CSV or XLSX? Picking a file format for data imports

For most business data imports, either format works. Pick CSV when the data is plain and you want to see exactly what the importer will read. Pick XLSX when your data lives in a spreadsheet already and contains accents, semicolons or codes a CSV export might mangle.

The choice sounds technical, but it decides where your import goes wrong, if it does. A CSV fails in visible ways: a broken character, a column split in two. An XLSX fails in hidden ways: a formula, a date stored as a number, a second sheet nobody noticed. Knowing which trap you are walking toward is most of the work.

What each format actually is

A CSV (comma-separated values) file is plain text. Each line is a row, and a separator splits the cells. There is no formatting, no formula and no second sheet. Open it in a text editor and you see every character the importer will see.

An XLSX file is a spreadsheet workbook: a zipped package that stores typed cells, formulas, formatting and any number of sheets. What you see on screen is the cell's display format, which is not always the value stored underneath.

	CSV	XLSX
What is inside	Plain text, one row per line	A workbook with typed cells and sheets
Separator	Comma, semicolon or tab, depending on who saved it	None needed
Accents and symbols	Depends on the encoding you save in	Stored safely
Leading zeros and long codes	Kept in the file, often lost when a spreadsheet opens it	Kept if the cell is text
Hidden problems	Few: what you see is what is there	Formulas, merged cells, hidden rows, extra sheets
Easy to check by eye	Yes, in any text editor	Not without a spreadsheet app

When CSV is the better choice

• Your old tool exports CSV. Many business tools export records as CSV. Importing that file as it is, without opening and resaving it, avoids a round of reformatting.
• You want to inspect the data. A text editor shows the exact separator, quotes and characters. Nothing is hidden behind a display format.
• The data is simple. Names, emails, addresses and plain numbers travel well as CSV.

The usual CSV failures are encoding and separators. Save as UTF-8 so accented names survive. Know that many European spreadsheet apps write a semicolon instead of a comma, because the comma is their decimal mark. And avoid double-clicking a CSV to open it: the spreadsheet then guesses each column's type, which can drop leading zeros, shorten long numbers or turn a code into a date.

When XLSX is the better choice

• Your data already lives in a spreadsheet. Exporting it to CSV adds a step where accents, separators or codes can break.
• The data has awkward characters. Commas inside addresses, semicolons in notes, or a mix of languages need no quoting rules in a workbook.
• Codes must stay text. Format the column as text, and a customer number like 00412 stays 00412.

The usual XLSX failures are the things you can't see. A cell with a formula, a merged header, a hidden row or a second sheet can all surprise an importer. Dates deserve care too: a date cell stores a serial number behind its display format, and some importers read that number rather than the date you see. Typing dates as text in year-month-day form, such as 2026-08-13, avoids the question.

A quick checklist for either format

1. One record type per file. Customers in one file, their contacts in another.
2. One header row. Row one holds the column names, exactly as the importer expects them, and nothing sits above it.
3. One record per row. No subtotals, blank spacer rows or notes at the bottom.
4. Plain values. Replace formulas with their values, and unmerge merged cells.
5. Codes as text. IDs, postal codes and phone numbers in text columns, so nothing is rounded or trimmed.
6. A small test. Import ten rows before the full file. Errors in ten rows are quick to read.

The guide to cleaning your data before an import covers duplicates and missing values in more depth.

How it works in Swiftly Workspace

In Swiftly Workspace, Data Import sits in Administration and takes either a CSV or an XLSX file. Both arrive as the same table, so the rules are the same:

• Column names come from the template. Download CSV Template gives the exact column names for a record type. Fill it in as a CSV, or open it in your spreadsheet app and save it as XLSX; the header row works the same way. If a required column is missing, the upload is refused, so you know straight away.
• CSV files are read as saved. The importer accepts commas, semicolons or tabs, and UTF-8 (with or without a byte-order mark), UTF-16 or Windows-1252, and works out which one your file uses.
• XLSX files: one sheet per file. The importer reads a single worksheet and takes row one as the column names, skipping empty rows. Keep each record type in its own file, in a workbook with one sheet.
• Nothing is saved until you commit. Rows are staged and checked so you can fix problems before they reach your records.

A good place to start is customers, then their addresses and contacts. Leave reference columns such as customer type or status empty, and set those values in the app after the import. The step-by-step guide to importing data on a Mac walks through a full session, and preparing a CSV for import covers templates and the errors to expect. Moving off spreadsheets altogether? See migrating from spreadsheets.

Frequently asked questions

Why do my accented names turn into strange symbols after an import?

The file was saved in one encoding and read in another. Save the CSV as UTF-8, or import the workbook as XLSX, where characters are stored safely.

Is a semicolon-separated file still a CSV?

Yes, in practice. The name says comma, but many spreadsheet apps in Europe write semicolons because the comma is used for decimals. A good importer detects the separator; check that yours does before you convert anything.

Can I import a workbook with several sheets?

Many importers read a single sheet and ignore the rest. Split a multi-sheet workbook into one file per record type to be sure every row is read.

Do formulas survive an import?

No. An importer stores values, not formulas. Some read a formula cell's last calculated value, but to be safe, replace formulas with their values before you upload.