🗺 Google Maps Scraper — Leads, Emails, Phones & Reviews

Turn any Google Maps search, URL or place ID into clean, export-ready business data.

Paste what you want, press Start, get a spreadsheet you can actually use: business name, category, rating, review count, phone, website, public email addresses, social profiles, address broken into fields, coordinates, opening hours, photos, search rank and ready-made lead scores.

No Playwright, no browser, no captchas to solve — the actor talks to Google Maps' own data endpoints, so it is fast and cheap.

---

✨ What makes this one different

One smart input box	Search phrases, Maps URLs, place IDs and data IDs all go into the same field. The actor detects each type automatically.
Three depth levels	⚡ Quick, 📇 Contacts, 🔎 Full — one switch, no fiddling with ten checkboxes.
Column presets	✨ Essentials, 📇 Leads, ⭐ Reviews, 🗂 Everything. Your CSV is readable the moment it downloads, with stable headers in a fixed order.
📊 Visual run report	Every run writes an HTML dashboard to the key-value store: coverage %, pipeline health, top categories and per-search-term results. Open it straight from the run's Output tab.
Smart deduplication	Beyond place ID: collapse chains by domain or phone number so you get one row per company.
Real filters	Rating, review count, include/exclude keywords, website presence, "must have a phone or email", open-now, closed businesses.
Handy deep links	whatsappLink, directionsUrl, reviewsUrl — generated for every row.
Flat text columns	emailsText, phonesText, categoriesText, photoUrlsText — no more JSON arrays inside your spreadsheet cells.

---

🚀 Quick start

1. Put your targets into 🔍 What do you want to scrape?

   coffee shops in New York, NY
   dentists in Austin, TX
   https://www.google.com/maps/place/Blue+Bottle+Coffee/...
   ChIJN1t_tDeuEmsRUsoyG83frY4
   

2. Leave everything else on its default and press Start.
3. Download the dataset as CSV / Excel / JSON, and open the 📊 Run report for a summary.

Need emails? Switch ⚙️ How deep should we go? to 📇 Contacts and set 📊 Which columns to 📇 Leads.

---

⚙️ Data depth

Mode	What you get	Speed
⚡ Quick	Places, categories, ratings, Google-published phone, address, coordinates, website, photos, opening state, search rank	Fastest, cheapest
📇 Contacts	Everything above + public emails and social profiles crawled from each business website (up to 3 pages, hard time budget)	Medium
🔎 Full	Everything above + place detail page + Google reviews with aggregated review statistics	Slowest, richest

---

🎚 Filters

All optional, all combined with AND, all applied before the row is saved:

• ⭐ Minimum Google rating and 💬 minimum number of reviews
• ✅ Keep only these categories — keyword match on category and title
• 🚫 Drop places matching these words — cut chains, franchises, noise
• 🌐 Website presence — with / without a website (the classic agency lead list)
• 📇 Contact requirement — must have a phone, an email, either, or both
• ⛔ Skip closed businesses and 🟢 only places open right now

---

📊 Output columns

Pick a preset in 📊 Which columns do you want?

Preset	Columns	Best for
✨ Essentials	23	A clean CSV to eyeball or hand to someone
📇 Leads	39	Cold outreach: emails, phones, socials, WhatsApp links, lead scores
⭐ Reviews	37	Reputation monitoring: star split, owner-reply rate, review recency
🗂 Everything	~150	Analysis and pipelines — nothing is dropped

Fixed presets always emit the same columns in the same order, even when a value is missing, so appended CSV exports line up across runs.

