# LinkedIn Lead Scraper

**Skill:** `/scrape-leads <keyword>`
**Type:** Python CLI tool
**API:** RapidAPI `fresh-linkedin-profile-data`

---

## What It Does

Finds qualified leads from LinkedIn post interactions. Given a keyword, it searches for recent LinkedIn posts, extracts everyone who commented or reacted, then runs each person through a triple-gate filtering system to identify decision-makers worth reaching out to.

The output is a CSV of qualified leads with full profile data, ready to import into a CRM or outbound tool.

---

## How It Works

```
Keyword (e.g. "revenue operations")
        │
        ▼
┌─────────────────────┐
│  Search LinkedIn     │
│  Posts (last N days) │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Extract Commenters  │
│  & Reactors          │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────────────────────────────────┐
│           TRIPLE-GATE FILTERING                  │
│                                                  │
│  Gate 1: Negative Title Filter                   │
│  ├── Exclude: SDR, Intern, Student, Recruiter,  │
│  │   Designer, Developer, Support, Freelance     │
│  └── If title contains any → REJECT              │
│                                                  │
│  Gate 2: Positive Title Filter                   │
│  ├── Include: CEO, CRO, VP Sales, Director,     │
│  │   Head of Growth, Founder, RevOps             │
│  └── If title matches none → REJECT              │
│                                                  │
│  Gate 3: Firmographic Filter                     │
│  ├── Company size >= 10 employees                │
│  ├── Exclude: staffing, recruiting, non-profit   │
│  └── If fails either check → REJECT              │
│                                                  │
└──────────────────────┬──────────────────────────┘
                       │
                       ▼
              Qualified Leads CSV
```

---

## Usage

```bash
# Basic: search for keyword, process last 7 days of posts
python main.py "revenue operations"

# Look back 30 days, process up to 100 posts
python main.py "sales enablement" --days 30 --max-posts 100

# Skip reactions (faster, comments only)
python main.py "outbound sales" --no-reactions

# Process a specific post URL
python main.py "keyword" --post-url "https://www.linkedin.com/feed/update/urn:li:activity:123456"

# Verbose logging
python main.py "GTM strategy" -v
```

---

## Triple-Gate Filtering Logic

### Gate 1: Negative Title Filter (Exclusion)

Catches common non-decision-maker roles before any further processing. If a person's title contains any of these terms, they are immediately rejected.

| Category | Terms |
|----------|-------|
| Junior roles | intern, student, trainee, junior, assistant |
| Outbound roles | sdr, bdr, sales development |
| Support roles | recruiter, talent acquisition, hr specialist, support, customer service, helpdesk |
| Technical roles | developer, engineer, software, backend, frontend |
| Creative roles | designer, graphic, ux, ui |
| Independent | freelance, consultant |

### Gate 2: Positive Title Filter (Inclusion)

Only people with decision-maker titles pass through. The title must contain at least one of these terms.

| Level | Terms |
|-------|-------|
| C-Suite | ceo, coo, cro, cmo, cto, cfo, chief executive, chief operating, chief revenue, chief marketing, chief technology |
| VP Level | vp, vice president, vp sales, vp marketing, vp growth, vp revenue |
| Director Level | director of sales, director of marketing, director of growth, director of revenue |
| Head Level | head of sales, head of marketing, head of growth, head of revenue, head of demand gen |
| Founders | founder, co-founder, owner, partner |
| Revenue Ops | revops, revenue operations, sales operations, marketing operations, growth operations |

### Gate 3: Firmographic Filter

Company-level checks to ensure the lead works at a viable target account.

- **Minimum company size:** 10 employees
- **Excluded industries:** staffing, recruiting, human resources, education, non-profit, government

---

## Output Format

The CSV contains the following columns:

| Column | Example |
|--------|---------|
| `name` | Jane Smith |
| `title` | VP of Revenue Operations |
| `company` | Acme Corp |
| `linkedin_url` | https://linkedin.com/in/janesmith |
| `employee_count` | 250 |
| `industry` | SaaS |
| `location` | San Francisco, CA |
| `interaction_type` | comment |
| `post_url` | https://linkedin.com/feed/update/... |
| `qualified` | True |
| `gate_passed` | all |

---

## Sample Run Output

```
2026-02-20 14:32:01 - INFO - Searching for 'revenue operations' posts from last 7 days
2026-02-20 14:32:05 - INFO - Found 34 posts matching keyword
2026-02-20 14:32:45 - INFO - Extracted 412 unique profiles from comments and reactions
2026-02-20 14:33:12 - INFO - Filtering: 412 total → 47 qualified (G1 rejected: 189, G2 rejected: 152, G3 rejected: 24)
2026-02-20 14:33:12 - INFO - Saved to data/runs/revenue_operations_20260220_143312.csv
```

---

## Configuration

Set these in `.env` or `config.yaml`:

```
RAPIDAPI_KEY=your_key_here
RAPIDAPI_HOST=fresh-linkedin-profile-data.p.rapidapi.com
```

---

## Key Files

| File | Purpose |
|------|---------|
| `main.py` | CLI entry point — keyword search, post processing, CSV export |
| `filters.py` | Triple-gate filtering system with configurable title lists |
| `linkedin_scraper/scraper.py` | Core scraper logic (API calls, deduplication) |
| `linkedin_scraper/config.py` | Configuration loader (API keys, filter lists) |