Daily Nationwide Facebook Auto Leads

I need a reliable way to capture every car listing that appears on Facebook Marketplace across the entire United States and extract the seller’s direct phone number from each ad. The listings change quickly, so I want a fresh pull every 24 hours and the results delivered as a clean CSV file I can drop straight into my CRM.

Here’s what the file should include each day:
• Listing URL
• Ad title exactly as posted
• Asking price
• City / state shown in the ad
• Date and time the ad was scraped
• Seller’s name (if visible)
• Seller’s phone number in a normalized format

The CSV must overwrite the previous day’s file or arrive as a clearly time-stamped new file; either way, I should always know I’m working with the latest nationwide data set.

I’m open to any stack—Python, Node, Selenium, headless browsers, API work-arounds—so long as the final feed is dependable, free of duplicates, and keeps pace with Facebook’s anti-scraping changes. If you already have proxy management, rotating user-agents, or other mitigation in place, even better; I don’t want interruptions.

For acceptance, I’ll import a sample day’s CSV into my system, spot-check random listings against Marketplace, and verify that at least 95 % include valid phone numbers. After that, we’ll switch to the daily schedule.