Email Scraper: Website Emails, Phones & Contact Details avatar

Email Scraper: Website Emails, Phones & Contact Details

Pricing

from $0.79 / 1,000 page crawleds

Go to Apify Store
Email Scraper: Website Emails, Phones & Contact Details

Email Scraper: Website Emails, Phones & Contact Details

Find email addresses, phone numbers, social profiles and postal addresses on any website. Paste URLs or domains; several pages per site are visited. One deduplicated row per site. No API keys needed.

Pricing

from $0.79 / 1,000 page crawleds

Rating

1.0

(1)

Developer

NanoScrape

NanoScrape

Maintained by Community

Actor stats

5

Bookmarked

334

Total users

30

Monthly active users

2 hours ago

Last modified

Features

Limited permissions

Share

Website Email Finder

Email extractor for company websites on Apify: finds the business email, phone numbers, social profiles and postal address of each site. Give it website URLs, or companies with an id, a name and an address, and it returns one row per site with a chosen primary_email, an email_status, the page it came from and the contact persons the site names. Price: $0.001 per page crawled, no monthly fee. Run it from the Console, the API, n8n, Make or Zapier.

Integration stopped on Nov 11? After a security incident, Apify rotated the API tokens of affected accounts. If yours was rotated and you call this Actor from n8n, Make, Zapier or your own code, paste your new token (Console → Settings → API & Integrations). Schedules inside Apify were not affected.

What this Actor is for

This email scraper reads the pages of a company's own website, the domain the business controls, and returns the contact details it finds. Feed it a homepage, a contact page URL, a branch page of a larger organisation, or a batch of company domains.

Designed for:

  • Company homepages and contact pages (acme-corp.com, firma-gmbh.de, entreprise.fr)
  • Impressum pages on German, Austrian and Swiss business sites
  • Branch pages of a larger organisation, such as one daycare centre or one office on a parent site
  • About and team pages on SaaS, agency or B2B websites
  • Enrichment of Google Maps results (the Google Maps Scraper calls this Actor for the websites it finds)

Not designed for:

  • Social media pages such as Instagram profiles, Facebook pages or TikTok accounts. They sit on someone else's domain. Use instagram-scraper for Instagram profiles.
  • Marketplace listing pages (Yapo, OLX, MercadoLibre, Craigslist, Marktplaats). The URL belongs to the marketplace operator, so you would get the operator's address, not the seller's. Such rows carry a marketplace_note.
  • Directories and review sites that list other businesses. The address on the page belongs to the directory. When you give the place (name and address) the Actor follows the place's own link on that page once and reads that website instead (website_type_original directory). Without such a link the row gets website_type directory and no primary_email.

How it works

For each website the Actor:

  1. Fetches the page you gave it.
  2. Fetches the pages that carry the legal contact first: contact and legal links under your input path (for a branch URL), the site root, the footer link to the Impressum or Imprint, and the paths /impressum, /kontakt, /contact and /imprint. These come before any other page, so a small maxPagesPerSite still reaches them.
  3. Spends the remaining page budget on other contact, about and team pages found in the navigation, the sitemap and language alternates.
  4. Decodes addresses that are written out or encoded, for example name(at)host(dot)de, name [at] host [dot] de, mailto: links in any letter case or URL-encoded, Cloudflare-protected addresses and JSON-LD or microdata emails.
  5. Chooses one primary_email and records why (email_status) and where (email_source_url).
  6. Reads the persons named on the Impressum, Kontakt, Team and About pages (see Contact persons).
  7. Deduplicates everything and writes one row per input.

Pages are fetched in parallel. If a site refuses the standard connection, the Actor retries it on more capable connection levels automatically and keeps using the level that worked for that site. This does not change the price per page. A site whose certificate is invalid is still read, and its row has tls_verified set to false.

Choosing the primary email

emails still lists every valid address found, exactly as in earlier versions. primary_email is the address likely to reach the business you asked about:

  1. Branch mailbox. When the URL points into a branch path, or you pass name and address, a mailbox that matches the branch (by local part, or by the text around it) wins over the general mailbox of the parent organisation.
  2. Own domain. Otherwise a mailbox on the site's own domain.
  3. Parent or other mailbox. Used only when it appears on a contact or legal page or in structured data, and email_status says parent_only.

Addresses of the website agency, the data protection officer (for example datenschutz@), certification bodies, hosting or platform providers and unrelated foreign domains are never chosen. They stay in emails_other and in emails_detail.

Several places on one website

A parent organisation often has one website for many branches. Pass the place context (name, address, place_id) for each place and the Actor crawls the site once, then answers each place separately in place_results. Branch pages are found through links that name the place, the mailbox is scored against the place name, street and postcode (the local part, the domain name for a branch on its own domain, and the text next to the address), and a mailbox that clearly belongs to another place is not given to this one. contact_scope says whose mailbox the primary email is: branch, parent (the organisation or the whole site) or unknown. matched_branch_url is the page of the site that belongs to the place.

