Two lead generation systems (technical breakdown)
Ghiles Moussaoui
July 26, 2025
I've been testing two lead generation systems that solve different problems.
The first finds and analyzes businesses at scale. The second does personalized LinkedIn outreach with higher response rates than anything I've seen.
Here's exactly how they work and what they cost to run.
System 1: Business Discovery & Intelligence
This system maps markets and extracts business intelligence automatically.
Technical stack:
n8n for workflow orchestration (4 separate workflows)
SERP API for Google Maps business discovery
OpenAI GPT-4o for website analysis
Custom scrapers for contact extraction
FindyMail for email verification
Process flow:
Business Discovery: SERP API queries Google Maps with search terms like "accounting firms Chicago" and returns 200-500 businesses with basic info (name, address, website, phone).
Website Analysis: Each website gets crawled by AI agents that extract:
Company description and services
Team size and key personnel
Recent news and updates
Technology stack indicators
Contact information
Social Media Discovery: Separate crawler finds LinkedIn, Twitter, Facebook profiles by analyzing website source code and common URL patterns.
Data Verification: FindyMail verifies email addresses. Invalid emails get filtered out.
Output: Clean dataset in Google Sheets with 15-20 data points per company.
Costs (monthly for 1,000+ companies):
SERP API: $3 (Serper at $0.30/1K searches)
OpenAI: $15-25 (depends on website complexity)
FindyMail: $49
Total: ~$70/month
Performance:
Processes 200-400 companies/day
70% email discovery rate
85% LinkedIn profile discovery rate
3-4 minutes per company
This works well for market research, competitive analysis, or building prospect databases in industries where LinkedIn targeting is limited.
System 2: LinkedIn Outreach Automation (n8n self-hosted)
This system finds LinkedIn prospects and sends personalized emails based on company research.
Technical stack:
Apollo.io for prospect identification
Apify for LinkedIn data scraping
Tavily API for company research
OpenAI GPT-4.1 for message personalization
Instantly.ai for email delivery
Process flow:
Prospect Discovery: Form input (job titles, company size, keywords, location) generates Apollo search URL. Apify executes the search and returns 500 prospects with verified emails.
Company Research: For each prospect, Tavily searches for:
Company background and mission
Recent news and press releases
Product offerings and services
Customer reviews and sentiment
Message Generation: GPT-4.1 analyzes prospect's role + company research and writes personalized email (100-150 words) that references specific company details.
Campaign Deployment: Lead gets added to Instantly.ai campaign with the custom message as a variable.
Costs (monthly for 1,000 personalized emails):
Apollo.io: $49 (Basic plan)
Apify: $49 (Starter plan)
Tavily: $8 (1,000 searches at $0.008 each)
OpenAI: $3-5 (message generation)
Instantly.ai: $47 (Growth plan)
Total: ~$150/month, can be run for $100
Performance:
15-25% response rates (vs 2-5% for generic cold email)
3-8% meeting booking rate
500 prospects researched and messaged in 2-3 hours
95% inbox delivery rate
The personalization quality is what drives the response rates. Instead of "I saw your company online," messages reference specific company initiatives, recent news, or business challenges.
Implementation Notes
Both systems require technical setup:
API key configuration for 5-6 services
n8n workflow imports and customization
Google Sheets template setup
Error handling and monitoring
The workflows handle rate limiting, data validation, and failure recovery automatically.
System 1 is better for broad market analysis. System 2 is better for targeted sales outreach.
Most B2B companies benefit from running both - use System 1 to map your market, then System 2 to convert high-value prospects.
Want the complete implementation? Get the files and the written setup guides for both systems:
