You don’t need to write any code to map a CSV file. Upload it to a mapping tool, and the tool reads your address columns and drops a pin for every row. For a few hundred addresses, BatchGeo is the fastest choice. Google My Maps costs nothing as long as your file stays under its 2,000-row import cap. If you have tens of thousands of rows, or you need to route or report on your pins, a paid business tool such as Felt, Maptive, eSpatial or Mapline is the better fit.
How mapping tools read a CSV file
All of these tools handle a CSV in roughly the same way. They scan the column names, pick out the ones that hold a location, and pass those values to a geocoder, which is the service that converts an address into map coordinates.
A geocoder only has the contents of those columns to go on. That makes a tidy header row and complete addresses more important than your choice of tool.
Pin accuracy comes down to what each row contains:
- With a full street address, the pin lands on the building itself.
- With only a city or ZIP code, the pin lands somewhere within that area. That works for a regional overview and fails for a delivery list.
- With latitude and longitude columns, there’s no geocoding step, since the file already has the coordinates.
Prepare your CSV before you upload it
Exports from a CRM or a survey platform tend to come out in the correct format. Hand-assembled files often don’t, so give yours a check first.
- Confirm that row one contains the column names.
- Put only one field in each column.
- Get rid of merged or empty header cells. Upload screens rely on header names to work out which columns contain the street, city, state and ZIP code.
- Enter full street addresses for any row where the pin has to sit on a specific building.
- Count the rows and check that number against the limit of the tool you intend to use.
The tools differ widely on row limits:
- Google My Maps: don’t import files of more than 2,000 rows.
- Felt: geocodes up to 10,000 addresses per upload by default.
- eSpatial: 50,000 records per dataset on Mapping Pro, 100,000 on higher plans.
- BatchGeo: its own pages list two different figures for a Pro map, so try a large file before paying.
Map a short address list fast with BatchGeo
No tool gets a CSV onto a map faster than BatchGeo. If you have a file of a few hundred addresses to turn into pins, it usually covers the job.
There are three ways to load data: drag and drop the CSV, choose the file, or paste rows copied directly from a spreadsheet. BatchGeo geocodes addresses, intersections, cities, states, postal codes and coordinates.
Every saved map has its own link for sharing, and the finished map can be embedded on a web page. Grouping by a column lets you color and filter the pins.
To change a saved map, open the edit link BatchGeo sends by email and save the updated data again.
You can try the free plan without an account. It gives you up to 1,000 geocodes a month and 100 map views, and maps are kept for 30 days. The Pro plan costs a flat $99 a month and includes 10 users, password protection, a heat map layer, and export to PDF, PNG and KML.
Share and co-edit a map for free with Google My Maps
My Maps, Google’s free tool for building custom maps, works well when several people need to work on the same CSV. Co-editing a map works like co-editing a Google Doc, and you can save the map to Google Drive and embed it on a website.
It brings CSV, TSV, XLSX and Google Sheets files in as layers. It takes latitude and longitude, addresses, or WKT, a text format for describing points, lines and areas.
The limits:
- A single map holds up to 10,000 places.
- The 2,000-row cap applies per imported file.
- Any imported file other than KML can be up to 40 MB.
The analysis goes no further than styling pins by a column. When you swap a layer for a newer copy of the file, those style rules stay in place.
Map ZIP codes, counties and census tracts with Felt
Felt takes CSV, Excel and Numbers files and geocodes the addresses inside them. It also handles files that contain no street addresses.
Given a column of states, counties, ZIP codes or census tracts, Felt can match it to Census boundary files and produce shaded areas in place of points. ZIP codes are matched as ZCTAs, which are the Census Bureau’s approximations of ZIP code areas.
Each layer can contain up to 50 million features, and uploads through the browser can reach 5 GB. Mapping a CSV requires the Professional plan or its 7-day free trial, and all plans include 10 full seats.
Clean up messy exported files with Maptive
Maptive is designed for exports that come out messy. A location column can contain full street addresses, cities, bare ZIP codes or coordinates. According to Maptive, its engine fixes garbled addresses and fills in missing entries in the ZIP column during upload.
Maps can be exported as images, and you can send the rows back out as a spreadsheet file, CSV included, either all of them or a filtered selection.
The main constraint is file size. On every self-serve plan, a single data set tops out at 100,000 locations. There’s also a separate limit on total locations geocoded, ranging from 150,000 on the one-time pass to 1,000,000 on Team. Markers per map have their own count, with a maximum of 200,000.
Map CRM exports and sales territories with eSpatial
eSpatial is a sales territory and mapping tool, and it fits a CSV that began life as a CRM export. Its spreadsheet upload maps account lists from Excel or other exports.
Teams that would prefer a direct connection over a file can buy a Salesforce integration as a $5,995 add-on.
Mapping Pro handles 50,000 records per dataset and 100,000 geocodes a month, with a daily import limit of 200,000 records. All three go up on higher plans. Mapping Pro also comes with 10 Auto-Sync Tasks that keep a dataset up to date after the initial upload, and it draws drive-time and radius polygons around the plotted accounts. The 7-day trial doesn’t require a credit card.
Turn uploaded stops into routes with Mapline
Mapline is a platform for web mapping and routing. It accepts a CSV or Excel upload, or rows pasted from a spreadsheet, and it can build routes from the uploaded stops.
You get one route on the base paid plan. Pro adds multi-vehicle routing.
Pin-density heat maps reveal where your list bunches up, and territories drawn around the uploaded pins can also be colored as a heat map. Mapline doesn’t publish its row limits. Its free Starter plan never expires and doesn’t ask for a card.
Fix rows that won’t map
Most of these tools give you a report of rows they couldn’t place. Begin there.
- Open the tool’s list of rows it failed to place.
- Check for the common culprits: typos, missing ZIP codes, and suite numbers stuffed into the street field.
- Correct those mistakes in the spreadsheet instead of on the map. Fixing pins one by one takes longer.
- Upload the fixed file.
When an address is beyond repair, step down a level. A row with a valid ZIP code or city will still map, with less precision, and that’s often good enough for regional counts.
If the list needs to be exact, add latitude and longitude columns from a source that already holds them, such as your CRM or a field app, and upload the file again.
Tools that match ZIP codes or counties to boundaries can count those rows toward area totals without placing a point.
Frequently asked questions
How many rows can I import into Google My Maps?
Google advises against importing files of more than 2,000 rows. One map can hold up to 10,000 places across all its layers.
Can I try BatchGeo for free?
Yes. You don’t need an account for the free plan, which gives you up to 1,000 geocodes a month and 100 map views. Maps are kept for only 30 days.
Can I map a CSV that has ZIP codes or counties but no street addresses?
Yes. Felt can match a column of states, counties, ZIP codes or census tracts to Census boundary files. Other tools will also map a row that has only a ZIP code or city, but less precisely than a full address.
Which tool handles the largest files?
Felt supports up to 50 million features per layer and browser uploads of up to 5 GB. Maptive limits a single data set to 100,000 locations on self-serve plans, and eSpatial allows 100,000 records per dataset on its higher plans.