Skip to content

Guide

How to import customers from a spreadsheet into field service software

Moving your customer list is mostly cleanup, not typing. Fix the sheet first, test with ten rows, then import the rest. Here is the order that avoids duplicates and bad phone numbers, and how CalSpark handles each step.

By CalSpark team. Published . Updated .

Clean the sheet before you import anything

Most import problems start in the spreadsheet, not in the software. Twenty minutes of cleanup saves an hour of fixing records later.

  • One row per customer. If a customer appears on three rows, merge them or decide which row wins.
  • Split names. Separate first name and last name columns are the cleanest. A single full name column can work if the software splits it for you.
  • Pick one phone format. Ten digits with no brackets or dashes is the safest.
  • Put the street, city, state and ZIP in their own columns, not one long address cell.
  • Keep the gate codes, dog warnings and access notes in a notes column. Dispatchers need these on the day.
  • Delete blank rows and test entries.

A customer with several properties needs care. CalSpark customers can hold multiple addresses, but the import creates one customer with one address per row. For a landlord with six buildings, import the main address and plan to add the others by hand afterward. If you have a lot of these, check how the import treats repeat rows on a small test first.

Getting the list out of where it lives now

Almost every tool can export customers to a CSV file. QuickBooks, Jobber, Housecall Pro and Google Contacts all have some way to do it, and Excel can save any sheet as CSV. The menus change from version to version, so search the tool's help for "export customers" rather than trusting a screenshot from a blog post.

Export more columns than you think you need. You can ignore columns at the mapping step, but you cannot add one that was never exported. Open the file once before you move on and look at it. If the phone numbers show as 5.55E+09, Excel has mangled them, and you should format that column as text and export again.

What the CalSpark customer import does

The importer lives in the customers area and walks you through five steps.

  1. Choose your CSV file. It is parsed in your browser.
  2. Map your columns to CalSpark fields: first name, last name or a full name that gets split, company, email, phone, street, city, state, ZIP and notes.
  3. Preview the result. You see counts of duplicates and invalid rows before anything is saved.
  4. Import. A progress bar shows how far along it is.
  5. Read the report: created, skipped and failed. Failed rows come in a downloadable error CSV.

Duplicates are found by matching email or phone against customers already in CalSpark and against earlier rows in the same file. The import runs one row at a time, so it suits a few hundred customers. It is not built for tens of thousands. If your list is that large, trim it first (see below).

Test with ten rows first

Copy ten representative rows into a new file. Include a company, a person with no email, a messy phone number and one obvious duplicate. Run the whole import. Then open each created customer and check the address, phone and notes.

What to check after the test import
CheckWhat good looks like
NamesFirst and last name in the right places, company filled where it should be
PhoneSame number as the sheet, no missing digits
AddressStreet, city, state and ZIP all landed in the right fields
DuplicatesThe planted duplicate was skipped, not created twice
Error CSVFailed rows list a reason you can fix in the sheet

If the test is clean, import the rest. Delete the test customers first if you do not want them counted as duplicates. For a deeper look at how customers fit into scheduling, see customer and job scheduling.

What not to import

A new system is a fresh start. Do not carry over every name you have ever typed.

  • Dead leads who never became customers
  • Customers you have not served in many years, unless you actively market to them
  • Vendors, suppliers and your own staff mixed into the customer list
  • Records with no phone, no email and no address

A rule of thumb: if you would not call them for a job tomorrow, leave them in the old file. You can always import a second batch later. When the list is loaded, the next job is learning the board, so read the dispatcher daily checklist. To see everything CalSpark does today, look at features, or start a trial and test with your own file.

Questions

How many customers can I import at once?

The import processes one row at a time, so it is meant for a few hundred customers. It is not built for tens of thousands. For a very large list, import your active customers first and add older records only if you need them.

Does CalSpark remove duplicates?

It detects them, by matching email or phone against existing customers and against earlier rows in the file. You see the duplicate count in the preview before importing, and skipped rows are reported afterward.

Can I import a customer with several addresses?

Customers can hold multiple addresses, but the import creates one customer with one address per row. Import the main address, then add the extra properties to the customer afterward.

What happens to rows that fail?

Failed rows are counted in the final report, and you can download an error CSV listing them. Fix those rows in your sheet and import just that file again.

Do I have to use CSV?

Yes, the importer reads CSV files. Excel and Google Sheets can both save or download a sheet as CSV, so any spreadsheet works once you export it.

Try it with your own jobs

14-day free trial. No card. $49 a month flat for up to 10 users after that.