Website Contact Scout

Turn a list of business websites into a compact, source-backed contact report. Find published email addresses, clickable phone numbers and linked social profiles, then see exactly which page supplied each result.

Use it to enrich website URLs from your own business directory, an authorized lead export or a CRM cleanup workflow. It does not search Google Maps itself, send messages, guess email addresses or test mailboxes.

What makes this useful

• One consolidated result per website, rather than a separate row for every page.
• Source URL and extraction method for every email, phone link and social profile.
• Follows linked contact, about and imprint pages within the page limit.
• Labels common role inboxes such as sales@ and support@; other addresses remain unclassified.
• Deduplicates repeated website domains and contact values.
• Distinguishes a successful scan with no contacts from a blocked, unavailable or disallowed website.
• Reports partial coverage when secondary pages fail. No invented contacts or confidence scores.

Run it

1. Add up to 20 public HTTPS website URLs in Website URLs.
2. Choose 1–4 pages per website. The starting page counts toward the limit. Only discovered contact/about/imprint links are considered for additional pages.
3. Start the run. The example checks one public Python.org help page; it is a demonstration, not a customer or sales lead.
4. Open Output → Website contacts and export JSON, CSV or Excel using Apify's export controls.
5. Open Evidence and failures → REPORT for failed and unattempted websites. SITE-001, etc. hold the saved per-website evidence.

{
  "startUrls": [{"url":"https://www.python.org/about/help/"}],
  "maxPagesPerSite": 1
}

Output

Field	Meaning
website	Supplied website after query/fragment removal.
status	contacts_found or no_contacts_found for a successfully checked page.
emails	Deduplicated published addresses. No mailbox/deliverability verification.
phones	Values from explicit tel: links, not guesses from arbitrary numbers.
socialProfiles	Linked social URLs; their contents are not crawled or verified.
evidence	Value, kind, source URL, extraction method and optional inbox/platform label.
pagesChecked	Successfully read HTML pages.
warnings	Explicit failures of secondary pages.
checkedAt	UTC observation time.

If every website fails, the run fails with no chargeable dataset results. Read REPORT to see why. If some websites succeed, their results remain available alongside a report of failures. A site that loads but publishes no detectable contacts still counts as a successfully checked website.

Pricing

Launch pricing: USD 0.005 per successfully checked website ($5 per 1,000 websites), including up to the selected page limit and platform usage. One dataset item equals one website scan, not one email or one page. Blocked/unavailable starting pages are not appended to the dataset and do not trigger a result charge. There is no startup charge or AI API fee. Check the live Pricing tab before running.

The maximum run cost limits the number of successful website results. For example, a $0.01 budget permits two websites at this price. Set at least $0.005. If a final storage request has an uncertain outcome, inspect the dataset before starting another run. A run with existing results cannot be resurrected to avoid duplicate result charges. Starting a new run is a new scan and may be charged again.

Coverage and boundaries

• Static HTML only. No JavaScript rendering, login, CAPTCHA solving, proxy rotation or anti-bot bypass. JS-only contact details may be missed.
• robots.txt is checked before page requests. Disallowed pages are skipped; unavailable or restricted robots files produce an explicit failure. Requests are paced at least one second apart per domain and honor supported crawl delays up to ten seconds.
• One bounded retry for HTTP 429/503. Longer retry instructions end that check rather than ignoring the site's requested delay.
• HTTPS only, with public IPv4 DNS. Local/private IPs, credential-bearing URLs and nonstandard ports are rejected. Sites reachable only by IPv6 are unsupported.
• Redirects are limited and remain on the same hostname or its www counterpart. Other domain redirects are reported for review.
• URLs lose query strings and fragments; signed/token-based or query-routed pages are unsupported.
• Response limit: 2 MiB per request and a 12-second response deadline. No guarantee that all contacts on a site are found.
• Common HTML entities and percent-encoded mailto: addresses are supported. JavaScript/Cloudflare obfuscation and image-based contact details are unsupported.
• Common placeholder addresses and image-file false positives are excluded. Phone values are not normalized to E.164. A social URL is a published link, not proof of profile ownership.

Only process websites you are authorized to access and use results appropriately. Published contact information does not establish consent to receive marketing. No emails are sent, and extracted addresses are not tested using SMTP. Raw webpage HTML is not stored; extracted results remain in your Apify storage under its retention/access settings.

Support

Open an issue with an error code and a public or fictional reproduction. Do not put confidential lists, credentials or identity documents into public issues.