Three ways to pass the context, all accepted:

  • companies[] entry with name (or company_name), address and place_id.
  • companies[] entry with a place object or a places list, when one website serves several places.
  • Top level places[] list with website_url per place. Places on the same website are grouped automatically.

Contact persons

contacts lists the named persons a site publishes, each as {name, role, email, source_url}. The Actor reads the Impressum, Kontakt, Team and Ueber uns pages and looks for a name next to a role word, for example Leitung, Kita-Leitung, Einrichtungsleitung, Ansprechpartner, Geschaeftsfuehrung, Inhaber, Vertreten durch, Verantwortlich i.S.d. MStV (and English equivalents such as managing director, owner, contact person, represented by). The email is only set when it is written on the same line or in the same block. Nothing is guessed or built from a mailbox name: a site that names nobody returns an empty list. A data protection officer is listed last with role dpo. For a shared website the persons of the matching branch page come first, then the persons of the parent (for example the managing director named in the Impressum). No key and no extra price: contacts come from the pages already crawled.

Per email detail

emails_detail explains every email of the row: domain_match is own (on the site's domain), branch (names the requested place), foreign (another domain) or freemail (a free mail provider), and type is generic, personal, privacy or role, with the source_url where it was seen. kind is branch (the mailbox of the requested place), carrier (the parent organisation, an agency or another body) or person. is_role_mailbox is true for a shared mailbox such as info@ or kita@ and false for a named person. confidence runs from 0 to 1 and says how sure the match to the place is. The list is ordered by relevance: the place's own mailbox first, then the site's shared mailbox, then persons, then carriers, and the data protection officer last. primary_email is the first entry. A mailbox on an umbrella domain with a daycare-specific local part that is listed on the place's own Impressum counts as the place's own mailbox.

Email status values

email_statusMeaning
branch_contactA mailbox of the requested branch was found.
own_domainA mailbox on the site's own domain was found.
parent_onlyOnly a parent organisation's mailbox was found.
not_foundThe pages were read and contain no usable email.
fetch_errorThe site, or an important contact or legal page, could not be fetched. The email may exist, so retry later. A failed fetch is never reported as not_found.
no_websiteThe company had no website URL. No request was made.
directory_pageThe URL is a directory, listing or social page and no link to the place's own website was found. No primary_email is taken from it.

website_type is one of own, directory, social or unknown.

Social platforms extracted

Profile links to nine platforms are picked up when they appear on the crawled pages: LinkedIn, Facebook, Twitter / X, Instagram, YouTube, TikTok, WhatsApp (normalized wa.me link or group invite), Telegram and Signal.

Use with AI agents (MCP)

Connect this Actor to any MCP-compatible AI client, such as Claude Desktop, Claude.ai, Cursor, VS Code, LangChain or LlamaIndex.

Apify MCP server URL:

https://mcp.apify.com?tools=santamaria-automations/website-email-scraper

Example prompt once connected:

"Use website-email-scraper to find the primary email of https://www.example-company.ch and show email_status and email_source_url."

Input example

{
"urls": ["https://www.example-company.ch", "another-business.de"],
"companies": [
{
"company_id": "acme-123",
"website_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"name": "Kita Matthaeusgemeinde",
"address": "Platz der Deutschen Einheit 1, 65183 Wiesbaden"
}
],
"maxPagesPerSite": 10
}

Input parameters

ParameterTypeDefaultDescription
urlsstring[]noneWebsite URLs or bare domains to scan.
companiesobject[]noneEntries with company_id and website_url, plus optional place context: name (or company_name), address, place_id, a place object or a places list. A company_id with an empty website_url returns a no_website row.
placesobject[]noneAlternative list of places, each with website_url, name, address, place_id and optional company_id. Places on the same website are crawled once.
maxPagesPerSiteinteger10Pages to crawl per site (1-50).
maxSecondsPerSiteinteger45Time limit per website in seconds (5-600). When it runs out, the crawl of that site stops, everything found so far is kept and error_message says partial: time budget. The contact and legal pages are read first, so only the other pages are cut. A site that answers very slowly is also given up early.
followExternalLinksbooleanfalseFollow links to other domains.
enableBrowserFallbackbooleanfalseRe-run blocked sites on the sister browser Actor (separate billed run).

