Geocoding

Geocoding converts addresses and place names into geographic coordinates that can be displayed on maps. Atlas makes it easy to upload spreadsheets with location data — simply upload and Atlas automatically detects address columns, converts them to coordinates, and shows the data on your map.

What is Geocoding?

Geocoding is the process of converting:

• Addresses ("123 Main Street, Denver, CO") → Latitude/Longitude (39.7392, -104.9903)
• Place names ("Paris", "Mount Everest") → Map coordinates
• Postal codes (80202) → Representative location points

Once geocoded, your location data becomes mappable and analyzable on a geographic basis.

Supported File Formats

Geocoding works with:

• CSV files (.csv)
• Excel spreadsheets (.xlsx, .xls)

How Geocoding Works

Automatic Detection

When you upload a spreadsheet, Atlas analyzes column headers and data patterns to automatically identify location information. The system recognizes:

• Street addresses in single columns ("123 Main St, Denver, CO") or multiple columns (street, city, state, zip)
• City and country names ("London", "Tokyo", "United Kingdom")
• Place names and landmarks ("Eiffel Tower", "Central Park")
• Coordinate pairs (latitude/longitude columns)
• Alternative coordinate systems:
  • UTM coordinates (Universal Transverse Mercator)
  • WKT geometry strings (Well-Known Text)
  • H3 hexagonal indexes
  • MGRS coordinates

In the upload preview, geocoded columns are highlighted with visual indicators showing which fields contain location data.

Supported Location Data Formats

Full addresses (single column):

123 Main St, Denver, CO 80202, USA
456 Oak Avenue, London, UK SW1A 1AA
789 Elm Road, Sydney, NSW 2000, Australia

Separate address components (multiple columns):

Place names:

Paris
New York City
Mount Kilimanjaro
Golden Gate Bridge

Coordinates (latitude/longitude):

40.7128, -74.0060  (New York)
51.5074, -0.1278   (London)
35.6762, 139.6503  (Tokyo)

Other formats:

• UTM: 33U 500000 5000000
• WKT: POINT (-104.9903 39.7392)
• MGRS: 13SDG000000

Quick Start: Geocoding a Spreadsheet

1. Prepare your spreadsheet with address, city, or location name columns
2. Click "Add Data" in the Layers panel
3. Select the "Upload Data" tab
4. Drag and drop your CSV or Excel file (or click to browse)
5. Review the preview table:
   • Atlas highlights detected location columns
   • Verify the correct columns are selected
   • Adjust column mapping if needed (click the column selector)
6. Click "Add" to upload and geocode
7. Wait for processing — Atlas geocodes all locations in the background
8. Review on the map — geocoded points appear as a new layer
9. Check the data table — click the layer to inspect which rows were successfully geocoded

Your original spreadsheet data is preserved — geocoding adds coordinate columns automatically.

Geocoding Accuracy & Quality

Factors Affecting Accuracy

Address completeness is the primary factor:

• Complete addresses (street + city + state/province + country) = highest accuracy
• Partial addresses (city only, or "Main St, Colorado") = moderate accuracy, may be ambiguous
• Place names only ("Denver") = lower accuracy for disambiguation

Examples:

• ✓ "123 Main St, Denver, CO 80202, USA" → High confidence
• ◐ "Main St, Denver" → Medium confidence (which Main St?)
• ◐ "Denver" → Low specificity (downtown or suburbs?)

Verifying Your Geocoded Data

After upload, always review your results:

1. Check the map — Do points appear in the correct locations?
2. Spot-check the data table — Click a few rows and verify the addresses match the geocoded locations
3. Look for outliers — Points appearing far from expected locations may be:
   • Incorrectly formatted addresses
   • Ambiguous place names (e.g., "Oxford" exists in UK and USA)
   • Typos in the original data

Rows That Failed to Geocode

Rows without valid location information or that couldn't be matched are:

• Still imported into your dataset
• Visible in the data table without coordinates
• Not displayed on the map

To find them:

1. Click the layer to open the Data Table
2. Filter by geometry or sort by location to find empty coordinate rows
3. Manually fix addresses in the data table and try geocoding again

Best Practices for Successful Geocoding

Prepare Your Data

• Use consistent formatting — Don't mix "Street Address", "street", and "addr" as column names; pick one
• Spell country names correctly — "United States" vs. "USA" vs. "US" all work, but typos ("Unites States") will fail
• Complete addresses when possible — Include city, state/province, and country, not just street address
• Standardize postal codes — Format consistently (e.g., all uppercase for UK postcodes)

Column Naming

Atlas detects columns better when they follow standard naming conventions. Use:

• address or street_address for full addresses
• street, city, state, zip, country for components
• latitude, longitude for coordinates
• place_name or location for place names

Common Mistakes to Avoid

• Mixing address formats (some rows have "Street Address", others have "Address")
• Missing country information for international addresses
• Abbreviating inconsistently ("St" vs. "Street", "CA" vs. "California")
• Uploading files with address columns hidden or formatted as formulas

Performance and Processing

Optimizing Performance

• Split large files into smaller batches before uploading
• Upload during off-peak hours if available
• Use high-quality internet connection for faster uploads
• Prepare data locally to reduce upload time

Troubleshooting

Address Columns Not Detected

Problem: Atlas didn't highlight your address columns in the preview.

Solutions:

• Rename columns to standard names: address, street, city, state, zip, country
• Ensure the column has actual address data (not formulas or images)
• Check for leading/trailing spaces in column headers
• Try manually selecting the column using the column selector dropdown

Incorrect or Unexpected Locations

Problem: Some points appear in wrong locations.

Causes:

• Ambiguous addresses (e.g., "Main St, Texas" — which town?)
• Abbreviations or misspellings ("Denver" instead of "Denver")
• Missing country information for international addresses
• Different address formats in different rows

Solutions:

1. Review the original addresses for typos
2. Add city/state/country to clarify ambiguous addresses
3. Use consistent formatting throughout the spreadsheet
4. Check the data table for rows that failed to geocode
5. Manually correct and re-upload if needed

Some Addresses Not Geocoding

Problem: Many rows are missing coordinates.

Causes:

• Invalid or incomplete addresses
• Outdated addresses or closed businesses
• Non-standard address formats
• Addresses in non-English languages (some regions may be less supported)

Solutions:

• Review failed rows in the data table
• Fix formatting/spelling and re-upload
• For international addresses, include the full country name
• Use external geocoding tools (Google Maps, ArcGIS) to pre-check addresses

Slow Geocoding Performance

Problem: Upload is taking a very long time.

Causes:

• Large file size (100,000+ rows)
• Poor internet connection
• High server load

Solutions:

• Wait — geocoding continues even if the browser appears stuck
• Split the file and upload in chunks
• Try uploading at a different time
• Check your internet connection speed
• Contact support if uploads consistently timeout

Advanced: Manual Coordinate Entry

If you already have latitude/longitude coordinates, you don't need geocoding — just upload the file directly:

1. Ensure your spreadsheet has latitude and longitude columns
2. Upload the file as usual
3. Atlas detects the coordinates automatically and skips geocoding

This is faster for pre-geocoded data.

Next Steps

• Customize layer appearance — See Layer Styles to change marker colors and sizes
• Add popups — See Popups to show address details when features are clicked
• Filter data — See Data Table to filter and analyze geocoded locations
• Download results — See Download Data to export geocoded coordinates in GIS formats

Getting Help

For persistent geocoding issues, see the Troubleshooting Guide or contact support at help@atlas.co.