LLM Lead Classification

Goal

Classify leads using Claude for complex distinctions that keyword matching can't handle.

When to Use

• Product SaaS vs IT consulting agencies
• High-ticket vs low-ticket businesses
• Subscription vs one-time payment models
• NOT for simple categories (dentists, realtors)

Scripts

• ./scripts/classify_leads_llm.py - Main classification script
• ./scripts/update_sheet.py - Update sheets
• ./scripts/read_sheet.py - Read from sheets

Usage

python3 ./scripts/classify_leads_llm.py .tmp/leads.json \
  --classification_type product_saas \
  --output .tmp/classified_leads.json

Performance

• ~2 minutes for 3,000 leads
• ~$0.30 per 1,000 leads
• Default: includes "unclear" classifications (medium confidence)

Classification Types

• product_saas: Product companies vs service/consulting
• Custom types can be added

Output

JSON file with classification added to each lead record.

---

Schema

Inputs

Name	Type	Required	Description
leads_file	file_path	Yes	JSON file with lead data
classification_type	string	Yes	Classification type (e.g., 'product_saas')
output	file_path	No	Output JSON file path

Outputs

Name	Type	Description
classified_file	file_path	JSON file with classification added to each lead

Credentials

Name	Source
ANTHROPIC_API_KEY	.env

Composable With

Skills that chain well with this one: scrape-leads, gmaps-leads, instantly-campaigns

Cost

$0.30 per 1,000 leads