Output example

{
"company_id": "acme-123",
"url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"domain": "matthaeus-wiesbaden.de",
"status": "success",
"website_type": "own",
"primary_email": "kita.matthaeusgemeinde.wiesbaden@ekhn.de",
"email_status": "branch_contact",
"email_source_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"emails": [
"kita.matthaeusgemeinde.wiesbaden@ekhn.de",
"matthaeuskirchengemeinde.wiesbaden@ekhn.de"
],
"emails_other": ["matthaeuskirchengemeinde.wiesbaden@ekhn.de"],
"contact_scope": "branch",
"matched_branch_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"contacts": [
{"name": "Anna Muster", "role": "Kita-Leitung", "source_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette"}
],
"emails_detail": [
{"email": "kita.matthaeusgemeinde.wiesbaden@ekhn.de", "domain_match": "branch", "source_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette", "type": "generic"}
],
"phones": ["0611 522546"],
"email_sources": {
"kita.matthaeusgemeinde.wiesbaden@ekhn.de": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"matthaeuskirchengemeinde.wiesbaden@ekhn.de": "https://www.matthaeus-wiesbaden.de/"
},
"social_urls": {},
"address": null,
"tls_verified": true,
"pages_crawled": 3,
"contact_page_url": "https://www.matthaeus-wiesbaden.de/kindertagesstaette",
"checked_at": "2026-11-13T09:30:00Z"
}

New fields since v1.8, all additive: primary_email, emails_other, website_type, email_status, email_source_url, checked_at. New in v1.9: contacts, emails_detail, contact_scope, matched_branch_url, and (only when applicable) website_type_original, directory_url and place_results. Every earlier field is unchanged.

place_results is present when places were given. Each entry holds company_id, place_id, name, source_url (the page of the email, or the page that was checked when nothing was found), emails, primary_email, contact_scope, matched_branch_url, email_status, email_source_url, checked_at, emails_detail and contacts for one place.

The run status message reads "N websites checked, M with emails, K with a contact name."

Pricing

You pay per page scanned. The price per page does not depend on how many connection levels a site needed.

EventPriceDescription
Page crawled$0.001Each page visited on a website

How much does it cost?

Pages per site10 websites100 websites1,000 websites
5 pages$0.05$0.50$5.00
10 pages$0.10$1.00$10.00
20 pages$0.20$2.00$20.00

No monthly fees. No minimum spend. Rows for companies without a website, dead domains and parked domains cost nothing.

How many pages should I crawl?

SettingSpeedEmail discoveryGood for
5 pagesFastAbout 34% of sitesQuick checks, large batches
10 pagesMediumAbout 50% of sitesTypical use
20 pagesSlowerAbout 70% of sitesSmaller batches on larger sites

More pages usually means more emails found. German, Austrian and Swiss sites usually have an Impressum with an email and phone, and the Actor fetches it before other pages.

Tips

  • Pass name and address in companies when the site belongs to a larger organisation and you want the branch mailbox.
  • Treat fetch_error as "try again", not as "no email".
  • Use email_source_url to check where an address came from before you send mail.
  • No API keys are needed. The Actor uses pattern matching and HTML parsing, not an LLM.

Find businesses to enrich

Real estate

E-commerce

Enrichment

Changelog

v1.9.31 (2026-11-16)

  • Websites whose server does not answer are given up sooner (8 seconds on the first attempt instead of 20) and retried through the next connection route, so one unreachable website no longer holds a worker for up to a minute.

v1.9.30 (2026-11-16)

  • Stops crawling a website once its own email address is found on the home, contact, legal-notice or first contact-style pages, so fewer pages are charged per email found. Websites without an own email address on those pages are crawled up to the page limit as before. Inputs that list several branches of one organisation are unchanged and are still crawled in full.

v1.9.29 (2026-11-15)

  • Faster page processing, so more pages fit in each website's time budget on small memory settings. Results are the same.

v1.9.28 (2026-11-15)

  • More detailed per-site log lines: pages crawled, time used and the reason the site stopped.
  • Sites whose home page cannot be loaded are now logged as a warning instead of an error.
  • Results are now saved in small groups during the run: after 25 finished sites, or after 15 seconds, whichever comes first, and always when the run ends or is stopped. Only the sites in a saved group are charged.

v1.9.27 (2026-11-15)

  • Results are delivered as each site finishes. A run that is stopped or runs out of time keeps every website it had finished, and only those are charged. The rows in the dataset now appear in the order the sites finish, not in the order of your input list.
  • A run that is stopped, or close to its time limit, starts no new website, gives the websites in progress a short margin, and ends with a status message that says how many websites were delivered.
  • A run that is restarted or resumed skips the websites already in its dataset. They are not crawled, stored or charged a second time.
  • With the browser retry switched on, rows that may still be handed to the browser retry are stored at the end of the run, as before. All other rows are stored at once.

v1.9 update (2026-11-14)

