Website Lead Intelligence

Turn any company website into actionable lead intelligence — with AI-powered industry classification, technology detection, contact extraction, company size estimation, and ICP-based lead scoring. No external API keys required.

What It Does

Give this Actor a list of company domains or URLs. It crawls each website (homepage + key subpages), then returns a rich, structured dataset for every lead:

Feature	Description
Industry Classification	Auto-classifies into 20+ industries using AI (zero-shot) or keyword analysis
Technology Detection	Identifies 50+ technologies across 8 categories (frameworks, CMS, analytics, marketing, etc.)
Contact Extraction	Finds emails, phone numbers, and social media profiles (LinkedIn, Twitter, Facebook, etc.)
Company Size Estimation	Estimates employee count from text patterns, careers pages, team pages, and tech sophistication
Lead Scoring	Scores each lead 0-100 against your Ideal Customer Profile (ICP)

How It Works

Input: ["stripe.com", "hubspot.com", "shopify.com"]
                    |
        +-----------+-----------+
        |           |           |
     Crawl       Crawl       Crawl
   (homepage    (homepage    (homepage
   + /about     + /about     + /about
   + /contact   + /contact   + /contact
   + /careers)  + /careers)  + /careers)
        |           |           |
        +-----+-----+-----+----+
              |           |
     Extract Contacts   Detect Tech Stack
              |           |
     Classify Industry  Estimate Size
              |           |
         Score vs ICP
              |
     Structured Output (sorted by score)

1. Crawl — Fetches the homepage plus up to 10 subpages (/about, /contact, /careers, /pricing, /team, etc.)
2. Extract — Pulls emails (from text + mailto: links), phones (from tel: links), and social profiles from anchor tags
3. Detect — Matches 50+ technology signatures in HTML and HTTP headers (React, WordPress, HubSpot, Stripe, AWS, etc.)
4. Classify — Determines the company's industry using a hybrid approach:
   • AI mode (default): Runs a zero-shot classification model (distilbert-base-uncased-mnli via transformers.js) locally — no API keys needed
   • Keyword mode: Fast regex-based matching against 20+ industry keyword dictionaries
   • Results are cross-validated between both methods for maximum accuracy
5. Estimate — Determines company size from employee count patterns, job listings, team profiles, and tech stack sophistication
6. Score — Calculates a 0-100 lead score based on how well the company matches your ICP criteria

Input

Field	Type	Default	Description
urls	string[]	(required)	Website URLs or domains to analyze (e.g., stripe.com, https://hubspot.com)
maxPagesPerDomain	integer	5	Max subpages to crawl per domain (1-10)
concurrency	integer	3	Number of domains to process in parallel (1-10)
enableAiClassification	boolean	true	Use AI model for industry classification. Disable for faster keyword-only mode
targetIndustries	string[]	[]	ICP: Target industries for lead scoring (e.g., ["SaaS", "E-commerce"])
targetSizeMin	integer	10	ICP: Minimum preferred employee count
targetSizeMax	integer	500	ICP: Maximum preferred employee count
targetTechnologies	string[]	[]	ICP: Target technologies (e.g., ["React", "HubSpot"])
requiredContactTypes	string[]	["email"]	ICP: Required contact types — email, phone, linkedin, twitter

Example Input

{
  "urls": ["stripe.com", "hubspot.com", "shopify.com", "coursera.org"],
  "enableAiClassification": true,
  "maxPagesPerDomain": 5,
  "concurrency": 3,
  "targetIndustries": ["SaaS and Software"],
  "targetSizeMin": 50,
  "targetSizeMax": 1000,
  "targetTechnologies": ["React", "Node.js"],
  "requiredContactTypes": ["email"]
}

Output

Each lead is returned as a structured JSON object with the following fields:

{
  "domain": "stripe.com",
  "url": "https://stripe.com",
  "title": "Stripe | Financial Infrastructure for the Internet",
  "description": "Stripe powers online and in-person payment processing...",
  "emails": ["sales@stripe.com"],
  "phones": ["+1-888-926-2289"],
  "socialLinks": {
    "linkedin": "https://www.linkedin.com/company/stripe",
    "twitter": "https://twitter.com/stripe",
    "facebook": "https://www.facebook.com/StripeHQ",
    "instagram": null,
    "github": "https://github.com/stripe",
    "youtube": null
  },
  "industry": {
    "primary": "SaaS and Software",
    "secondary": "E-commerce and Retail",
    "confidence": 0.485,
    "allScores": [
      { "label": "SaaS and Software", "score": 0.485 },
      { "label": "E-commerce and Retail", "score": 0.389 }
    ],
    "method": "keyword"
  },
  "companySize": {
    "estimate": "enterprise",
    "employeeRange": "1000+",
    "confidence": 0.25,
    "signals": ["Enterprise-level language detected"]
  },
  "technologies": ["React", "Next.js", "AWS", "Vercel", "Fastly", "Bootstrap", "Stripe"],
  "techCategories": {
    "frameworks": ["React", "Next.js"],
    "cms": [],
    "analytics": [],
    "marketing": [],
    "infrastructure": ["AWS", "Vercel", "Fastly"],
    "libraries": ["Bootstrap"],
    "ecommerce": ["Stripe"],
    "payments": []
  },
  "leadScore": 34,
  "scoreBreakdown": {
    "industryMatch": 15,
    "sizeMatch": 0,
    "techMatch": 10,
    "contactQuality": 0,
    "webPresence": 9
  },
  "pagesAnalyzed": 3,
  "analyzedAt": "2026-03-08T08:03:53.138Z"
}

