AI-personalized cold outbound at 5x reply rate
Run 500-lead cold campaigns weekly with personalized first lines based on each lead's LinkedIn, hitting 5-8% reply rates instead of 1-2%.
Tools required
- Apollo / Lemlist
Lead sourcing + sequence sending
- Apify ↗
LinkedIn scraping
- Claude or GPT-4o
Personalized first-line generation
- Google Sheets
Mid-pipeline review
Steps
- 1
Build the list
Export 500 ICP leads from Apollo. Include LinkedIn URL.
- 2
Scrape LinkedIn About + last 3 posts
Use Apify's LinkedIn scraper. Pull About, headline, last 3 post titles.
- 3
Generate personalized openers
Send each row to GPT-4o with a strict prompt: 'In 1 sentence, react to one specific thing about this person. No flattery, no compliments.'
- 4
Review 10% sample by hand
Spot-check 50 rows. Keep tone consistent.
- 5
Load into sequence tool
Personalized line goes in {{custom1}}. Rest of email is templated.
- 6
Send 50/day per inbox
Stay under deliverability thresholds. Use multiple inboxes if scaling.
Expected results
Reply rates from 1-2% baseline → 5-8%. For a B2B SaaS, that's a 3-4x increase in pipeline per dollar.