Business Opportunity Finder — AI Lead Generation

Find businesses that are likely to need your service, understand why they need it, and get the exact information to contact them.

Business Opportunity Finder is a production-ready Apify Actor that transforms generic business searches into qualified, pitch-ready B2B sales leads.

Instead of outputting raw, unstructured directories, it autonomously discovers local businesses, inspects their websites, analyzes their digital tech stacks, extracts verified public contact details, calculates explainable opportunity and data quality scores, and synthesizes tailored sales pitches with AI.

---

🌟 Why This Actor Is Different

Most business scrapers merely dump raw phone numbers and addresses into a spreadsheet. Sales teams and agency owners are still forced to manually click through each website, spot broken designs or missing features, and write pitches by hand.

This Actor automates the complete qualification pipeline:

BUSINESS SEARCH
      ↓
BUSINESS METADATA (Name, Location, Phone, Reviews)
      ↓
WEBSITE DISCOVERY & HTTP INSPECTION
      ↓
TECH STACK DETECTION (CMS, Ecommerce, Booking, Analytics, Live Chat)
      ↓
OPPORTUNITY SIGNAL DETECTION (No SSL, Archaic Code, Broken Mobile, Missing Booking/Ordering)
      ↓
PUBLIC CONTACT ENRICHMENT (Emails, Direct Phones, Contact Form, Social Media Profiles)
      ↓
DATA QUALITY & OPPORTUNITY SCORING (Distinct 0–100 Scores)
      ↓
AI SALES QUALIFICATION & CUSTOM OUTREACH PITCH
      ↓
READY-TO-USE SALES LEAD DATA (JSON / CSV / CRM Ready)

---

👥 Who Should Use It?

1. Web Development Agencies & Freelancers: Target businesses with outdated websites, no SSL, non-responsive mobile designs, or no online presence at all.
2. SEO Specialists & Marketers: Find local businesses with missing title tags, absent meta descriptions, no H1 structure, or weak organic presence.
3. AI Automation Agencies (AAA): Identify companies lacking 24/7 customer chat, automated booking systems, or automated inquiry workflows.
4. B2B Appointment & Booking Integrators: Discover restaurants lacking online reservations or clinics without digital scheduling.
5. E-Commerce & Digital Ordering Consultants: Pitch turnkey Shopify, WooCommerce, or pickup/delivery ordering systems to retail shops and restaurants.
6. Lead Generation & Cold Outreach Teams: Feed rich, personalized lead intelligence into cold email sequences and CRM pipelines (HubSpot, Close, Instantly, Smartlead).

---

🚀 Example Use Cases

• "Restaurants in London needing Web Development & Online Booking"
  • Searches for restaurants in London, checks whether they have modern responsive websites, flags outdated copyright dates (e.g., 2018), detects lack of OpenTable/Resy/SevenRooms, extracts contact email (bookings@...), and drafts a tailored pitch.
• "Dentists & Clinics in Manchester needing AI Chat & Automated Scheduling"
  • Identifies clinics without Calendly/Acuity or live chat assistants, providing direct cold outreach copy offering an automated front-desk assistant.
• "Retail Boutiques in Berlin needing E-Commerce Storefronts"
  • Flags brick-and-mortar merchants with no shopping cart or checkout integration.

---

⚙️ Input Parameters

The Actor is configured through an intuitive UI in the Apify Console.

