Google Maps Scraper: businesses, reviews, popular times, contacts, jobs

Scrape Google Maps search results anywhere in the world. You get business names, addresses, phone numbers, websites, reviews, ratings, opening hours, GPS coordinates, and 30+ structured fields per place. Fast, reliable, and priced per result.

Optionally chain enrichment on every website found: emails and phones via a regex extractor (no API key), or AI-powered contacts and job listings using your own Gemini, Groq, or OpenRouter key. All merged into a single dataset row per business.

Why use this scraper

• All-inclusive pricing. 30+ fields per place at one flat rate. No extra charges for reviews, photos, hours, or address parsing.
• Fast. Results in seconds, not minutes.
• Volume-friendly. $1.00 per 1,000 places.
• Built-in enrichment. One input flag adds emails, phones, socials, contacts, or job listings from each business's website.
• Bring your own LLM key. Use Gemini (recommended, free tier available), Groq, or OpenRouter. Automatic fallback between providers when you supply multiple keys.
• Works worldwide. Any country, any language. Pass an ISO language code and get localized results.

Use with AI agents (MCP)

Connect this actor to any MCP-compatible AI client: Claude Desktop, Claude.ai, Cursor, VS Code, LangChain, LlamaIndex, or a custom agent.

MCP server URL:

https://mcp.apify.com?tools=santamaria-automations/google-maps-scraper

Example prompt once connected:

"Use google-maps-scraper to find pediatric dentists in Zurich and put the results in a table."

Clients that support dynamic tool discovery (Claude.ai, VS Code) will receive the full input schema automatically via add-actor.

Quick start

Simple search

Pass plain search strings. Include the location in the string itself:

{
  "searchStrings": [
    "restaurants in Paris",
    "coffee shops in Tokyo",
    "dentists in New York"
  ],
  "maxResultsPerQuery": 20
}

Structured queries

Use queries when you want to attach your own company_id for join-back, or split the search term from the location:

{
  "queries": [
    {
      "query": "software companies",
      "location": "Berlin",
      "country": "DE",
      "company_id": "my-internal-id-123"
    },
    {
      "query": "plumbers",
      "location": "London",
      "country": "GB"
    }
  ],
  "maxResultsPerQuery": 50,
  "language": "en"
}

Add emails, phones, and social links (no API key)

The email extractor pulls contact details straight from each place's website using regex. No LLM key needed, no AI cost.

{
  "searchStrings": ["altenpflegeheim konstanz"],
  "maxResultsPerQuery": 20,
  "enableEmailExtraction": true
}

Each place row gets four new fields: emails_from_website, phones_from_website, social_urls, and address_from_website.

Add AI-powered contacts and job listings

Bring your own LLM API key. Structured contacts (name, position, email, phone) and job listings get merged into the same dataset row as the Maps data.

{
  "searchStrings": ["IT companies in Munich"],
  "maxResultsPerQuery": 20,
  "enableContactExtraction": true,
  "enableJobExtraction": true,
  "enableEmailExtraction": true,
  "groqApiKey": "gsk_your_own_key",
  "waitForAddOns": true
}

The scraper waits for each add-on to finish and merges the results into each place row. You end up with one row per business, everything in the same dataset. That's the flow that works nicely for n8n, Zapier, or Make workflows where you don't want to join multiple datasets.

