Google Maps Email Scraper 📇 Contact Extractor (emails, phones)

Search Google Maps the way you would in the app, and get emails, phone numbers and social profiles for every business that comes back. Each business's own website is read for you: the homepage, and then the pages that actually carry a contact, which on a French site is "mentions légales" and on a German one is the Impressum.

One flat row per business. Empty where the data genuinely is not there, never a guessed address.

Google Maps scraper vs this actor

	plain Maps scraper	Google Maps Contact Extractor
Business name, address, phone, website	yes	yes
Rating, category, coordinates, place id	yes	yes
Advertisements filtered out of the results	usually not	yes, always
Email address from the business's own site	no	yes
Extra phone numbers published on the site	no	yes, in international format
LinkedIn, Facebook, Instagram and X profiles	no	yes, real profiles only
Contact and legal pages read, not just the homepage	no	yes, up to 3 per site
Says why a business has no email	no	yes, websiteStatus

A Maps scraper gives you the shopfront. This one gives you the mailbox you can actually write to.

What you get per business

Field	What it is
name, category, categories	The business as Google lists it
address, street, city, postalCode, country	Split out, ready for a mail merge
phone, phoneNational	The number Google holds, international format first
website, websiteDomain	Tracking parameters stripped off
rating, latitude, longitude, placeId, cid, googleMapsUrl	The usual Maps fields
email	The best address found on the site, general mailbox first
emails	Every address found, deduplicated and lowercased
websitePhones	Numbers published on the site, as they are written there
websitePhonesE164	The same numbers as +33…, when they could be normalised confidently
phoneSources	tel-link or page-text, one entry per phone
linkedinUrl, facebookUrl, instagramUrl, twitterUrl	One profile per network
socialProfiles	Every distinct profile found
contactsFound	true when the website gave up at least one contact
websiteStatus	ok, no-contact-found, unreachable, unreadable or no-website
websiteError	The reason a website could not be read
pagesCrawled, pagesCrawledUrls	Exactly which pages were read for that row

Fields are empty when the data genuinely is not there. Nothing is guessed.

How much does it actually find

Straight answer, because it decides whether this is useful to you.

Measured on 2026-08-19 against three real searches (plumbers in Lyon, dentists in Manchester, Zahnarzt in Berlin): 80 businesses, 77 of which had a website, 75 distinct websites.

	France	United Kingdom	Germany	All
Websites read successfully	28 of 29	19 of 19	19 of 19	73 of 75
At least one email	59%	68%	100%	73%
At least one phone from the site	86%	100%	100%	91%
At least one social profile	28%	84%	53%	48%
Any contact at all	90%	100%	100%	93%

Per delivered row, counting the businesses with no website at all: 71% came back with an email address and 90% with some contact from the website.

Germany hits 100% because the Impressum is a legal requirement and it always carries a mailbox. France is the hardest of the three: a good number of small trades publish a contact form and a mobile number and no address at all.

Two things worth knowing before you buy:

• Averages 1.4 pages per website. It reads the homepage, and only goes looking further when there is no address on it. That is deliberate: it is somebody else's small business site.
• This trade mix is website-heavy. Plumbers and dental practices nearly all have a site. Search market stalls or barbers and the "no website" share will be much larger, and those rows come back with their Maps details only.

What is deliberately not extracted

• Addresses assembled by JavaScript. Some sites build info@… from string fragments in a script. Reading those means running the page, which this actor does not do. Written-out obfuscations are covered: name [at] domain [dot] fr, (at), {at}.
• Share buttons. A link to facebook.com/sharer is not the business's Facebook page, and a link to instagram.com/p/… is a post, not a profile. Only real profile URLs are kept.
• Form placeholders. votre@email.fr, you@example.com and the rest of the theme defaults are filtered out, along with retina image artefacts like logo@2x.png, which parse as perfect email addresses and are not.
• Numbers scraped from running text that could not be checked. A phone in a tel: link is trusted as published. A number found in the page text is kept only when it normalises to a valid international number for that country, so a SIRET or a VAT number never turns up in the phone column.