  • More obfuscated spellings are read: (at), [at], {at}, a bare at between address-like parts, (dot), ät, HTML entities, split or reversed script strings, and the symbol spelling used by some city sites. Ordinary prose such as "look at that" is not turned into an address.
  • Team, Kontakt, Impressum, Leitung and Ansprechpartner pages are crawled before other content pages.
  • A directory entry is followed once to the place's own site. A moved directory path gets one extra try, otherwise the row is fetch_error with a reason.
  • Placeholder addresses are dropped. Agency, footer and data protection addresses are never the primary email.
  • New per email fields kind, is_role_mailbox and confidence. place_results entries carry name and source_url.
  • Contact names are read when the name stands above the role or after "Leitung & Kontakt".

v1.9.0 (2026-11-13)

  • New contacts: named persons with their role from the Impressum, Kontakt, Team and About pages. Never guessed, data protection officers last with role dpo.
  • Place context per website (name, address, place_id) as flat fields, a place object, a places list or a top level places[] list. Sites that serve several places are crawled once and answered per place in place_results, with contact_scope and matched_branch_url.
  • New emails_detail with domain_match and type for every email.
  • An address on an unrelated foreign domain (web agency, certification body, umbrella body) is never the primary email.
  • Directory pages: with a place context the place's own outbound link is followed once and the linked site is read as the place's own site (website_type_original directory).
  • The status message adds how many rows have a contact name.

v1.8.1 (2026-11-13)

  • Hosts that do not exist or do not answer are asked once and then left alone; they no longer trigger the more capable connection levels. Guessed subdomains are only requested when their name exists.
  • The page budget is spent on pages that exist: a guessed path that returns 404 no longer uses up a slot.
  • Soft hyphens, zero-width characters and <wbr> inside an address no longer cut it into a shorter false address.
  • A mailbox on the branch page of your input URL now wins over the general mailbox on the site's contact page.
  • The run status message shows how many websites were checked and how many returned emails.

v1.8 (2026-11-13)

  • New fields: primary_email, emails_other, website_type, email_status, email_source_url, checked_at. The emails field is unchanged.
  • Branch versus parent: the URL path, and the optional name and address per company, decide which mailbox is primary. Agency, data protection and platform addresses are never primary.
  • Directory and social pages are labelled and never yield a primary_email.
  • A failed fetch is reported as fetch_error, never as not_found.
  • Written-out and encoded addresses are decoded: name(at)host(dot)de, name [at] host [dot] de, mailto: in any case and URL-encoded.
  • The legal page, contact page and footer legal link are fetched before any other page.
  • Connections: the shared connection ladder replaces the older connection logic. Each site keeps the level that worked.
  • Companies with a company_id and no website now return a no_website row.

v1.5.4 (2026-09-15)

  • JavaScript static-string extraction: recovers emails hidden in <script> blocks as concatenation, String.fromCharCode(...) sequences and atob('...') base64.
  • mailto: form action extraction.
  • Meta, Open Graph and Schema.org scan: og:email, itemprop="email", itemprop="telephone".

v1.5.3 (2026-09-15)

  • Homepage fetch retries up to 3 times on 403, 429, 503, soft blocks and connection resets.
  • Dead-domain detection marks status: failed without wasted retries.
  • Parked-domain detection returns status: no_contacts with an explanatory message.
  • www versus apex fallback before giving up.
  • pages_crawled=0 is reported as failed, not no_contacts.

v1.5.2 (2026-09-15)

  • Cloudflare data-cfemail decoder verified.
  • Subdomain probing on zero-yield apex domains (mail., contact., info., karriere.).

v1.4 (2026-09-15)

  • Optional browser fallback (enableBrowserFallback) that re-queues blocked sites on the sister browser Actor. The run ID is stored in BROWSER_FALLBACK_RUN_ID and in browser_fallback_run_id.
  • SPA state extraction (__NEXT_DATA__, __NUXT_DATA__, __APOLLO_STATE__, window.__INITIAL_STATE__).
  • Hreflang alternate-URL probing, capped at 8 URLs per site.
  • Sitemap sub-file traversal cap raised from 3 to 5.

v1.3 (2026-09-15)

  • WhatsApp, Telegram and Signal extraction.
  • JSON-LD contact extraction.
  • Soft-404 and gate-page detection.
  • marketplace_note for known marketplaces.
  • Contact paths in ES, PT, IT, NL, PL, TR and JA.

v1.2 (2026-09-15)

  • Prefetch error fallback, date-string phone filter, wider contact page discovery, sitemap fallback and malformed URL repair.

Issues and feature requests

If something does not work or a field is missing, please open an issue.

Found this Actor useful or not? A rating on its Apify Store page, with a short note on what you used it for, helps other users judge whether it fits their case.