B2B Company Data API - Firmographics and Tech Stack

Search a B2B company database and export the matches as clean rows, with the firmographics and the technology stack in the same record. Filter by domain, company name, specialties, NAICS and SIC code, country, state, postal code and a radius around it, employee count, annual revenue, and by the technology a company actually runs, at four levels: category, subcategory, product and vendor. You pay only for the companies you actually receive.

This actor is built for technographic targeting, building competitive-displacement lists, account list building, and enriching a CRM with verified firmographics.

What you can do with it

• Find every company in a market running a specific product, so you can target its users or its competitors.
• Build an ideal customer profile list by industry code, headcount and revenue band.
• Search around a location: a postal code plus a radius in miles.
• Restrict a list to human-verified records only, for outbound where accuracy matters most.
• Pull the full detected technology stack for each account alongside its firmographics.
• Export everything as JSON, CSV or Excel, or push it into a CRM or warehouse.

Features

• Every documented company filter is a real input field, grouped into company, location, size and revenue, technology, and quality and order.
• The provider's two list separators are handled correctly: ordinary filters are comma joined, and the four technology filters are tilde joined because technology and vendor names contain commas themselves.
• Technology targeting works at all four levels the provider exposes: category, subcategory, product and vendor.
• Page pagination is walked automatically and stops on a short page, on the reported total, or on the provider's 10,000-record depth limit, which is logged rather than hit silently.
• Full pages are always requested, because the provider bills its own credits per page accessed rather than per record.
• The headquarters address is picked out of the address list by the provider's own head-office flag rather than by taking whichever came first, and every address is kept.
• Revenue is exposed three ways where the provider has it: the value used, the actual reported figure and the predicted figure, plus the band label.
• Duplicate companies are removed across pages, keyed on company id then domain then name, before anything is charged.
• The untouched provider payload is kept under raw, so nothing is lost.
• Rate limits are handled with retry-after aware backoff, and a run that hits a limit keeps everything already collected.
• Pay-per-result pricing with a hard result cap so a run can never surprise you.

Input

Field	Type	Description
companyDomains	array	Company domains to match. The most reliable key.
companyNames	array	Company names to match.
specialties	array	Keywords describing what the company does.
naicsCodes / sicCodes	array	Classification codes.
countryCodes / states / zipCodes	array	Location filters.
zipCodeRadiusMiles	integer	Widen the postal code filter by a radius. United States only.
internationalOnly	boolean	Only companies flagged international.
minEmployees / maxEmployees	integer	Employee count range.
minRevenueMillions / maxRevenueMillions	integer	Revenue range, in millions USD.
techCategories / techSubcategories / techProducts / techVendors	array	Technographic filters.
verifiedOnly	boolean	Only human-verified records.
sortBy / sortDirection	string	Order by company name or location, ascending or descending.
rawParams	object	Any extra query parameter, merged last.
maxResults	integer	Stop after this many companies (1 to 10,000).
pageSize	integer	Companies per API call (1 to 50).
apiKey	string	Your API key. Required, stored securely.

At least one filter is required.

Example output

Each dataset item is one company. Fields are populated where the provider has data, and the untouched payload is kept under raw:

{
  "companyId": "c_8842197",
  "name": "Example Analytics",
  "displayName": "Example Analytics, Inc.",
  "domain": "example.com",
  "domains": ["example.com", "example.io"],
  "industry": "Computer Software",
  "sector": "Technology",
  "naicsCode": "541511",
  "sicCode": "7372",
  "employees": 310,
  "employeeRange": "201-500",
  "revenue": 44000000,
  "actualRevenue": 44000000,
  "predictedRevenue": 46500000,
  "revenueRange": "$25M-$50M",
  "street1": "88 Example Avenue",
  "street2": "Suite 400",
  "city": "Boston",
  "state": "MA",
  "zip": "02110",
  "country": "US",
  "location": "Boston, MA, US",
  "allAddresses": [
    { "street_1": "88 Example Avenue", "city": "Boston", "state": "MA", "zip": "02110", "country": "US", "is_hq": true }
  ],
  "isInternational": false,
  "specialties": ["revenue forecasting", "subscription analytics"],
  "technologies": ["Salesforce", "HubSpot Marketing Hub", "Snowflake", "dbt"],
  "technologyCount": 4,
  "linkedinUrl": "https://www.linkedin.com/company/example-analytics",
  "logoUrl": "https://logos.example-provider.com/example.com.png",
  "companyStatus": "Active",
  "verified": true,
  "raw": { "company_name": "Example Analytics", "company_domains": ["example.com"] }
}