Field	Type	Default	Description
searchQueries	Array of strings	["restaurants"]	Business categories or keywords (e.g. restaurants, dentists, lawyers).
locations	Array of strings	["London, UK"]	Geographic targets (e.g. London, UK, Manchester, UK, Berlin, Germany).
country	String	United Kingdom	Country filter to focus geographic discovery.
serviceToSell	Select / String	web development	The service you are pitching: web development, SEO, AI automation, digital marketing, mobile app development, online booking integration, ecommerce solutions, reputation management.
maxResults	Integer	50	Maximum number of qualified leads to produce (1 to 500).
analyzeWebsite	Boolean	true	When true, fetches and parses the business website.
findContacts	Boolean	true	Extracts public emails, direct telephone numbers, and contact page links.
aiQualification	Boolean	true	Uses AI to synthesize findings and generate tailored cold outreach pitches.
includeSocialProfiles	Boolean	true	Discovers LinkedIn, Instagram, Facebook, and X/Twitter profiles.
minimumRating	Number	0.0	Minimum customer review rating filter (0.0 to 5.0).
minimumReviews	Integer	0	Minimum number of customer reviews.
crawlDepth	Integer	2	1 inspects the homepage; 2 also crawls the /contact or /about page to uncover hidden emails.
concurrency	Integer	5	Concurrent website analysis workers (1 to 20).
language	String	"en"	Language code for search results and pitches.
discoverySource	Select	"osm"	Discovery engine: osm (OpenStreetMap Overpass - built-in, free, zero keys needed), apify_google_places (delegates to an Apify Google Maps Actor), or direct_seeds.
aiApiKey	Secret String	null	API key for AI qualification (supports OpenAI, Gemini, Groq, Anthropic, OpenRouter, DeepSeek). Can also be set via AI_API_KEY env var.
aiProvider	Select	"openai"	AI provider (openai, gemini, anthropic, groq, openrouter, deepseek).
aiModel	String	"gpt-4o-mini"	Model identifier (e.g., gpt-4o-mini, gemini-2.0-flash, claude-3-5-haiku-20241022).

---

📊 Output Schema

Each item pushed to the Apify dataset contains:

1. Business Information

• business_name: Commercial name.
• category: Business niche or category.
• address: Physical street address.
• city, country, postal_code: Location details.
• phone: Primary telephone number.
• website: Resolved website URL.
• business_rating: Average customer review rating.
• review_count: Total customer reviews.
• source_url: Discovery source profile URL.

2. Website Information

• website_found: Boolean indicating website existence.
• website_url: Final redirected homepage URL.
• website_status: HTTP response status code (e.g. 200, 404).
• ssl_enabled: HTTPS security certificate status.
• mobile_friendly: Mobile viewport tag validation.
• website_title: HTML <title> tag.
• website_description: Meta description content.
• technology_stack: Array of all detected tools and libraries.
• cms: Detected CMS (e.g. WordPress, Shopify, Wix, Squarespace, Webflow).
• ecommerce_platform: Platform detected (e.g. Shopify, WooCommerce, Magento).
• analytics: Array of analytics suites (Google Analytics 4, Meta Pixel, Hotjar, etc.).
• booking_system: Detected booking tool (Calendly, OpenTable, Resy, Fresha, etc.).
• payment_system: Payment gateways (Stripe, PayPal, Square, Klarna, etc.).
• social_links: Key-value map of LinkedIn, Instagram, Facebook, and Twitter/X profiles.

3. Contact Details

• email: Primary verified business contact email.
• contact_page: URL of the contact or inquiry form.
• linkedin, instagram, facebook, twitter_x: Direct social profile links.
• other_public_contact_information: Complete list of secondary emails and alternate numbers.

4. Opportunity & AI Analysis

• opportunity_score: Normalized score (0 to 100) reflecting urgency of service need.
• opportunity_level: Categorical rating (low, medium, high).
• data_quality_score: Record completeness score (0 to 100).
• data_quality: Completeness category (low, medium, high).
• why_this_is_a_lead: Factual explanation of why this business represents an immediate opportunity.
• detected_problems: Concrete observed issues (e.g. ["Missing HTTPS / SSL", "Outdated website (copyright 2019)", "No online reservation system"]).
• recommended_service: Tailored service offering to pitch.
• recommended_pitch: Ready-to-send personalized outreach opening.
• confidence: Confidence score (0.0 to 1.0).
• ai_summary: Optional executive summary.
• signals: Detailed breakdown of individual weighted signals.