Input parameters

Parameter	Type	Default	Description
searchStrings	string[]	[]	Simple search strings. Include the location in the string for geo-targeting.
queries	object[]	[]	Structured queries with optional query, location, country, company_id.
maxResultsPerQuery	integer	20	Results per query. Max 120 (Google Maps page limit).
maxTotalResults	integer	(none)	Optional global cap across all queries combined. Leave empty for no cap. Useful for budget control.
maxResults	integer	(none)	Deprecated alias for maxResultsPerQuery. Still works for backward compatibility.
language	string	"en"	ISO 639-1 language code (en, de, fr, ja, es, pt, it, ko, zh, ar, etc.).
includeReviewCount	boolean	true	Enrich each result with review_count via the Maps Embed API. Adds about 150 ms per place.
includePopularTimes	boolean	false	Opt-in. Enrich each result with the visitor-traffic histogram by weekday × hour (0–100). Adds about 1.5 s per place.
includeReviews	boolean	false	Opt-in. Fetch featured reviews for each place (author, rating, date, text). Adds about 300 ms per place.
maxReviewsPerPlace	integer	25	Maximum reviews per place. Set to 0 for no cap.
reviewsSort	string	most_relevant	Sort order: most_relevant, newest, highest_rating, lowest_rating.
reviewsNewerThan	string	(none)	Only return reviews on or after this date (YYYY-MM-DD). Auto-switches sort to newest for an efficient cutoff.
enableEmailExtraction	boolean	false	Regex extractor for emails, phone numbers, and social URLs (LinkedIn, Facebook, Instagram, Twitter, YouTube). No LLM key required.
enableContactExtraction	boolean	false	AI-powered structured contact extraction (name, position, email, phone). Requires your own LLM API key.
enableJobExtraction	boolean	false	AI-powered job-listing extraction from careers pages. Requires your own LLM API key.
enableBrowserFallback	boolean	false	Re-scrape JS-rendered sites (React, Vue, Angular) with a full browser. Catches about 25% more sites.
outputLanguage	string	"en"	Language for AI-extracted fields (en, de, fr, it, es, pt, nl, auto).
geminiApiKey	string	(none)	Your Gemini API key. Free tier is generous, so this is the recommended starting point.
groqApiKey	string	(none)	Your Groq API key. Ultra-fast inference with a free tier.
openrouterApiKey	string	(none)	Your OpenRouter API key. Access to 100+ models. Optional fallback.
llmModel	string	(provider default)	Optional model override. Useful when your OpenRouter key doesn't have access to the default (google/gemini-2.0-flash-001) and you need to point at a free model like meta-llama/llama-3.1-8b-instruct. Also works for Gemini and Groq.
waitForAddOns	boolean	true	When true (default): the scraper waits for each add-on and merges its output into the same dataset row as the Maps result. When false: add-ons run fire-and-forget and write to their own datasets, and you handle the join yourself.
addOnWaitCapMs	integer	1800000	Safety cap on how long to wait for add-ons before flushing remaining places with a timed_out status. Default 30 min.
excludeCids	string[]	[]	Google CIDs to skip. Useful for excluding already-scraped places.
maxConcurrency	integer	10	Parallel requests (1–20).
requestDelay	integer	300	Delay between pagination requests in milliseconds (100–10,000).
proxyConfiguration	object	Apify proxy	Proxy settings. Datacenter proxies work well and are cost-efficient.

Use either searchStrings or queries, or both. searchStrings sends each string directly to Google Maps. queries lets you separate the search term from the location and attach your own company_id for linking results back to your data.

About the LLM API key fields. They're shown as plain text (not masked) in the Apify UI. That's a technical necessity for how Apify forwards fields to sub-actors right now: secret-marked fields get encrypted on the parent side and the sub-actor can't decrypt them. If the visibility bothers you, use a scoped or throwaway key.

Output

Each result has 30+ fields. Here's a shortened example (real Paris restaurant, some URLs abbreviated):

{
  "company_id": "restaurants in Paris",
  "title": "Le Petit Cler",
  "category": "French restaurant",
  "categories": ["French restaurant", "Restaurant"],
  "address": "3 Rue Cler, 75007 Paris, France",
  "complete_address": {
    "street": "3 Rue Cler",
    "city": "Paris",
    "postal_code": "75007",
    "state": "Ile-de-France",
    "country": "France"
  },
  "latitude": 48.8571,
  "longitude": 2.3007,
  "plus_code": "8FW4V87C+X3",
  "timezone": "Europe/Paris",
  "phone": "+33 1 45 51 24 52",
  "website": "https://www.lepetitcler.com/",
  "emails": [],
  "review_rating": 4.5,
  "review_count": 1823,
  "reviews_per_rating": { "1": 42, "2": 31, "3": 89, "4": 312, "5": 1349 },
  "reviews_link": "https://search.google.com/local/reviews?placeid=...",
  "user_reviews": [
    {
      "name": "Marie Dupont",
      "profile_link": "https://www.google.com/maps/contrib/...",
      "rating": 5,
      "text": "Excellent cuisine, warm atmosphere...",
      "published_at": "2025-12-10",
      "response_from_owner": null
    }
  ],
  "status": "OPERATIONAL",
  "price_range": "$$",
  "open_hours": {
    "Monday": ["12:00-14:30", "19:00-22:30"],
    "Tuesday": ["12:00-14:30", "19:00-22:30"],
    "Wednesday": ["12:00-14:30", "19:00-22:30"],
    "Thursday": ["12:00-14:30", "19:00-22:30"],
    "Friday": ["12:00-14:30", "19:00-23:00"],
    "Saturday": ["12:00-23:00"],
    "Sunday": []
  },
  "thumbnail": "https://lh3.googleusercontent.com/places/...",
  "images": [
    { "title": "Interior", "image": "https://lh3.googleusercontent.com/..." }
  ],
  "owner": { "name": "Le Petit Cler", "link": "https://maps.google.com/..." },
  "about": [
    { "id": "dine_in", "name": "Dine-in" },
    { "id": "reservations", "name": "Reservations" }
  ],
  "reservations": [
    { "link": "https://www.thefork.com/...", "source": "TheFork" }
  ],
  "order_online": [],
  "menu": { "link": "https://www.lepetitcler.com/menu", "source": "lepetitcler.com" },
  "description": "Cozy French bistro in the heart of the 7th arrondissement.",
  "link": "https://www.google.com/maps/place/Le+Petit+Cler/...",
  "cid": "12345678901234567",
  "data_id": "0x47e671a4e...",
  "searchString": "restaurants in Paris",
  "_rank": 1,
  "scraped_at": "2026-03-05T10:30:00.000Z"
}

When you enable add-ons with waitForAddOns: true, the same row gets these extra fields merged in:

{
  "emails_from_website": ["info@lepetitcler.com", "reservations@lepetitcler.com"],
  "phones_from_website": ["+33 1 45 51 24 52"],
  "social_urls": {
    "facebook": "https://www.facebook.com/lepetitcler",
    "instagram": "https://www.instagram.com/lepetitcler"
  },
  "address_from_website": "3 Rue Cler, 75007 Paris",
  "contacts": [
    { "name": "Marie Dubois", "position": "Head Chef", "email": "marie@lepetitcler.com" }
  ],
  "job_listings": [
    { "title": "Sous Chef", "location": "Paris", "job_url": "https://...", "ats_system": "personio" }
  ],
  "_enrichment_status": { "emails": "success", "contacts": "success", "jobs": "no_data" }
}

Complete field reference

Field	Type	Description
company_id	string	Your reference ID (from queries) or the search string
title	string	Business name
category	string	Primary Google Maps category
categories	string[]	All assigned categories
address	string	Full formatted address
complete_address	object	Parsed components: street, city, postal_code, state, country
latitude	number	GPS latitude
longitude	number	GPS longitude
plus_code	string	Google Plus Code
timezone	string	IANA timezone identifier
phone	string	Phone number
website	string	Website URL
emails	string[]	Email addresses found on the listing
open_hours	object	Opening hours grouped by day of week
popular_times	object	Visitor traffic patterns (opt-in via includePopularTimes)
price_range	string	Price level ($ to $$$$)
status	string	OPERATIONAL, CLOSED_TEMPORARILY, or CLOSED_PERMANENTLY
review_count	integer	Total number of reviews
review_rating	number	Average rating (1.0 to 5.0)
reviews_per_rating	object	Breakdown of reviews by star rating (1 through 5)
reviews_link	string	Direct URL to all reviews
user_reviews	array	Sample reviews with name, rating, text, date, owner response
thumbnail	string	Main business photo URL
images	array	Additional photos with titles
reservations	array	Reservation links with source (TheFork, OpenTable, etc.)
order_online	array	Online ordering links with source
menu	object	Menu link and source
owner	object	Business owner name and profile link
about	array	Amenities and features (dine-in, Wi-Fi, wheelchair access, etc.)
description	string	Business description from Google
link	string	Full Google Maps URL for this place
cid	string	Google Maps CID: a stable identifier that persists across name and address changes
data_id	string	Google internal data ID
searchString	string	The search query that produced this result
_rank	integer	Position in search results (1-based)
scraped_at	string	ISO 8601 timestamp
emails_from_website	string[]	Emails extracted from the company's website. Add-on: email extractor
phones_from_website	string[]	Phone numbers extracted from the company's website
social_urls	object	Social profiles from the website: linkedin, facebook, instagram, twitter, youtube
address_from_website	string	Postal address from schema.org data on the website
contacts	array	AI-extracted structured contacts from the website (name, position, email, phone). Add-on: contact extractor
job_listings	array	AI-extracted job listings from the careers page. Add-on: job extractor
_enrichment_status	object	Per-add-on status: success, no_data, failed, timed_out, or skipped. Present when any add-on is enabled and waitForAddOns is true

About popular_times

Set includePopularTimes: true to enrich each result with Google's visitor-traffic histogram. The output shape is Day → Hour → Popularity(0-100) with English day keys (Sunday, Monday, ...) regardless of the input language, so the data stays portable across runs.

