Contact and spreadsheet CSV workflows

Convert Excel Contacts to CSV Without Losing Fields

Export an Excel contact sheet to CSV while preserving phone text, field boundaries and Unicode names, then check the destination’s required column mapping.

By Uddhav Gopani, FounderPublished 27 September 2026Updated 29 September 2026

Convert Excel Contacts to CSV Without Losing Fields

To convert Excel contacts to CSV, arrange one contact per row, map the columns to the receiving app’s expected fields and export the prepared worksheet as a CSV file. Saving the extension alone is not enough: a structurally correct CSV can still put a phone number into the wrong contact field if its headers or mapping are unsuitable.

Choose the receiving contact application first

Identify where the contacts will be imported before changing the workbook. A generic customer directory, Google Contacts and a desktop address book may use different column names or provide different mapping screens. Use that destination’s template when one exists, and retain a copy of the original workbook.

For example, the generic header Phone describes a column to a person, but it is not a promise that every importer will recognize it. The Google Contacts formatting guide explains that particular destination’s numbered field headers. Do not apply those headers indiscriminately to unrelated applications.

Make the worksheet a plain contact table

Use one header row followed by records with the same column order. Remove presentation-only title rows from the export copy, resolve merged cells and place notes in a notes column rather than below the table. A total or separator row can otherwise become an unwanted contact.

Keep each person’s details together. Sorting only the names column can detach names from their phone numbers while leaving a CSV that parses perfectly. Select the complete table for sorting, and compare several known name-and-number pairs afterward.

Preserve phone numbers as text

A telephone number is an identifier, not a quantity to calculate. Country-code plus signs and leading zeros can be meaningful. Prepare the phone column as text before entering or importing those values, then inspect the actual exported characters.

Changing a cell’s format after a leading zero has already disappeared does not recover the missing digit. Likewise, a custom display format can make a numeric cell look different from its stored value. Compare the CSV against the original contact source instead of assuming its spreadsheet appearance proves preservation.

Inspect a small generic CSV example

Name,Email,Phone,Notes
Asha Rao,asha@example.com,+12025550123,"Accounts, east office"
René Park,rene@example.com,0012345,"Says ""hello"""

This fictional fixture has four fields in every record. The comma in Accounts, east office stays inside one quoted field, and the quotation marks around hello are doubled inside the quoted value. René retains its accented letter. The short identifier 0012345 deliberately tests leading-zero preservation; it is not a validated real telephone number.

A CSV-aware parser should read two contacts, not treat the comma inside the note as a fifth column. Splitting each line on every comma would misread this example. Use a CSV-aware preview or CSV Viewer to inspect field boundaries with non-sensitive sample data.

Export the intended worksheet

Microsoft’s CSV export instructions use Save As with a text or CSV file type. The export covers the current worksheet, so select the contact sheet and export other sheets separately if required. Keep the workbook version for formulas and formatting that a CSV cannot retain.

Choose a Unicode-compatible export option for names outside plain ASCII and check the resulting file’s encoding. Also inspect the actual delimiter: a file named .csv can be produced with a regional separator that the receiving app does not expect. The destination’s requirements govern the final file.

Check values as well as the record count

Compare the exported record count with the intended contacts, excluding the header. Inspect an accented name, a phone beginning with a plus sign, a leading-zero identifier, an empty optional field and a note containing punctuation. These cases reveal different mistakes that an ordinary name-only row can hide.

Before a large import, use the destination’s preview or a small deliberate test selection, then inspect the resulting fields. Repeatedly importing the full file can create duplicates. Correct the export or mapping first, and keep enough information to identify the specific records from a test import.