Wanted to share a workflow I built for automating local lead generation, in case it’s useful to others here or sparks ideas.
The problem: I was manually searching Google Maps for local businesses in a given niche, then copy-pasting details into a spreadsheet — slow and error-prone.
What I built:
**•** Manual (or scheduled) trigger kicks it off
**•** SerpAPI node searches Google Maps for a given query (e.g. “real estate agents in Calgary, AB”)
**•** A Code node flattens/cleans the raw results into a consistent structure
**•** Loops through each lead
**•** Each one gets sent to Claude, which scores it 1–10 for quality and writes a short custom outreach angle
**•** Everything gets written to Airtable as a new CRM row — name, address, phone, website, category, score, pitch angle, status
Result: one click → a clean, scored, ready-to-work lead list in a few minutes, zero manual research.
A few things I learned building it:
**•** \[optional: add a real technical detail here — e.g. how you handled rate limits on SerpAPI, or a prompt-tuning tip for the Claude scoring step, or how you structured the Code node\]
Happy to answer questions if anyone’s building something similar or wants to compare notes on the AI scoring prompt. I also packaged the finished version (workflow file + setup guide) for anyone who’d rather skip the build: Lead Scraping + CRM Auto-Entry — AI-Powered n8n Automation