Website Contact Scraper & Lead Enrichment

Turn a list of company websites into structured contact records for your CRM, lead research, or sales workflow. Find public email addresses, phone numbers, company social profiles, contact forms, website technologies, and careers links. Every email and phone includes the page where it was found.

$5 per 1,000 enriched websites, plus $0.001 per run start. An enriched website has at least one public email, phone, supported social link, or contact form. Up to six pages can be included in one website result. Invalid inputs, duplicate domains within a run, inaccessible websites, and websites with no contacts produce no result charge. The start fee still applies. Platform usage is included in the listed customer price.

Start in under a minute

1. Paste company domains or website URLs into Websites.
2. Set Maximum websites to attempt and, if needed, the phone country.
3. Run the Actor. Open the Company contacts dataset view or download LEADS_CSV.
4. Review UNRESOLVED for websites that did not produce a billable record and REPORT for completion details.

You do not need an external API key, proxy subscription, browser cookies, or an email account. This Actor reads publicly accessible web pages. It does not send emails, submit forms, log in to websites, or guess unpublished email addresses.

Why use company-level enrichment?

Many lead lists already contain a website but lack a usable contact route. A generic homepage scraper can miss the contact or imprint page, return the same address repeatedly, or separate one company's information across many rows. This Actor prioritizes contact, about, imprint, support, team, and careers links, then combines the collected information into one company record.

The unit you pay for is an enriched website, rather than every visited page or every email. By default the crawler attempts four pages per website. You can increase that to six without changing the result price. A record can contain multiple emails, multiple phones, several social profiles, and technology evidence. The output preserves source row numbers so it can be joined back to an existing list.

Input options

Provide exactly one input source. Clear example URLs before selecting CSV or dataset input.

Paste websites

{
  "websites": ["https://www.python.org/", "https://www.scrapy.org/"],
  "maxWebsites": 100,
  "maxPagesPerWebsite": 4,
  "phoneRegion": "US",
  "checkEmailDns": true
}

Both bare domains and complete HTTP or HTTPS URLs are accepted. The initial path is preserved, which is useful when you already know a contact page. URLs for the same registered company domain are combined within the run. For example, a homepage and its www contact page produce one record. Separately hosted businesses on supported private suffixes, such as different github.io sites, remain separate.

CSV input

Paste comma, semicolon, or tab-separated CSV into CSV text. Choose the exact top-level column name with Website column / field. The default is website. CSV input is limited to 4 MB. The output includes sourceRow, a one-based data-row index excluding the CSV header. Duplicate rows appear in duplicateSourceRows on the company's record.

{
  "csvText": "website,company\nhttps://www.python.org/,Python\nhttps://www.scrapy.org/,Scrapy",
  "websiteField": "website"
}

Existing Apify dataset

Select an accessible source dataset containing website URLs. For example, a Google Maps business dataset commonly provides a website field. This Actor enriches the company websites in that dataset; it does not scrape Google Maps itself. Select the resource in the input form to grant read-only access to that dataset. The source is never modified.

Use websiteField to select a different top-level field, such as url or domain. Nested property paths are not supported. Output source row numbers refer to dataset item order. Keep the original dataset if you need to join other columns later.

Output you can use immediately

Each result includes:

Field	Meaning
recordId	Stable identifier derived from the company domain
domain, website	Company domain and fetched entry URL
companyName, companyNameSource	Name from business structured data, published site name, or a page-title fallback
description	Published page description, when present
bestEmail	A selected published address on the company domain, prioritizing sales and general contact mailboxes
emails	Addresses, source pages, role flags, same-company-domain flags, and DNS status
phone, phones	Parsed phone numbers in international format with source pages
socials, linkedinUrl	Social links published by the website; LinkedIn output is restricted to company, school, and showcase pages
contactPages, contactForms	Discovered contact pages and forms; forms are never submitted
technologies	Recognized markers, categories, source URLs, and matching evidence
hiringLinks	Published careers or supported ATS links
addresses	Business addresses explicitly present in structured data
pagesScraped, sourceUrls, warnings	Coverage and any failed subpages
sourceRow, duplicateSourceRows	References back to the original input
partial, checkedAt	Incomplete subpage coverage and extraction timestamp

The Company contacts view is convenient for CRM imports. Technology & hiring helps segment websites by observed software and careers links. Coverage & evidence helps audit where the information came from. Full nested JSON remains available through the dataset API. A flat, formula-protected CSV is saved as LEADS_CSV in the run's key-value store.

