Facebook Marketplace Scraper (Apify Actor)

Custom Apify actor that scrapes Facebook Marketplace search results by intercepting GraphQL responses in a real browser. Replaces apify/facebook-marketplace-scraper to eliminate per-event billing (~$0.0062/listing).

How it works

Uses PlaywrightCrawler (Crawlee) with Apify residential proxies (RESIDENTIAL, US). Navigates to a search URL, intercepts /api/graphql responses as they stream in, and extracts listing data from data.marketplace_search.feed_units.edges[].node.listing (falls back to data.viewer.marketplace_feed_units). Images, fonts, and media are blocked to reduce proxy bandwidth.

Input

{
  "startUrls": [{ "url": "https://www.facebook.com/marketplace/nyc/search?query=bike" }],
  "resultsLimit": 20
}

Field	Type	Default	Max
startUrls	array of { url }	required	—
resultsLimit	integer	20	100

Output

Each run pushes an array of objects to the Apify dataset:

{
  "facebookUrl": "https://www.facebook.com/marketplace/nyc/search?query=bike",
  "listingUrl": "https://www.facebook.com/marketplace/item/123456789",
  "id": "123456789",
  "marketplace_listing_title": "Trek Road Bike",
  "primary_listing_photo": { "photo_image_url": "https://..." },
  "listing_price": {
    "amount": "350",
    "formatted_amount": "$350",
    "amount_with_offset_in_currency": "35000"
  },
  "location": { "reverse_geocode": { "city": "New York", "state": "NY" } },
  "is_hidden": false,
  "is_live": true,
  "is_pending": false,
  "is_sold": false,
  "delivery_types": ["PICKUP"]
}

If no listings are collected, a debug screenshot is saved to the actor key-value store under debug-screenshot.

Development

npm install       # install deps + generate package-lock.json (commit this)
npm run build     # tsc → dist/

Deployment

apify-cli is a local dev dependency — no global install needed.

# First time only
./node_modules/.bin/apify login

# Build + push in one step
npm run deploy

After pushing, copy the actor ID from the Apify console (your-username~facebook-marketplace-scraper) and update ACTOR_ID in apps/api/src/adapters/facebook.ts.

Infrastructure

• Base image: apify/actor-node-playwright-chrome pinned to digest in Dockerfile — update by pulling the new tag and re-running docker inspect ... --format='{{index .RepoDigests 0}}'
• Proxies: Apify residential pool, US only — requires an Apify plan with residential proxy access
• Memory: 1024 MB minimum (Playwright + Chrome)
• Runtime: Node 20, ESM ("type": "module")