Google Geocoding Batch Scraper

Convert a bulk list of addresses to latitude/longitude — or coordinates back to clean, normalized addresses — using the official Google Geocoding API. Stop running one-off cURL calls; paste your list and export.

What you get

• Forward geocoding — address → lat/lng, placeId, formatted address, and split components (street, locality, country, postal code, etc.)
• Reverse geocoding — lat/lng → formatted address, placeId, and the full component breakdown
• One dataset row per result — Google sometimes returns multiple candidates per input; each one is its own row with a rank
• All address components extracted — street number, route, neighborhood, locality, admin areas, country, country code, postal code, plus code
• Per-input error handling — bad inputs get a row with status set to INVALID_INPUT, ZERO_RESULTS, etc.; the run continues
• Region & component biasing — pin forward geocoding to a country, restrict by admin area, or filter reverse geocoding by location type

Use cases

• CRM cleanup — turn a column of free-text addresses into normalized fields with lat/lng for mapping
• Real estate / property data — resolve street addresses to coordinates for distance, school zone, and amenity analysis
• Logistics & routing — pre-geocode delivery addresses before bulk routing
• Reverse-lookup analytics — turn GPS pings into city/neighborhood for cohort analysis
• Address validation — flag inputs that return partial_match: true or ZERO_RESULTS for cleanup

How to use

1. Pick a Mode:
   • Forward — address strings, one per line.
   • Reverse — lat,lng pairs (e.g. 40.7128,-74.006), one per line.
2. Paste your list into Inputs.
3. (Optional) Set Language for the formatted address (en, es, fr, …).
4. (Forward only, optional) Set Region to bias by country (us, gb, de) and Components to constrain (country:US).
5. (Reverse only, optional) Set Result type and Location type filters.
6. Run — every result becomes its own dataset row.

Forward vs reverse examples

Forward input:

1600 Amphitheatre Pkwy, Mountain View, CA
221B Baker Street, London

Reverse input:

37.4220,-122.0841
51.5237,-0.1585

Output schema

Field	Type	Description
input	string	Echo of the original input
mode	string	forward or reverse
rank	number	1-based rank when Google returns multiple candidates
status	string	OK, ZERO_RESULTS, INVALID_INPUT, OVER_QUERY_LIMIT, REQUEST_DENIED, INVALID_REQUEST, ERROR
formattedAddress	string | null	Google's canonical address string
latitude / longitude	number | null	Coordinates
placeId	string | null	Google place ID — stable identifier
locationType	string | null	ROOFTOP, RANGE_INTERPOLATED, GEOMETRIC_CENTER, APPROXIMATE
partialMatch	boolean | null	true if Google guessed at part of the input
types	string[]	Place types (e.g. street_address, locality)
streetNumber	string | null	Street number component
route	string | null	Street name
neighborhood	string | null	Neighborhood
subLocality	string | null	Sub-locality (e.g. NYC borough)
locality	string | null	City
adminArea1Long / adminArea1Short	string | null	State / region (full and abbreviated)
adminArea2Long	string | null	County
country	string | null	Country name
countryCode	string | null	Two-letter country code
postalCode	string | null	ZIP / postal code
plusCodeGlobal	string | null	Global plus code
plusCodeCompound	string | null	Compound plus code
error	string | null	Error message if status is not OK

Tips

• Use Components (country:US) to prevent ambiguous addresses from matching the wrong country (e.g. "Springfield" matches dozens of places worldwide).
• locationType matters — ROOFTOP is a precise address match; APPROXIMATE means Google fell back to a city center.
• partialMatch: true is a strong signal the input had typos or missing parts — flag these for human review.
• Google rate-limits aggressive batches; if you see OVER_QUERY_LIMIT consistently, slow down or raise the quota in Google Cloud.
• For reverse geocoding of a point with multiple result types, the first row is the most specific (street address) and later rows broaden up to country.