api-token request header rather than as a URL parameter, so it never appears in a log line or a redirect.

FAQ

What is a lookalike company search API?

A lookalike company search API takes one or more companies you already value and returns others that resemble them, rather than making you write filters that approximate what made them a fit. This provider compares website content semantically, so two companies match when they describe similar products and services, not merely because they share an industry label. This actor wraps that API and turns its results into a dataset you can export as CSV, JSON or Excel, or pull from another tool.

How do I build an ideal customer profile list from my existing customers?

Put up to ten of your best customer domains in lookalikeDomains and run company-preview. Read the total match count, then tighten with hqIncludeCountries, companySizes and excludeDomains for the customers you already have. When the count looks right, switch to company-search and raise maxResults.

What is the difference between precise and broad matching?

precise compares what the companies actually sell, using semantic similarity of their website content, and returns a tight set. broad matches anything inside the same industries and returns a far larger set. Start precise. Move to broad only when precise returns too few companies to work with.

Can I filter companies by the software they use?

Yes. technologiesAny, technologiesAll and technologiesNone filter on software detected on the company website, and technologyCategoriesAny filters on broader groupings such as Analytics or Security. Excluding a competitor's product with technologiesNone is the usual way to strip out accounts that are already committed elsewhere.

How do I find decision makers at the companies I just found?

Run people-search with the same company filters plus seniorities set to Founder, C-Level or VP, and jobTitleKeywordsAny for the function you want. Set peoplePerCompany to cap how many contacts come back from each account so one large company does not consume the whole run.

How does pagination work?

Paging is a cursor, not an offset. Each response carries a searchAfter value that the next request sends back, and a response without one means the result set is finished. The actor handles this for you and simply stops when the provider runs out of pages or maxResults is reached.

Why are some of my lookalike domains ignored?

A seed domain is only usable if the provider already holds enough content for it. Domains that are missing, still being crawled, or lacking a usable content vector are listed in the run log with the reason, and the search proceeds on the remaining seeds. Adding more seeds is the fix.

How many results can one run return?

Up to 25,000 per run, paged at up to 1,000 records per request. Preview mode is capped by the provider at 50 sample companies, but reports the full match count regardless of that cap.

Do I need my own API key?

Yes. This actor is bring your own key, and your usage is billed on your own provider plan. The actor charges separately for each result it returns.

Keyword map

lookalike company search API, similar companies API, ideal customer profile API, ICP builder API, B2B company data API, company search API, company enrichment API, domain to company API, technographics API, technology stack lookup API, firmographics API, company funding data API, B2B prospecting API, account based marketing data, sales prospecting API, contact data API, people search API