Website Contact Extractor

Extract emails, phone numbers, social profiles, and addresses from one or more websites.

Website Contact Extractor is an Apify Actor built with TypeScript and Crawlee CheerioCrawler. Point it at company homepages — it crawls the landing page plus contact/about/footer links, deduplicates contacts per domain, and writes a structured dataset plus CSV and Markdown summaries.

Features

• Email extraction — mailto: links, page text regex, schema.org; filters noise (example.com, Sentry, Wixpress, noreply, image extensions)
• Phone extraction — tel: links, international patterns, itemprop="telephone"
• Social profiles — LinkedIn, Twitter/X, Facebook, Instagram, YouTube, TikTok, GitHub
• Physical addresses — schema.org PostalAddress, microdata, footer <address> blocks
• Contact page discovery — follows contact, about, team, support, locations, impressum links
• Company name — from og:site_name, page title, or JSON-LD Organization
• Per-domain deduplication — one dataset item per domain with merged contacts
• Rate limiting & retries — configurable concurrency, delay, and request retries
• Proxy-ready — optional Apify Proxy configuration
• Exports — CONTACTS.csv and CONTACTS.md in the default key-value store

Installation

# Clone / open the Actor directory
cd website-contact-extractor

# Install dependencies
npm install

# Run locally (uses storage/key_value_stores/default/INPUT.json or --input)
apify run

On the Apify platform, deploy with apify push or use the published Actor from the Store.

Input

Field	Type	Required	Default	Description
startUrls	object[]	Yes	—	Homepage URLs to extract contacts from
maxPagesPerDomain	integer	No	10	Max pages crawled per domain
maxConcurrency	integer	No	5	Parallel request limit
maxRequestRetries	integer	No	3	Retries for failed requests
requestDelayMs	integer	No	200	Delay hint for rate limiting
followContactLinks	boolean	No	true	Crawl discovered contact/about links
includeEmails	boolean	No	true	Extract emails
includePhones	boolean	No	true	Extract phones
includeSocials	boolean	No	true	Extract social profiles
includeAddresses	boolean	No	true	Extract physical addresses
proxyConfiguration	object	No	Proxy off	Apify Proxy settings

Example input

{
    "startUrls": [
        { "url": "https://www.apify.com" },
        { "url": "https://www.shopify.com" },
        { "url": "https://stripe.com" }
    ],
    "maxPagesPerDomain": 5,
    "maxConcurrency": 5,
    "requestDelayMs": 200,
    "followContactLinks": true,
    "includeEmails": true,
    "includePhones": true,
    "includeSocials": true,
    "includeAddresses": true
}

See also: examples/input.json

Output

Each dataset item is one domain:

Field	Type	Description
domain	string	Normalized host without www
url	string	Start URL for the domain
companyName	string | null	Detected company/site name
emails	string[]	Deduplicated emails
phones	string[]	Deduplicated phones
socials	object	Map of platform → profile URL
addresses	string[]	Physical addresses when found
contactPageUrls	string[]	Discovered contact/about URLs
pagesScraped	integer	Pages successfully scraped
scrapedAt	string	ISO timestamp
error	string | null	Present if the domain failed entirely

Example output

{
    "domain": "apify.com",
    "url": "https://www.apify.com",
    "companyName": "Apify",
    "emails": ["hello@apify.com"],
    "phones": [],
    "socials": {
        "linkedin": "https://www.linkedin.com/company/apifytech",
        "twitter": "https://twitter.com/apify",
        "github": "https://github.com/apify"
    },
    "addresses": [],
    "contactPageUrls": ["https://apify.com/contact"],
    "pagesScraped": 3,
    "scrapedAt": "2026-07-13T12:00:00.000Z"
}

Key-value store exports

Key	Content type	Description
CONTACTS.csv	text/csv	Flat CSV summary of all domains
CONTACTS.md	text/markdown	Human-readable Markdown report
OUTPUT	application/json	Run summary stats

See also: examples/output.json

How it works

1. Start URLs — Each unique domain is queued once (first URL wins as the primary).
2. Homepage scrape — Cheerio parses the page for emails, phones, socials, addresses, and company name.
3. Contact link discovery — Same-domain links matching contact/about/team/support/locations patterns are enqueued (up to maxPagesPerDomain).
4. Merge & dedupe — Contacts from all pages are merged per domain.
5. Export — One dataset item per domain, plus CONTACTS.csv and CONTACTS.md.

Pricing

$0.003 per domain result using Apify pay-per-event (apify-default-dataset-item). Charged once per domain written to the dataset.

Run locally

npm install
npm run build
npm test
apify run -p -i examples/input.json

Limitations

• Static HTML only — Uses Cheerio (no browser). Contacts injected purely via client-side JavaScript may be missed.
• Obfuscated emails — Cloud-encoded or image-only emails are not decoded.
• Address detection — Best with schema.org / microdata; free-form footer text is heuristic.
• Phone false positives — Aggressive phone regex can occasionally match non-phone number sequences; noise filters reduce this.
• Same-domain only — Contact links on other domains are not followed.
• Rate limits — Aggressive settings may trigger site WAF/bot protection; enable Apify Proxy for production.

FAQ

How many pages are crawled per site?
Up to maxPagesPerDomain (default 10): the start URL plus discovered contact/about links.

Can I extract only emails?
Yes. Set includePhones, includeSocials, and includeAddresses to false.

Does it work without a proxy?
Yes for small local tests. For bulk production runs, enable proxyConfiguration.useApifyProxy.

Where is the CSV?
In the run’s default key-value store under key CONTACTS.csv.

One result or many?
One dataset item per unique domain, regardless of how many pages were scraped on that domain.

Changelog

1.0.0

• Initial release
• Email, phone, social, address extraction with CheerioCrawler
• Contact/about link following and per-domain deduplication
• CSV and Markdown exports
• Input flags for selective extraction
• LEAD_GENERATION + DEVELOPER_TOOLS categories

License

Apache-2.0