Output Field Reference

Field	Type	Description
domain	string	Normalized domain name
url	string	Full URL of the homepage
title	string	Page title from <title> tag
description	string	Meta description
emails	string[]	Extracted email addresses (deduplicated, validated)
phones	string[]	Phone numbers from tel: links (max 10, deduplicated)
socialLinks	object	Social media profile URLs (LinkedIn, Twitter/X, Facebook, Instagram, GitHub, YouTube)
industry.primary	string	Top industry classification
industry.secondary	string | null	Second-best industry (if confidence > threshold)
industry.confidence	number	Classification confidence (0-1)
industry.method	string	"ai" or "keyword" — which method produced the final result
companySize.estimate	string	Size category: micro, small, medium, large, enterprise, or unknown
companySize.employeeRange	string	Human-readable range (e.g., "51-200")
companySize.signals	string[]	Evidence used for estimation
technologies	string[]	All detected technologies
techCategories	object	Technologies grouped by category
leadScore	integer	ICP match score (0-100), higher = better fit
scoreBreakdown	object	Score components: industryMatch, sizeMatch, techMatch, contactQuality, webPresence
pagesAnalyzed	integer	Number of pages successfully crawled
analyzedAt	string	ISO 8601 timestamp

Supported Industries

The classifier recognizes 20+ industries:

Industry	Example Companies
SaaS and Software	Stripe, HubSpot, GitHub, Salesforce
E-commerce and Retail	Shopify, Amazon, Etsy
Marketing and Advertising	Creative agencies, ad networks
Finance and Payments	Banks, investment platforms, fintech
Healthcare and Biotech	Hospitals, pharma, telehealth
Education and Training	Coursera, universities, edtech
Manufacturing and Industrial	Factories, industrial equipment
Consulting and Professional Services	McKinsey, Deloitte
Media and Entertainment	Streaming, publishing, podcasts
Travel & Hospitality	Airbnb, hotels, tourism
Real Estate	Property management, mortgage
Food & Beverage	Restaurants, food delivery
Telecommunications	Telecom, ISPs
Energy	Solar, oil & gas, utilities
Non-Profit	Charities, foundations, NGOs
Legal	Law firms, legal services
Construction	Contractors, architecture firms
Transportation & Logistics	Shipping, freight, fleet management
Cybersecurity	InfoSec, threat detection
AI & Machine Learning	AI research, ML platforms

Detected Technologies (50+)

Category	Technologies
Frameworks	React, Next.js, Vue.js, Nuxt.js, Angular, Svelte, Gatsby, Remix
CMS	WordPress, Shopify, Squarespace, Wix, Webflow, Drupal, Ghost, Contentful
Analytics	Google Analytics, Google Tag Manager, Facebook Pixel, Hotjar, Mixpanel, Segment, Amplitude, Plausible
Marketing	HubSpot, Mailchimp, Intercom, Drift, Zendesk, Crisp, Salesforce, Marketo, ActiveCampaign
Infrastructure	Cloudflare, AWS, Vercel, Netlify, Heroku, Google Cloud, Azure, Fastly
Libraries	jQuery, Bootstrap, Tailwind CSS, Material UI, Font Awesome, GSAP, Three.js, D3.js
E-commerce	Stripe, PayPal, WooCommerce, BigCommerce, Magento
Payments	Stripe Payments, PayPal Checkout, Square, Braintree

Lead Score Breakdown

The lead score (0-100) is composed of five weighted dimensions:

Dimension	Max Points	What It Measures
Industry Match	30	How well the company's industry matches your target industries
Size Match	20	How close the company size is to your target range
Tech Match	20	How many of your target technologies the company uses
Contact Quality	15	Availability of emails, phones, and social profiles
Web Presence	15	Tech sophistication, analytics usage, marketing tools, active website

Company Size Categories

Category	Employee Range	Midpoint
Micro	1-10	5
Small	11-50	30
Medium	51-200	125
Large	201-1000	600
Enterprise	1000+	2000

Use Cases

• Sales Prospecting — Score and prioritize leads from cold outreach lists
• Market Research — Map competitors' tech stacks and company sizes at scale
• Lead Qualification — Auto-filter leads that match your ICP before CRM import
• Data Enrichment — Append industry, size, and tech data to existing lead lists
• Competitive Intelligence — Monitor which technologies competitors are adopting

Integration Tips

Chain with Other Scrapers

Use this Actor as a post-processing step after collecting domains from:

• Google Maps Email Extractor
• LinkedIn Company Scraper
• Any domain list scraper

Export to CRM

Results can be exported as JSON, CSV, or Excel directly from the Apify dataset. Key fields map naturally to CRM fields:

• emails → Contact Email
• phones → Contact Phone
• industry.primary → Company Industry
• companySize.employeeRange → Company Size
• leadScore → Lead Score / Priority

Webhook Integration

Set up an Apify webhook to automatically send enriched leads to your CRM, Slack, or any HTTP endpoint when a run completes.

Performance

Metric	Value
Speed	~10-15 seconds per domain (with AI), ~5-8 seconds (keyword-only)
Memory	Recommended 4096 MB (for AI model). Minimum 1024 MB (keyword-only)
Concurrency	Up to 10 domains in parallel
AI Model	distilbert-base-uncased-mnli (~100MB, loaded once, runs locally)

Limitations

• Contact extraction relies on publicly visible information (emails in HTML, tel: links, social media links)
• Company size estimation is approximate — based on text patterns, careers pages, and tech sophistication
• Some websites may block automated crawlers, resulting in fewer pages analyzed
• AI classification requires 4096 MB memory; use keyword-only mode for lower memory usage

Cost

This Actor runs on the Apify platform. Costs depend on compute usage:

• Memory: 4096 MB recommended (AI mode) or 1024 MB (keyword mode)
• Compute units: ~0.01-0.02 CU per domain analyzed
• No external API keys or additional costs required