Each enrichment is one extra request to Google's place-detail endpoint, run sequentially (concurrency 1) with a 300 ms delay. Budget roughly 1.5 s per place on top of the SERP scrape. Only places with enough foot traffic for Google to compute the histogram will have non-null data. Smaller venues keep popular_times: null, which is expected, not an error.

Example:

"popular_times": {
  "Sunday":   {"12": 41, "13": 56, "14": 63, "15": 70, "16": 73, "17": 83, "18": 76},
  "Monday":   {"11": 43, "12": 61, "13": 60, "14": 53}
}

About user_reviews

Set includeReviews: true to enrich each result with the place's featured reviews. Each review has the author's name, star rating, publication date, profile link, and text when the reviewer wrote one. Adds one extra request per place (about 300 ms).

• maxReviewsPerPlace caps the number of reviews per place (default 25). Set to 0 for no cap. The actual number returned depends on how many featured reviews Google surfaces, typically up to 8 for high-traffic places.
• reviewsSort controls ordering: most_relevant (default), newest, highest_rating, lowest_rating.
• reviewsNewerThan (YYYY-MM-DD) filters to reviews published on or after that date. It automatically switches sort to newest so scraping stops as soon as older reviews are reached.

Reviews come from Google's place-preview endpoint, which returns the same featured reviews shown in the Google Maps overview panel. These are the reviews Google has selected as most helpful, typically the most recent positive reviews with photos or detailed text. Only places with enough review activity will return non-empty user_reviews; low-traffic places get user_reviews: [].

"user_reviews": [
  {
    "name": "Marie Dupont",
    "profile_link": "https://www.google.com/maps/contrib/114798109881286611060",
    "rating": 5,
    "text": "Excellent cuisine, warm atmosphere.",
    "published_at": "2025-12-10",
    "response_from_owner": null
  }
]

Enrichment add-ons

After the Maps scrape finishes, this actor can trigger enrichment sub-actors on every website it found and merge everything back into a single dataset. One row per place, with contacts, jobs, emails, phones, and socials attached.

Single-dataset mode (waitForAddOns: true, the default) is designed for n8n, Zapier, and Make workflows where you don't want to join multiple datasets. The scraper waits for each add-on to finish, merges the results by CID and domain, and pushes the enriched row once everything is in. Every row also gets an _enrichment_status object showing what happened per add-on: success, no_data, failed, timed_out, or skipped.

Multi-dataset mode (waitForAddOns: false) fires the add-ons and exits. You'll find their run IDs in the key-value store under CONTACT_EXTRACTOR_RUN_ID, JOB_EXTRACTOR_RUN_ID, and EMAIL_SCRAPER_RUN_ID, and you can consume them however you want.

Emails, phones, and social links (no API key)

Enable with enableEmailExtraction: true. Pure regex, no LLM key, no AI cost. Powered by the Website Email & Phone Scraper.

Populates these fields on each place row:

• emails_from_website: email addresses from mailto: links, plain text, and structured data
• phones_from_website: phone numbers from tel: links and international-format text
• social_urls: linkedin, facebook, instagram, twitter, youtube
• address_from_website: postal address from schema.org JSON-LD when available

AI-powered contact extraction

Enable with enableContactExtraction: true and provide your own LLM API key. Powered by the Website Contact Extractor.

Populates contacts on each place row with structured records (name, position, email, phone) extracted from team, about, impressum, and contact pages.

AI-powered job listing extraction

Enable with enableJobExtraction: true and provide your own LLM API key. Powered by the Website Job Extractor.

Populates job_listings on each place row with open positions (title, location, department, job_url, application_url, posted_at, ats_system) extracted from careers pages and ATS portals: Personio, Greenhouse, Ashby, Workable, and others.

Browser fallback for JavaScript-rendered sites

Some company websites are built with React, Vue, or Angular and need a full browser to render. When enableBrowserFallback: true, the AI-powered add-ons re-scrape these sites with a real browser. Catches about 25% more sites but increases cost and runtime.