Use it for

• Cold email lists. Search a trade and a town, filter to rows where email is set, and you have a list nobody had to be paid to type up.
• Multichannel prospecting. linkedinUrl plus email plus a phone in one row, so an outreach sequence has all three touchpoints from the start.
• Filling gaps in a CRM. Search the same trade and town you already sell to, join on phone or websiteDomain, and take the emails you were missing.
• Agency prospecting. Filter to websiteStatus of no-website for businesses with nothing to sell them but a website, or to no-contact-found for the ones whose site is impossible to contact through.
• Local market research. Every row carries the category, the rating and the social profiles, so you can see who is actually marketing themselves.

Input

Three fields. Nothing else to configure.

{
    "searchQueries": ["plumbers in Lyon"],
    "maxResults": 10,
    "country": "fr"
}

Several searches in one run, up to ten:

{
    "searchQueries": [
        "plumbers in Lyon",
        "electricians in Villeurbanne",
        "roofers in Lyon"
    ],
    "maxResults": 60
}

A German list, where the Impressum makes the email hit rate highest:

{
    "searchQueries": ["Zahnarzt in Berlin", "Zahnarztpraxis in Potsdam"],
    "maxResults": 120,
    "country": "de"
}

Put the town in the search text, the way you would when typing into Maps. The country field is only there for the case where your search has no place in it at all.

Pricing

Pay per result, and only for the work that produced something.

Event	When it fires	Price
business-row	One business delivered, with its Google Maps details	$0.004 ($4 per 1,000)
contact-found	That business's website gave up an email, a phone or a social profile	$0.005 ($5 per 1,000)
actor-start	Once per run	$0.001 per run

So 1,000 businesses of which 400 hand over a contact cost $4.00 + $2.00 + $0.001 = $6.00. business-row and contact-found both carry the 10 / 20 / 30 % Apify Store discount tiers (Bronze / Silver / Gold and up); actor-start is flat.

A business whose website was down, or had no contact anywhere on it, bills as a business row and nothing more. You never pay the contact rate for a row with no contact in it.

Platform usage is included, no subscription. There is nothing else to pay, no API key to buy, and no proxy to configure.

FAQ

Is this a Google Maps scraper? It contains one, and the Maps half of every row is what a good Maps scraper gives you. The reason to use this one is that it then goes and reads the business's website.

How many pages of each website do you load? The homepage, and up to three more if the homepage carried no email address. Ranked by how likely they are to hold a contact, which puts "mentions légales" and "Impressum" first, then "contact", then "about". In the measured set the average came out at 1.4 pages per site.

Why does a business come back with no email? Three different reasons and the row tells you which. no-website means Google holds no website for it. unreachable means the site did not answer, and websiteError says what happened. no-contact-found means the site answered and genuinely has no address, no phone and no social profile on the pages that were read, which usually means a contact form and nothing else.

Can I trust the phone numbers? The ones with phoneSources of tel-link are numbers the site owner marked as dialable. The page-text ones were found in the page text and kept only after they normalised to a valid international number for that country. Everything that did not normalise was dropped rather than shipped with a caveat.

Are advertisements included? No. Google returns two paid placements per page and every one of them is filtered out before any website is read. You never pay for an advert.

How many results can I get per search? Up to 120. Google Maps itself stops producing results for most local searches somewhere between 60 and 120, and the actor stops as soon as Google does.

Do you verify that the emails are deliverable? No. They are the addresses published on the businesses' own websites, as published. No mailbox is probed and no address is guessed from a pattern.

What about businesses that share a website? A chain with several branches in one search has its website read once, and every branch row gets the result. Those rows carry fromCachedDomain: true.

Related actors

• Google Maps Verified Leads if you want the same Maps search checked against the official company register instead: registered legal name, registration number, still trading or insolvent.
• B2B Lead Verifier if you already have a list and want it checked against SIRENE, Companies House and Handelsregister.
• French Company Contacts for contact details starting from a company name or a SIREN rather than from a map search.

Support

Open an issue on the actor's Apify Store page. Issues are answered within a day or two.

---

google maps email extractor, google maps contact extractor, google maps scraper with emails, business email finder, local business leads, email scraper from website, contact details scraper, b2b lead generation, linkedin facebook instagram profiles, mentions légales email, impressum email, cold email list, lead list building, google maps leads