Business Website Enricher — Source-Backed Company Contacts

Turn your own list of business websites into structured, auditable company metadata. This Actor only visits a company's publicly accessible own website. It is not a Google Maps, employee, personal-email, or reputation scraper.

Input

• websiteUrls (required): 1–100 domains or HTTP(S) URLs. Domain-only inputs use HTTPS.
• maxPagesPerSite: 1–10 successful HTML pages per business, default 3.
• crawlContactPages: include permitted Contact links and conventional Contact paths, default true.

Dataset output

One record per unique submitted website, including status, companyName, businessEmail, businessPhone, businessAddress, socialUrls, sourcePages, pagesCrawled, fieldEvidence, robotsStatus, and processedAt. Every populated field has a source page URL, method, and high/medium/low confidence. Social links also include their source page and confidence.

Business email means role mailbox on the website's domain (examples: info@, sales@, contact@, support@). Personal addresses such as jane@company.example and named employee profiles are excluded. Corporate social links are discovered from the company's own pages, not by visiting the linked social networks.

Charging: pay per useful website

Target price: $0.006 per successful website with at least one company contact detail or corporate social link ($6 per 1,000 useful sites). No per-page or startup event. Invalid, blocked, duplicate, or empty results are not charged. Apify's run spending cap applies. See the Store Pricing tab for the actual effective event price; the account's publishing configuration is authoritative.

Boundaries

• Robots.txt respected before fetching business HTML. Fail closed if robot rules are unavailable or deny access.
• Same-domain only; redirects cannot be used to escape the allowed site. No search engines, Maps, directory bypass, third-party API, login pages, browser automation, or proxies.
• 1 request per second per site; per-request timeouts; HTML size cap 2 MB; per-site page cap, 100 website cap, no unbounded crawl.
• HTML links and organization JSON-LD are parsed; JavaScript-render-only contact data may not be visible. Uncertain organization names are omitted rather than guessed.
• Some websites prohibit automated reuse in terms independent of robots.txt. Users must ensure their URL inputs and data use are permitted. Do not use results for unlawful unsolicited outreach.

Example input

{
  "websiteUrls": ["https://www.hetzner.com", "https://www.digitalocean.com"],
  "maxPagesPerSite": 3,
  "crawlContactPages": true
}

Example output (illustrative, not a real company)

{
  "inputUrl": "https://example.org",
  "status": "ok",
  "companyName": "Example LLC",
  "businessEmail": "info@example.org",
  "businessPhone": null,
  "businessAddress": null,
  "socialUrls": [{"platform":"linkedin","url":"https://www.linkedin.com/company/example","sourceUrl":"https://example.org","confidence":"medium"}],
  "fieldEvidence": {
    "companyName": {"value":"Example LLC","sourceUrl":"https://example.org","confidence":"high","method":"json-ld"},
    "businessEmail": {"value":"info@example.org","sourceUrl":"https://example.org/contact","confidence":"high","method":"mailto-link"},
    "businessPhone": null,
    "businessAddress": null
  },
  "sourcePages": ["https://example.org", "https://example.org/contact"],
  "pagesCrawled": 2
}

Tests, known limits and support

Five labeled site-structure fixtures include JSON-LD, footer contacts, personal-email false positives, company-only contact points, and JavaScript-only content. See reports/fixture_validation.json. They are synthetic tests, not proof of general internet extraction accuracy. Real-world evaluation is kept separately under reports/live_business_sites.json. Requests may be blocked, pages may exceed 2MB, or sites may load contact information only through JavaScript. This Actor leaves unverified data blank.

Source code and maintenance scope: Netcup /opt/APIFY/API09 only. No dependency on API06, no Google Places integration. If Places is added in a later version, its license, retention, and display requirements must be separately reviewed first.