The extraction glossary

Software that reads publicly displayed business listings from Google Maps and exports them as structured data (CSV or Excel): name, phone, website, address, rating, review count and category. Also marketed as a Maps data extractor, GBP scraper or lead extractor; the names vary, the job does not.

The practical ceiling on a single Google Maps search: roughly 120 listings are returned regardless of how many businesses exist in the area, because the limit applies per query rather than per city. Complete coverage requires issuing many smaller area-bounded queries and merging the results.

The technique that defeats the per-query ceiling: divide the target area into a grid of cells, search each cell separately, then merge and de-duplicate the results. Coverage scales with cell count, and dense cells are subdivided further when their result counts approach the cap.

Lead Finder’s implementation of adaptive grid extraction: automatic subdivision of the search area, per-cell progress checkpointing, and place-ID de-duplication during the sweep, so a metro extraction survives interruptions and returns each business once.

Google’s stable identifier for a specific listing, independent of the business’s name, phone or address. The correct key for de-duplication and for joining one extraction against a later one; name-based matching fails in both directions and phone-based matching fails on shared numbers.

The system businesses use to manage their listing, formerly Google My Business (GMB); Google Maps is where the profile is displayed. Scrapers read the profile’s public face only: dashboard data such as insights, messages and the owner’s account email is private and not extractable.

A business that serves customers at their location (plumbers, mobile locksmiths, cleaners) and may hide its street address on the listing. SAB rows legitimately carry blank or approximate addresses; coordinates and phone remain the reliable fields for territory and contact work.

A listing whose business has verified ownership through Google Business Profile, typically bringing maintained hours, photos, replies and a website link. Unclaimed listings are auto-generated from public data and often skeletal, which makes the claimed/unclaimed distinction a strong prospecting signal for local-marketing services.

The number of Google reviews on a listing; alongside star rating, the best free qualification signal in local data. Banding lists (0-9, 10-50, 51+) separates businesses by maturity and marketing investment, and count changes between extractions flag businesses actively investing in growth.

The outreach strategy of sorting a category by review count and pitching the under-reviewed tail on visibility and reputation services, using the public gap between the prospect and nearby leaders as evidence. Reliable because the claim is specific, verifiable and about the prospect.

The international phone format (+country code then number, no punctuation: +14155550123) that diallers, WhatsApp tooling and de-duplication all expect. Normalising extracted numbers to E.164 at export time also prevents Excel from mangling them into scientific notation.

The erosion of a lead list’s contactability as businesses close, move and change numbers: roughly 1-2% per month for local data, faster in food and beauty. The countermeasure is scheduled re-extraction, which also yields the openings-and-closures delta as its own prospect source.

The two-stage process behind every “Maps email extractor”: take the website URL from the listing, crawl that site for a published address. Maps itself has no email field, so realistic coverage is 20-40% of rows and claims of near-complete coverage indicate guessing or database resale.

Google’s official, licensed route to place data, billed per request. Stable and correct for product integrations; for lead lists it is constrained by cost at prospecting volume, a roughly 60-result cap per Nearby Search, and storage terms restricting retention of fetched data.

The map-plus-three-listings block Google shows for local-intent searches. Ranking in it is the commercial prize local SEO chases, and extracted rating/review data across a whole category is how consultants find businesses losing pack visibility to better-reviewed neighbours.

The pricing model where extraction software costs a fixed amount per period with unlimited rows, as against per-row metering. Above roughly 10,000-25,000 rows a year the flat model is cheaper, and it changes behaviour: complete sweeps and quarterly refreshes stop carrying marginal cost.