Scores the actor computes for you

Field	Meaning
leadScore	0–100 — how usable this row is as a sales lead
contactCompletenessScore	0–100 — how much contact data was found
digitalPresenceScore	0–100 — website, email, socials, description, photos
reputationScore	0–100 — rating, review volume, owner responsiveness
outreachOpportunityScore / outreachPriority	0–100 + high/medium/low, with outreachReasons explaining why
intelligenceConfidence	How much to trust an absence of contact data (high / medium / partial)

---

🧹 Deduplication

Places are always de-duplicated by Google place ID. Choose a stricter rule when you want one row per company:

• 🌐 …plus website domain — collapses every branch that shares a site
• ☎️ …plus phone number — collapses shared phone lines
• 🌐☎️ …plus domain or phone — strictest

---

🚀 Speed & cost

Setting	Parallel searches	Use when
🐢 Gentle	low	You see blocked requests, or you're on a weak proxy
⚖️ Balanced	medium	Default — good for almost everything
🚀 Turbo	high	Large keyword lists on residential proxies

Use 🛑 Stop after this many places to put a hard cap on a run, and 🔢 Max places per search term to control per-keyword depth.

Proxies: residential proxies are strongly recommended. Google Maps blocks datacenter IPs quickly.

---

📤 Output records

Record	What it is
Dataset	One row per business, shaped by your column preset
REPORT (key-value store)	📊 Visual HTML run report — coverage, pipeline health, per-term results
OUTPUT (key-value store)	🧾 JSON summary — counters, timings, warnings, dataset link

Sample row (✨ Essentials)

{
  "title": "Blue Bottle Coffee",
  "categoryName": "Coffee shop",
  "rating": 4.4,
  "reviewsCount": 1837,
  "priceLevel": "$$",
  "phone": "(510) 653-3394",
  "website": "https://bluebottlecoffee.com/",
  "email": "hello@bluebottlecoffee.com",
  "address": "300 Webster St, Oakland, CA 94607",
  "city": "Oakland",
  "state": "CA",
  "postalCode": "94607",
  "countryCode": "US",
  "latitude": 37.8007,
  "longitude": -122.2758,
  "businessStatus": "open",
  "openingHoursToday": "7 AM–5 PM",
  "imageUrl": "https://lh5.googleusercontent.com/...",
  "searchQuery": "coffee shops in Oakland, CA",
  "searchRank": 3,
  "placeId": "ChIJ...",
  "googleMapsUrl": "https://www.google.com/maps/place/?q=place_id:ChIJ...",
  "scrapedAt": "2026-09-03T10:21:44Z"
}

---

🛟 Mistakes are explained, not crashed on

A wrong setting never kills the run. The actor repairs what it can, keeps scraping, and tells you exactly what it did — in the run log, in the Input check section of the 📊 run report, and in OUTPUT.inputIssues.

You do this	What happens
Paste a Yelp or Facebook link	That entry is skipped: "is a link, but not a Google Maps link" — the rest of your list still runs
Require an email in ⚡ Quick mode	Warning: "Quick mode never collects emails — switch to 📇 Contacts or 🔎 Full, otherwise every place will be filtered out"
Require an email and only places without a website	Warning: emails come from websites, so this combination returns nothing
Put the same word in both keep and drop lists	Warning: exclusions win
Send a number where a list was expected, an unknown option, or an out-of-range value	Repaired to a sane value, with the valid options listed
Turn the proxy off	Warning: Google blocks datacenter IPs quickly

When a run saves nothing, the run status message says why — it repeats the specific problem instead of a generic "nothing found".

---

❓ FAQ

Does it scrape emails from Google Maps? Google Maps does not publish emails. In 📇 Contacts and 🔎 Full modes the actor visits the business's own website and collects the email addresses published there.

Why did I get fewer places than my limit? Google returns a finite result set per search — a narrow query in a small town simply has fewer businesses. Broaden the phrase, drop the location, or add more search terms.

Can I re-enrich a list I already have? Yes. Paste your Google Maps URLs, place IDs or data IDs into the same input field and pick a deeper mode.

Is this legal? The actor only collects publicly visible information. You are responsible for how you use it — in particular for complying with GDPR/CCPA and anti-spam rules when contacting businesses.

---

🛠 Local development

pip install -r requirements.txt
python -m pytest tests -q      # unit tests, no network
apify run                      # run locally with ./storage/key_value_stores/default/INPUT.json
apify push                     # deploy