5. Metadata

• source: Discovery source engine (osm, apify_google_places, direct_seeds).
• scraped_at: ISO 8601 UTC timestamp.
• errors: Non-fatal warning messages encountered during enrichment.

---

🎯 How Scoring Works

A. Business Opportunity Score (0–100)

The opportunity score measures need and sales urgency:

• No Website: +35 points (Prime candidate for a ground-up build)
• No SSL / Insecure HTTP: +25 points (Urgent security risk)
• Not Mobile-Friendly: +25 points (Fails Google mobile index)
• Outdated Codebase (Copyright ≤ 2021 / Archaic tags): +20 points
• No Online Booking (Appointment businesses): +25 points
• No Table Booking or Online Ordering (Restaurants): +25 points
• No E-Commerce Storefront (Retail merchants): +20 points
• Missing AI / Live Chat (When pitching AI automation): +20 points
• Weak SEO (Missing title/description/H1): +20 points
• Slow Response Latency (> 2.5s): +15 points
• Missing Social Presence: +15 points
• Sub-optimal Customer Rating (< 3.8/5.0): +20 points

B. Data Quality Score (0–100)

Calculated completely independently from Opportunity:

• Business Name present: +20 pts
• Street Address / City present: +15 pts
• Direct Phone Number verified: +20 pts
• Working Website (HTTP 200): +15 pts
• Business Email verified: +20 pts
• Social Media profile verified: +10 pts

Example: A business with an address, verified phone, and verified email has a Data Quality Score of 95, but if their website is modern and fast, their Opportunity Score might be 20. You get clean data without false positives!

---

🤖 AI Qualification & Strict Factual Grounding

When aiQualification: true is enabled and an API key is provided, the Actor invokes an LLM to evaluate the evidence:

• Strict Factual Guardrails: The prompt explicitly enforces:
  • Never invent or hallucinate facts.
  • Only use verified evidence supplied in the scrape.
  • Distinguish observed facts from recommendations.
  • If evidence is insufficient, explicitly state so.
• Multi-Provider Support: Supports OpenAI (gpt-4o-mini), Google Gemini (gemini-2.0-flash), Anthropic Claude (claude-3-5-haiku-20241022), Groq, OpenRouter, and DeepSeek.
• Fail-Safe Fallback: If an AI request times out or is rate limited, the Actor automatically falls back to deterministic rule scoring without interrupting the run.

---

💰 Cost Considerations & Monetization

Compute Cost

• Efficient async HTTP crawling (using httpx + BeautifulSoup) consumes minimal memory (~256 MB to 512 MB).
• No heavy headless browsers are spun up for standard HTTP inspection.
• Total compute cost per 1,000 leads: ~$0.10 - $0.25.

AI Cost

• Optimized prompts utilizing ultra-efficient models like gpt-4o-mini or gemini-2.0-flash cost approximately $0.0003 per lead (~$0.30 per 1,000 leads).

Apify Pay-Per-Event (PPE) Readiness

• The Actor is instrumented with Actor.charge(event_name="qualified-lead", count=1).
• You only charge users for successfully produced qualified leads.
• Suggested pricing model on Apify Store: $0.03 – $0.08 per qualified lead.

---

🔒 Responsible Scraping & Security

• Zero Hard-coded Secrets: All API keys and tokens are loaded strictly from environment variables or encrypted Actor secrets.
• Public Data Only: Only publicly displayed business information, canonical websites, and public contact pages are crawled.
• Rate-Limiting & Backoff: Requests employ randomized delays, exponential backoff, and respectful user agents.

---

🛠️ Local Development & Testing

1. Setup Virtual Environment

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
pip install pytest pytest-asyncio

2. Run Test Suite

$pytest tests -v

3. Run Actor Locally

$apify run

---

📄 License

Apache-2.0 License.