Email quality and what DNS means

Addresses are extracted from published text, mailto links, and business structured data. Common [at] and [dot] text obfuscation is normalized. Obvious placeholders and asset-like strings are excluded. Emails inside arbitrary script code are not harvested. Addresses from outside the company's domain remain clearly marked.

Optional DNS checks examine MX records for up to five distinct email domains per website. DNS requests send domain names to Google Public DNS; full email addresses are not sent. mx_present means a mail route exists. It does not prove that a particular mailbox exists or will accept a message. no_mx_record is also not a definitive invalid-mailbox result because some domains use address-record fallback. unknown means the DNS lookup could not establish a result.

mailboxStatus is always not_checked. This Actor does not perform SMTP verification, catch-all testing, deliverability guarantees, or message sending. bestEmail only selects addresses on the same registered company domain and is null when none qualifies. It is a deterministic ordering of published addresses, not a verified decision-maker identity or a promised conversion.

Technology and hiring evidence

The Actor recognizes a curated set of common CMS, ecommerce, analytics, marketing, support, payment, and framework markers. Examples include WordPress, Shopify, WooCommerce, Webflow, HubSpot, Klaviyo, Intercom, Google Analytics, Stripe, and Next.js. This is a focused detector, not an exhaustive technology census. Missing a marker does not establish that a company does not use that product.

Each detection includes a source page and matching marker. A linked provider or embedded script is evidence of that web integration, not proof of an active commercial subscription. Careers links recognize providers including Greenhouse, Lever, Ashby, Workable, SmartRecruiters, Workday, and BambooHR. A careers link does not establish that the company is currently hiring. No job counts or buying-intent scores are invented.

Limits, completion, and predictable charges

The defaults attempt up to 100 unique websites, four pages each, with eight websites processed concurrently. Maximum configurable limits are 2,000 websites, six pages per website, 10,000 input rows, and 480 seconds of processing. Large lists can therefore require multiple runs. The overall Apify run timeout should exceed the processing budget to allow exports to finish.

The Actor checks the customer's remaining charge allowance before starting another batch. Set Apify's maximum run charge to control spending. maxWebsites limits attempted websites, including unsuccessful attempts; it is not a promise of that many delivered results. Large pages, long response times, restrictive crawl delays, and unreachable sites are stopped within bounded resource limits.

REPORT.complete and REPORT.stopped describe whether the input was finished or processing stopped because of a limit. UNRESOLVED lists invalid inputs, inaccessible websites, and no-contact outcomes without charging a result fee. A partial record with useful contacts is billable and includes warnings about subpages that failed. Social-only and contact-form-only records are billable even when no email is found. Input duplicates are free within a run; running the same company again in a later run performs a fresh enrichment and is billed normally.

Coverage and responsible use

The crawler reads HTML returned by public websites and respects robots.txt. It does not execute JavaScript, bypass login walls, solve CAPTCHA challenges, rotate proxies to evade blocks, or crawl unrelated external domains. JavaScript-only contact pages and protected websites may be unresolved. Redirects are restricted to the same registered company domain. IP-address destinations and private networks are rejected.

Phone parsing uses the configured country for numbers without an international prefix. Choose the correct country for local business lists; use separate runs for mixed regions when local number formats are ambiguous. A published social link does not by itself prove ownership. Website contact details can become outdated, and automatically selected company names may require review.

Only collect and use information you are authorized to process. The Actor provides data collection and enrichment; your downstream use, retention, and outreach decisions remain yours. There is no built-in bulk messaging or external contact-data upload. Results are stored in your Apify run storage.

Automation and support

Run through the Apify API, use a task for a saved configuration, or connect the resulting dataset to your existing n8n, Make, or CRM workflow. Apify schedules can repeat enrichment when you need fresher information. Scheduling does not discover new companies automatically: supply the website list or dataset you want processed.

If an expected published contact is missing, open an Actor issue with the public website URL, the relevant source page, and the field you expected. Avoid sharing credentials or private customer lists in public issues. Reports that identify a reproducible parsing problem are the most useful way to improve coverage.