Company Website Profile: Social Links, Tech, VAT/CVR Numbers

Give it a list of company websites. For each one it returns a compact company profile: the company name, registration and VAT numbers printed on the site (Danish CVR, Norwegian and Swedish organisation numbers, Finnish Y-tunnus, EU VAT), the company's social profile links, the web-shop platform or CMS, and the marketing, consent and payment tools the site uses.

It is built for lead lists, B2B enrichment, market mapping and agency prospecting: "which of these 500 companies run Shopify, use a Meta Pixel, and what is their CVR number and LinkedIn page?"

What makes it different

• Registration numbers with check digits. CVR, Norwegian and Swedish organisation numbers and Finnish Y-tunnus are only returned when their official check digit is valid, so a random 8-digit number on a page is not mistaken for a company ID. The number is the key to join the profile with official registers.
• No contact data. E-mail addresses, phone numbers and street addresses are not collected at all. Personal social profiles (LinkedIn /in/) and names of employees are never collected. Sites that publish Person data instead of Organization data are flagged personal_site_suspected, and name and description are left out for them.
• Polite by design. The Actor reads each host's robots.txt first and follows it (a site that blocks us is skipped, not circumvented), makes at most one request per second per host, reads at most 6 pages per site, and never uses proxies, logins or disguises. A site that answers 403/429 is skipped and not charged.
• You only pay for sites that were read. One result per website. Sites that were unreachable, blocked by robots.txt or refused the request are not charged and are listed in the key-value store record FAILED_DOMAINS.

Input

Field	Description
domains	Websites, one per line (novonordisk.com or a full URL). IP addresses and internal names are rejected.
pagesPerSite	Home page plus sub-pages found through the home page's own links: contact, imprint, about, terms/privacy. Default 3, max 6.
includeDescription	Add the site's own meta description, cut to 200 characters. Default on.
maxConcurrency	Websites processed in parallel (default 5). Each host still gets at most 1 request per second.
delaySeconds	Pause between requests to one host (min 1).

Example:

{
  "domains": ["novonordisk.com", "matas.dk", "jysk.dk"],
  "pagesPerSite": 3
}

Output (one row per website)

domain, url, httpStatus, entityType, companyName, nameSource, legalName, description, language, languages, countryHint, registrationNumbers (country, type, value, checksumValid), socialProfiles (facebook, instagram, x, linkedin company pages, youtube, tiktok, pinterest, threads), platform, technologies (name, category), isWebshop, foundedYear, employeesHint, logoUrl (only when enabled), pagesAnalysed, flags, checkedAt.

Example (shortened):

{
  "domain": "novonordisk.com",
  "companyName": "Novo Nordisk",
  "countryHint": "DK",
  "registrationNumbers": [{"country": "DK", "type": "CVR", "value": "24256790", "checksumValid": true}],
  "socialProfiles": {"linkedin": "https://www.linkedin.com/company/novo-nordisk", "x": "https://x.com/novonordisk"},
  "technologies": [{"name": "OneTrust", "category": "consent"}],
  "isWebshop": false,
  "flags": []
}

Quality flags

• redirected_to_other_domain: the home page redirected to another domain (for example a regional or checkout host). Check url.
• little_text_in_html_page_may_need_javascript: the page has almost no text in its HTML (often a JavaScript-only site), so fewer fields could be found.
• name_guessed_from_domain: no company name was published; the name was taken from the domain.
• personal_site_suspected: the site publishes Person structured data; name and description are left out.
• subpage_blocked_by_robots: a contact or imprint page exists but robots.txt disallows it, so it was not read.

Limits and honest notes

• Only data that is printed on the site's public pages is returned. If a site does not print its registration number, none is returned.
• Sites built purely with JavaScript, or that block automated clients, return less or are skipped. The Actor does not use a browser and does not bypass blocks.
• Technology detection is a source-code pattern match, not a guarantee.
• Social links are taken from the pages and the site's structured data; large groups sometimes link to sister brands, so check the result.
• You are responsible for how you use the data. Only profile websites you are allowed to read, and check your own legal basis before you use the data (for example for outreach).
• The Actor identifies itself in the User-Agent as CompanyWebsiteProfile (Apify Actor) with a contact URL.