Facebook Marketplace Deal Radar

Find flip-worthy Facebook Marketplace deals before everyone else — multi-city USA fanout + smart query expansion.

Built for resellers, phone flippers, and local arbitrage operators who need new listings + a deal score, not another raw scrape dump.

Filters (curious_coder / memo23 parity)

• condition — new / used_like_new / used_good / used_fair
• sortBy — newest, price, or distance on search URLs
• deliveryMethod — local_pickup or shipping
• onlyWithPhotos — skip text-only cards
• allPhotos — emit photos[] (full gallery when available; else primary imageUrl)
• categories[] — Marketplace category fanout (vehicles, electronics, home, …); merged + deduped
• latitude / longitude — optional precise pin with radiusKm on search URLs
• onlyNewListings — alias of newOnly (seen-ID KV)
• proxyConfiguration — optional Apify proxy (cookies still required for live)

Why this Actor (not a generic Marketplace scraper)

You get	Most scrapers
Multi-city cities[] / locations[] fanout	One city per run
smartQuery + synonyms[] query expansion	Type one FB search and hope
dealScore 0–100 on every listing	Raw title/price only
newOnly across scheduled runs (KV seen-IDs)	Full re-dumps every time
Discord + webhook alerts for hot finds	Export CSV and hope
Vehicle fields parsed from titles	Manual regex later
Honest demoMode (no fake “success” without cookies)	Silent empty runs

Live mode uses your Facebook session cookies. Anonymous Marketplace scraping does not work reliably — we fail clearly instead of billing empty air.

Vehicles vertical (car / truck / SUV flippers)

Set vertical = vehicles to unlock flipper presets without changing the Greenville general schedule:

Input	Effect
vertical=vehicles	Vehicle-tuned dealScore + prefer vehicle listings
vehiclePresets	cars / trucks / suv query expansion (defaults to all three)
vehicleYearMin / vehicleYearMax	Filter by year parsed from listing title/description
vehicleMake	Filter by make parsed from listing text (e.g. Honda)

Example: vertical=vehicles, cities=["austin","houston"], vehiclePresets=["trucks","suv"], vehicleYearMin=2015.

PPE event names stay listing / listing-detail / actor-start. Store title stays Facebook Marketplace Deal Radar.

Pricing (PPE)

You pay for value delivered:

Event	What it is	Price
listing	One Marketplace listing pushed	$0.0012 (~$1.20 / 1k)
listing-detail	Optional detail-page enrichment	$0.002
actor-start	Run start	$0.005

demoMode is free of Actor charges (sample data only).

Vs Store comps (Bronze, approx): official Apify ~$5/1k · curious_coder ~$0.50–$1.50/1k · memo23 ~$1.50/1k · cheapest Discord scrapers ~$0.40/1k. Deal Radar sits in the smart-middle: cheaper than official, priced for deal intelligence (score + alerts), not race-to-bottom raw rows.

Quick start

1. Open the Actor → Try for free / create a Task.
2. Set searchQuery (e.g. iPhone) and either:
   • Multi-city: cities = ["greenville","austin","houston"], or
   • Single city: location = greenville (Marketplace city slug).
3. Optional: turn on smartQuery and/or add synonyms for query expansion.
4. Paste sessionCookies (c_user + xs from a logged-in Chrome → Application → Cookies → facebook.com). Treat like a password.
5. Turn on newOnly + optional discordWebhook.
6. Schedule the Task (cloud) so deals keep flowing while your PC sleeps.

Input highlights

• searchQuery / smartQuery / synonyms[] / startUrls
• location (single) · cities[] / locations[] (multi-metro)
• radiusKm · minPrice / maxPrice · daysSinceListed · maxResults
• includeDetails · newOnly · webhookUrl · discordWebhook
• vertical · vehiclePresets · vehicleYearMin/vehicleYearMax · vehicleMake
• demoMode · sessionCookies · storageStateJson

Output

Each dataset item includes title, price, location, citySlug, URL, image, recency signals, optional vehicle fields, searchQuery, and dealScore.

Auth notes

• Cookies expire — when you hit a login wall, refresh xs in the Task input (Console only; never commit cookies).
• Prefer a dedicated FB account you control for scraping.

Support

IntelScrape on Apify. Issues → Actor issues / Console support.

License

Apache-2.0