Contact Details Scraper

Extract public contact information from one website or a list of websites. The Actor follows a small set of high-value, same-site pages and returns one consolidated record per site.

The default crawler uses fast HTTP requests rather than a browser. It is designed for predictable cost at a planned pay-per-result price of $0.0015 per successful site, below the current $0.002 competitor baseline. Empty, blocked, invalid, robots-disallowed, and no-contact sites produce an uncharged diagnostic row.

What it extracts

• Public email addresses, including mailto: links and common [at] / [dot] obfuscation
• Phone numbers from tel: links and visible page text, with optional country normalization
• LinkedIn, Facebook, Instagram, X, YouTube, TikTok, Pinterest, GitHub, Threads, and Bluesky links
• Visible and JSON-LD postal addresses
• Contact forms with action, method, field names, and field types
• JSON-LD Organization, LocalBusiness, and Person data
• Website title and meta description
• Every source page for every extracted value
• Per-page crawl metadata and clear error diagnostics

The crawler prioritizes the home, contact, about, team, company, legal, privacy, terms, and impressum pages. It also discovers equivalent pages from link text and paths.

Input

Field	Type	Default	Description
websites	string array	Apify website in UI	Domains or full URLs. Values without a scheme use HTTPS.
startUrls	request-list sources	empty	Alternative Apify-compatible URL input. Can be combined with websites.
maxPagesPerSite	integer	6	Maximum page attempts per site, from 1 to 20.
maxConcurrency	integer	5	Websites processed in parallel, from 1 to 20.
requestTimeoutSecs	integer	12	Timeout for each HTTP request.
maxResponseSizeKb	integer	1536	Maximum downloaded HTML size per page.
respectRobotsTxt	boolean	true	Skip pages disallowed for the Actor user agent.
defaultCountry	string	empty	Optional two-letter country code for local phone numbers, such as US, GB, or CA.
proxyConfiguration	proxy	off	Optional Apify or custom proxy. Direct HTTP is the profitable default.

Example:

{
  "websites": [
    "apify.com",
    "https://www.example.org"
  ],
  "maxPagesPerSite": 6,
  "respectRobotsTxt": true,
  "proxyConfiguration": {
    "useApifyProxy": false
  }
}

startUrls also accepts objects:

{
  "startUrls": [
    { "url": "https://www.apify.com" },
    { "url": "https://example.org/contact" }
  ]
}

Output

One successful row is stored for each site with useful public contact data:

{
  "ok": true,
  "siteUrl": "https://example.org/",
  "requestedUrl": "https://example.org/",
  "domain": "example.org",
  "title": "Example Organization",
  "description": "Example website description.",
  "emails": [
    {
      "email": "hello@example.org",
      "sourcePages": [
        "https://example.org/contact"
      ]
    }
  ],
  "phones": [
    {
      "phone": "+1 202 555 0100",
      "e164": "+12025550100",
      "country": "US",
      "sourcePages": [
        "https://example.org/contact"
      ]
    }
  ],
  "socialLinks": [
    {
      "platform": "linkedin",
      "url": "https://www.linkedin.com/company/example",
      "sourcePages": [
        "https://example.org/about"
      ]
    }
  ],
  "addresses": [],
  "contactForms": [],
  "jsonLdOrganizations": [],
  "jsonLdPeople": [],
  "sourcePages": [
    "https://example.org/contact",
    "https://example.org/about"
  ],
  "crawledPages": [
    {
      "url": "https://example.org/",
      "status": 200,
      "title": "Example Organization",
      "contactsFound": 3
    }
  ],
  "pagesCrawled": 4,
  "pagesAttempted": 6,
  "emailCount": 1,
  "phoneCount": 1,
  "socialLinkCount": 1,
  "addressCount": 0,
  "contactFormCount": 0,
  "robotsTxtRespected": true,
  "pageErrors": []
}

Values contain their own sourcePages, which makes the result easier to verify than a flat list of contacts.

Billing and diagnostics

The code emits the pay-per-event event site once, and only after a real successful row has been stored. The intended future event price is $0.0015 per site.

These rows are never charged:

• _sample: true: empty input preview used by Apify automated tests
• BAD_INPUT or UNSAFE_URL: invalid or blocked target
• NO_CONTACTS: pages loaded, but no useful public contact signal was found
• ROBOTS_DENIED: relevant pages were excluded by robots.txt
• BLOCKED, RATE_LIMITED, TIMEOUT, NETWORK, or other request errors

Each requested site is isolated. One blocked or invalid site does not fail the rest of the run, and expected per-site failures exit successfully with diagnostic data.

Security

The Actor is designed for user-supplied URLs:

• Only http:// and https:// are accepted.
• URL credentials are rejected.
• Localhost and local network hostnames are rejected.
• Loopback, private, link-local, carrier-grade NAT, multicast, documentation, benchmark, unspecified, and reserved IP ranges are rejected.
• Every hostname and every redirect is resolved and validated.
• Direct requests use a DNS lookup guard at connection time.
• Redirects are manual, revalidated, and limited to the requested site.
• Crawling is limited by site, page count, timeout, and response body size.
• Binary and common document URLs are not crawled.

When a proxy is enabled, the destination is still resolved and validated before the proxy request. Direct mode provides the strongest DNS pinning and is the recommended default.

Robots.txt

respectRobotsTxt is enabled by default. The Actor fetches each site's robots.txt and checks every candidate URL against its own user agent. If robots.txt is unavailable, normal public-page crawling continues. Set the option to false only when you have permission and a specific reason.

Proxies and profitability

Most public company sites do not need residential proxies, so proxy use is off by default. This keeps the planned $0.0015/site price profitable:

• 512 MiB default memory
• no browser startup
• bounded HTTP responses
• six page attempts by default
• site-level concurrency
• successful-result-only billing

Enable a datacenter or residential proxy only for sites that block direct cloud requests. Residential proxy bandwidth can cost more than the result price on heavy pages, so it should not be the default for this Actor.

JavaScript-heavy websites

This version intentionally does not launch a browser. Browser rendering would make the planned $0.0015/site price unreliable when several pages need JavaScript and residential bandwidth. Many modern sites still expose contact details in server-rendered HTML, metadata, JSON-LD, or footer links, which this Actor extracts.

For a fully client-rendered site, use its public contact page as the start URL or enable a proxy if the issue is IP blocking rather than rendering. A browser fallback should only be added with a separate higher-priced event after measured cloud cost proves it remains profitable.

Run locally

Requires Node.js 20 or newer:

npm ci
npm test
npm start

For local Actor input, place INPUT.json in the default Apify key-value store or run with the Apify CLI. Results are written to the default dataset.

Responsible use

Extract only public business contact information you are allowed to collect and process. Follow website terms, robots rules, privacy law, anti-spam law, and data retention requirements applicable to your use case. The Actor does not bypass logins or access controls.