Skip to content
ChangelogBatches (v1)

November 12, 2026

Google Maps: Category Match by an AI classifier, 43 more Search Regions, 6 more Search Languages, whole-area tasks estimated before the run; LinkedIn post share links and French post ages

The v1 API, for batchesSquids, tasks and runs: many calls on a schedule, delivered where you work. For one call that answers in seconds, use the API v2.
November 12, 2026

Google Maps: Category Match by an AI classifier, 43 more Search Regions, 6 more Search Languages, whole-area tasks estimated before the run; LinkedIn post share links and French post ages

New

Google Maps Leads Scraper: Category Match by an AI classifier

For a search and each Google Maps category, a classifier answers how likely a business listed under that category is one you are looking for. A place's category_match_score is its best category's; a place named with the searched words, whole and in order, scores 1 (a brand, a chain). Category Match keeps the places at 0.5 or more. See Google Maps Leads Scraper.

Google Maps Leads Scraper: whole-area tasks are estimated before the run

A task over a whole area (a search and an area) gets its own estimate, read from the search's count in the country and the area's share of its places, instead of the default 100 rows. A search that names a brand is counted by its name, and an estimate is never below 5 places.

Improved

Google Maps Leads Scraper: 43 more Search Regions, and every region searches its own country

15 regions were spelled differently in the list and in the engine ("Türkiye", "Côte d'Ivoire", "Bosnia & Herzegovina"...) and searched Google as the United States. Each now searches its own country, and 43 regions Google serves are added: Monaco, Réunion, Martinique, Guadeloupe, New Caledonia, Bermuda...

Google Maps Leads Scraper: 6 more Search Languages

English (United Kingdom), English (Australia), Français (Canada), Chinese (Hong Kong), Serbian (Latin) and Belarusian: 80 of the languages Google Maps serves.

Google Maps Leads Scraper: categories matched as Google names them

A category from Google's list is matched word for word, British or American spelling alike (Hardware shop and Hardware store, Petrol station and Gas station). A listed category almost no place carries ("Moving company", where Google files movers as "Mover") is compared by meaning, as before.

Google Maps Leads Scraper: small dense areas run in parallel

An area of a few km² started from one window, searched one after the other. It now starts from up to 16 windows searched at once: the City of London, Shibuya, Monaco, Le Marais. An area across 180 degrees (the United States with the Aleutians, Russia, New Zealand) starts on its own land, where it returned 0 results.

Google Maps results: one row per business on a run without emails

A run without email extraction could return a business twice or more, once per email other runs had found for it. It now returns one row per business. A run with emails keeps one row per email.

Google Maps emails: contact form placeholders left out

Addresses like jane@email.com or jean.dupont@entreprise.fr, the examples written in contact forms, are no longer returned as a business's email.

LinkedIn Posts Commenters and Likers: share links and hidden posts

A share link (lnkd.in/...) is followed to its post, where it returned 0 rows; a URL with no post ends invalid_url. A post your account cannot see ends access_restricted instead of reading as a post with no comment. See LinkedIn Posts Commenters and Likers Scraper.

Fixed

LinkedIn Posts Scraper: Fetch Since on an account in French

A LinkedIn account in French shows a post's age as "3 mois", "2 sem", "5 j": posted_at stayed empty and Fetch Since kept every post. French ages are now read. See LinkedIn Posts Scraper.

File uploads to Google Maps Leads Scraper

Every CSV or TSV sent to Google Maps Leads Scraper failed with "something is wrong", in the app and through Upload Tasks. Uploads work again.

Capterra Reviews Scraper: products with more than one page of reviews

Every product past its first page of reviews failed. Its reviews come back, page after page.

Google News: articles with very long URLs

An article whose URL ran past 2,000 characters stopped the run. It is kept without its query string, which only named the site's experiments.

Contact Details Scraper: phones with hidden characters

A phone written with invisible characters on a site failed the result. They are now removed and the phone kept.

A run starting is not launched twice

A run could be started again in the 20 seconds before it said it was running, and two workers could take the same task. It now starts once.