Pricing and the credit model

This actor uses pay-per-result pricing. You are charged a fixed price for each company delivered to the dataset, after duplicates are removed. A search that matches nothing costs nothing. Set maxResults to cap how many rows a run can collect, and set a run spending limit as a second safety net.

Your own provider plan meters separately, and it does so per page accessed rather than per record, with at most 50 records per page. That is why the actor always requests full pages: a page of 5 costs the same as a page of 50. Set maxResults to a multiple of your page size to avoid paying the provider for a page you barely use.

Bring your own API key

This actor connects to a third-party B2B company data provider (SalesIntel) on your behalf and requires your own provider API key. Create a key in the provider's platform and paste it into the apiKey input. It is stored securely and never logged, and the actor never ships with or shares credentials.

Frequently asked questions

What data does this actor return?

Company id, name and display name, every known domain, industry and sector with NAICS and SIC codes, employee count and band, revenue as the value used plus the actual and predicted figures and the band label, the headquarters address broken into street, city, state, postal code and country along with every other address on file, whether the company is flagged international, specialties, the full detected technology stack with a count, LinkedIn URL, logo, company status and whether the record is human verified. Every record also carries the untouched provider payload.

How does technology targeting work?

Four levels, from broad to specific: category, subcategory, product and vendor. Use category to sweep a whole software class, product to name the exact tool, and vendor to catch every product from one supplier. Check the provider's technology taxonomy reference for the valid values, since these are controlled vocabularies rather than free text.

Why are technology filters separated differently from the others?

Because the provider says so, and for a good reason: product and vendor names frequently contain commas, so a comma-separated list would split them in the wrong place. Those four filters use a tilde as the separator instead. The actor handles this for you, whether you type a list or paste a delimited string.

What does verified mean?

The provider maintains a human-verified subset of its data alongside the machine-collected records. Turning on verifiedOnly restricts results to the verified set, which is smaller but more accurate. Every returned row also carries its own verified flag, so you can filter after the fact instead.

How does the postal code radius work?

Set one or more postal codes and a radius in miles, and the search widens to everything within that distance. The provider supports this for United States postal codes only, so a radius set alongside non-US codes is ignored by the provider rather than erroring.

Why does the run stop at 10,000 records?

The provider limits how deep pagination can go, at 10,000 records, regardless of how many companies match. The log says so explicitly when the limit is reached. Split the search, for example one state or one industry code at a time, to reach past it.

How should I set the page size?

Leave it at 50. The provider bills a credit per page accessed rather than per record, so smaller pages cost the same and simply use more of your quota. The only reason to lower it is to reduce response size on a slow connection.

Which is the better company key, domain or name?

Domain. It is unique and already standardised, so it matches cleanly. If you only have names, strip suffixes such as Inc, Corp, Co and LLC first, which is what the provider recommends, because those vary between records.

How does the actor handle rate limits?

It throttles requests, honours the retry-after signal on a rate-limited response, and backs off exponentially. A response saying the account is out of credits or the endpoint is not licensed stops the run early and keeps everything already collected.

How am I charged?

You pay a fixed price per company returned to the dataset. Duplicates are removed before charging and empty runs cost nothing. Use maxResults and the run spending limit to control cost.

What output formats are available?

The dataset can be exported as JSON, CSV, Excel, HTML or RSS, or read through the API for pushing into a CRM or warehouse.

Keyword map

b2b company data api, technographic data api, company tech stack api, technology install data, find companies using salesforce, competitive displacement list, firmographic data api, naics code company search, sic code company search, company revenue data api, employee count filter api, verified b2b data, account list building api, ideal customer profile data, crm enrichment api, postal code radius company search, company specialties search, b2b prospecting database, intent and technographic targeting, bulk company export api.