$299 → $0 · the Saudi coffee-shop list, rebuilt free
That SmartScrapers list is just Google Maps + OpenStreetMap scraped and resold — public data. Their own page: 8,984 cafés, only 138 with emails (1.5%). You’re paying for a scrape you can run yourself.
Every source below is free. Crack open the drops 
🪤 what your $299 actually buys
Their own numbers, out of 8,984 shops:
| Field |
Filled |
| Phone |
5,256 |
| Facebook |
972 |
| LinkedIn |
75 |
| Email |
138 (1.5%) |
So the “premium” product = a Google Maps scrape (scrape = a bot copying on-screen data into a spreadsheet) + a weak email crawl. Name, address, phone, rating, reviews, website, socials — all public on Google Maps + OpenStreetMap right now.
Set expectations: email is the one hard column. Most Saudi cafés are Instagram + phone only — even the paid vendor got 1.5%. Chase phone + socials, grab emails where they exist.
🗺️ the map that bills you nothing
OpenStreetMap (free volunteer world map — Wikipedia for places) tags every café amenity=cafe. Pull them all with Overpass Turbo (a search box for OSM). No signup, no limit → export CSV (plain spreadsheet file).
Paste this, hit Run, then Export → CSV:
[out:csv(name,"contact:phone",phone,"contact:website",website,"contact:email",email,"contact:instagram",::lat,::lon;true;",")]
[timeout:180];
{{geocodeArea:Saudi Arabia}}->.searchArea;
(
nwr["amenity"="cafe"](area.searchArea);
nwr["shop"="coffee"](area.searchArea);
nwr["cuisine"="coffee_shop"](area.searchArea);
);
out center;
- OSM stores contacts under two spellings (
phone and contact:phone) — query grabs both so you don’t lose half the numbers.
- Arabic turning to junk? Export “raw data directly from Overpass API” or import as UTF-8 (a text-encoding setting).
- Just want a count first? Swap the last line for
out count;.
Whole country in one file: Geofabrik GCC extract (.pbf = a region’s entire map in one download, refreshed daily). Filter with osmconvert/osmfilter, or drag it into free QGIS.
OSM has fewer cafés than Google (a few thousand). Free base layer — top up from Google below.
💣 106 million businesses, left unlocked
Late 2024, Foursquare dumped its entire places database open + free: Foursquare OS Places — 106M businesses worldwide, name/address/phone/website/socials, monthly refresh, commercial-use license. The sellers never mention this.
Filter it to Saudi cafés with DuckDB (free tool that reads giant files on your laptop) — no full download needed:
SELECT name, address, locality, region, tel, website, email, latitude, longitude
FROM read_parquet('s3://fsq-os-places-us-east-1/release/dt=2025-12-08/places/parquet/*.parquet')
WHERE country = 'SA'
AND (list_contains(fsq_category_labels,'Café')
OR list_contains(fsq_category_labels,'Coffee Shop'));
Sister dump: Overture Maps (Meta + Microsoft + Amazon + TomTom). One line grabs the Saudi map-box: overturemaps download --bbox=34,16,56,33 -f geojson --type=place.
🔧 rip google maps raw — every café, $0
The 8,984 count came from Google, so go to the same well.
No-code (start here):
Self-hosted (unlimited):
| Tool |
Free |
Note |
| omkarcloud |
200 searches/mo → 10k+ leads |
desktop app, 50+ fields, email+social built in |
| gosom |
forever |
add -email → crawls each site for the email |
Two make-or-break tricks:
- 120-result cap: Google returns max ~120 per search. Split by city/district, stack results (good scrapers auto-split if the city goes in the city field).
- Search Arabic or miss half:
coffee shop · cafe · مقهى · كافيه · قهوة · كوفي.
- Run self-hosted on your home internet (residential IP = a home address) — Google blocks server-farm IPs fast.
Official tap: Google Places API throws in thousands of free calls/month.
🥷 the saudi stash nobody scrapes
Every Saudi delivery app is a café directory with name, address, exact location, often phone — by city, served fresh. Catches shops that never touched Google: HungerStation, Jahez, Mrsool, ToYou, Careem Food, Talabat, Ninja (each loads its restaurant list from a hidden data feed the browser reads).
HungerStation is Saudi-locked + anti-bot → needs a Saudi home IP. Starter scraper on GitHub: scrape-do HungerStation — bend it to pull the store list. Bonus layer for phone gaps.
🏛️ the govt left the door open
- Maroof — Ministry of Commerce business directory. Each shop links its website + every social account on its page. Best free enrichment (enrichment = filling missing columns) — name in, Instagram + site out.
- Saudi Open Data Portal — 11k+ official datasets, CSV/JSON, no signup. The Ministry of Commerce open data has the commercial registry (name, activity, city, phone), searchable.
📬 the 2 columns google buries: emails + socials
Emails (ranked by free yield):
- Crawl the café’s own site — gosom
-email / omkarcloud built-in. Hits info@/contact@. Best.
- Free finder tiers: Hunter (25/mo) · Apollo (big free tier) · Snov (50) · QuickEnrich (300, no signup). Burn one, rotate.
- WHOIS (who-owns-this-site record) — sometimes leaks a contact.
Socials (Instagram is the real prize here):
- omkarcloud/gosom/gmplus/Apify already grab FB/IG/TikTok/X links off listings + sites.
- Apify IG email scraper pulls the email+phone behind the IG “contact” button.
- OSINT kit (OSINT = digging public info): Social-Media-OSINT-Tools.
- Or Maroof + Google
"café name" instagram.
🧵 fuse it into one clean sheet
- Dedupe by Google Place ID (Google’s unique code per business); else by phone; else same name + same spot.
- Normalize every phone to +966.
- gosom’s built-in “LeadsDB” auto-dedupes, or a one-line spreadsheet/pandas
drop_duplicates.
- Re-run every few months — cafés open/close by the hundreds yearly.
💸 where this actually prints
Same method, one payoff: a free contact list of any business type, any city, on demand.
- Sell espresso gear, beans, or POS registers → every café in Riyadh with a phone, cold-call Monday.
- Run web design / social media as a side gig → filter cafés with no website = warm leads who need exactly you.
- Building a delivery or loyalty app → the whole market mapped free before you spend a riyal on research.
- Swap “café” for gym, dentist, barber, car wash → any niche, any country, same three clicks.
- People resell these lists for $299 — you now make them for $0.
The stack, short:
OSM + Foursquare + Google Maps → dedupe (+966, Place ID) → your list
fastest: Overpass query (5 min) · fullest: omkarcloud · socials: Maroof
Public data is fair game, but it breaks Google’s ToS and Saudi PDPL covers personal info — keep it B2B, honor opt-outs, credit OSM/Foursquare.
They didn’t sell you data. They sold you the ten minutes you didn’t know you could skip.