Back to Templates

Qualify and draft LinkedIn outreach sequences for Airtable leads with Claude

Created by

Created by: Allan Vaccarizi || growthai
Allan Vaccarizi

Last update

Last update 2 days ago

Categories

Share

⚠️ Self-hosted only — This template uses a community node (@mendable/n8n-nodes-firecrawl) and cannot run on n8n Cloud.

Who it's for

This workflow is for B2B sales teams and growth operators who want to automatically enrich, qualify, and write personalized LinkedIn outreach messages for a list of leads stored in Airtable — without touching the workflow logic to change targeting criteria.

How it works

  1. A manual trigger sets the Airtable base and table IDs, then fetches the AI prompt used for qualification.
  2. Leads are retrieved from Airtable and fed into a batch loop, one at a time.
  3. For each lead, Apify scrapes their LinkedIn profile. Unavailable profiles are flagged in Airtable and skipped.
  4. If a company LinkedIn URL exists, Apify fetches company info. If a website URL is present, Firecrawl scrapes the homepage and a code node cleans the resulting markdown.
  5. LinkedIn posts are fetched via Apify, sorted, and consolidated with profile fields into a structured dataset.
  6. A Claude AI agent evaluates the lead against the qualification prompt stored in Airtable. Unqualified leads are flagged and the loop continues.
  7. For qualified leads, a second Claude AI agent drafts a three-part personalized outreach message sequence, which is saved back to Airtable.

How to set up

  • [ ] Connect your Airtable credentials and set the correct Base ID and Table ID in the Set Airtable Base and Table node
  • [ ] Create Airtable tables for prompts, leads, and message storage with the expected field names
  • [ ] Add your Apify API key and configure the three actor calls (LinkedIn profile, company info, posts)
  • [ ] Add your Firecrawl API key for the Scrape Company Homepage node
  • [ ] Add Anthropic API credentials for both Claude AI agent nodes
  • [ ] Review and update the structured output schemas in both Structured Output Parser nodes

Requirements

  • Airtable account with leads table and prompt table
  • Apify account (three actors: LinkedIn profile scraper, company scraper, post scraper)
  • Firecrawl account
  • Anthropic (Claude) API access
  • n8n self-hosted instance

How to customize

  • Update the qualification and message-drafting prompts directly in Airtable — no workflow changes needed to adjust targeting logic or tone.
  • Add extra enrichment steps (e.g., email finder, firmographic APIs) between the company URL check and the LinkedIn post fetch.
  • Swap the manual trigger for a scheduled trigger to process new leads automatically on a daily or weekly cadence.