LLM API keys: bring your own

The two AI-powered add-ons need an LLM API key. You provide your own. The key is forwarded through the scraper to the AI sub-actor, and all LLM calls happen against your account. The Maps scraper itself never uses your key for anything else.

Providers (any one works; multiple enables automatic fallback):

1. Gemini (recommended). Best quality-to-cost ratio, generous free tier. Get a key at Google AI Studio.
2. Groq. Ultra-fast inference, generous free tier. Get a key at Groq Console.
3. OpenRouter. Access to 100+ models. Get a key at OpenRouter.

If you're using OpenRouter and hit a "model not found" error, your key probably doesn't have access to the default paid Gemini model. Set llmModel to a free option like meta-llama/llama-3.1-8b-instruct, or browse available models at openrouter.ai/models.

Use cases

• Lead generation. Search for businesses by type and location. Get phone numbers, websites, and emails in one run.
• Market research. Map competitors in a geographic area with ratings, review counts, and price ranges.
• Company enrichment. Add addresses, phone numbers, GPS coordinates, and opening hours to your existing database.
• Local SEO monitoring. Track your business listing and competitor rankings across locations.
• Real estate analysis. Find nearby amenities, restaurants, and services for property listings.
• Hiring intelligence. Discover companies in a region and extract their open positions in a single pipeline.
• Review analysis. Collect review ratings and sample reviews across competitors for sentiment analysis.

Skip already-scraped places

Use the excludeCids parameter to skip places you already have. The CID (Customer ID) is a stable Google Maps identifier that doesn't change even if the business changes its name or address.

{
  "searchStrings": ["cafes in Zurich"],
  "maxResultsPerQuery": 100,
  "excludeCids": ["12345678901234567", "98765432109876543"]
}

This is useful for incremental scraping workflows where you only want new results.

Supported languages

Pass any ISO 639-1 language code: en, de, fr, ja, es, pt, it, ko, zh-CN, ar, nl, pl, sv, da, fi, no, cs, hu, ro, el, tr, th, vi, id, and more.

Tips for better results

• Include the location in your search. "plumber Berlin" works better than just "plumber".
• Use the local language for better results in non-English countries. "Zahnarzt Zurich" beats "dentist Zurich".
• Set language to match your target market: de for German results, fr for French, and so on.
• Use excludeCids to avoid re-scraping places you already have from previous runs.

Pricing

$1.00 per 1,000 places scraped. All 30+ fields are included per place. No extra charges for reviews, photos, opening hours, parsed addresses, or any other field.

Example costs:

• 100 places = $0.10
• 1,000 places = $1.00
• 10,000 places = $10.00

Add-on	Pricing
Emails, phones, socials (regex, no LLM)	Billed separately by the Website Email & Phone Scraper
AI-powered contact extraction	Billed separately by the Website Contact Extractor, plus your own LLM API cost
AI-powered job listing extraction	Billed separately by the Website Job Extractor, plus your own LLM API cost

Limitations

• About 120 results per query. Google Maps returns a maximum of roughly 120 places per search query. For broader coverage, split your search into multiple queries by neighborhood or sub-region.
• No individual place detail pages. Data comes from search results, not from visiting each place page individually. Most fields are populated, but some niche fields (full review text, all photos) may be limited to what Google includes in search results.
• Search relevance. Results depend on Google's ranking algorithm. The same query may return slightly different results over time.
• Rate limiting. Very high concurrency or very low request delays may trigger temporary blocks from Google. The defaults (10 concurrency, 300 ms delay) are tuned for reliability.

Related actors

Find businesses to enrich

• Kleinanzeigen.de Scraper: Germany
• Willhaben.at Scraper: Austria
• Tutti.ch Scraper: Switzerland
• Subito.it Scraper: Italy
• Marktplaats.nl Scraper: Netherlands

Real estate

• Homegate.ch Scraper: Swiss real estate
• Immowelt.de Scraper: German real estate
• Oikotie.fi Scraper: Finnish real estate

Vehicles and travel

• AutoScout24 Scraper: European car listings
• Booking.com Scraper: hotel listings

Enrichment

• Website Email & Phone Scraper
• Website Contact Extractor
• Website Job Extractor
• LinkedIn Scraper

Support

Found a bug or have a feature request? Open an issue.