Why Excel breaks your product CSV, and how to save a file your shop will import
You export products from your shop, open the file in Excel, fix a few prices and save. Then the import fails. Or worse, it succeeds, and now half your product names read “café” and every barcode ends in 000.
It isn't your fault. Excel is a spreadsheet, not a CSV editor: when it opens a CSV it guesses what every cell means, and when it saves one it writes the file the way your Windows region expects, not the way your shop does. Here is what changes, how to avoid it and how to check a file before you import it.
1. The encoding: “CSV” isn't UTF-8
Shopify and WooCommerce expect UTF-8. Excel's plain “CSV (Comma delimited)” option, though, saves in your Windows code page: Windows-1252 in Western Europe and the US, Windows-1250 in Central Europe, Windows-1251 for Cyrillic. Plain English letters survive, but every é, ü, č, ł or ж becomes a byte the shop reads differently.
On Shopify this often shows up as “Illegal quoting on line …”, even though nothing is wrong with your quotes. WooCommerce says the file is “using a different encoding than UTF-8”. The fix is to pick the type that starts with “CSV UTF-8” in the Save as type list.
2. Semicolons and decimal commas
In countries that write a decimal comma (most of Europe), Windows uses the semicolon as its list separator, and Excel's CSV follows it: 24,90;12 instead of 24.90,12. Shopify then reads each row as one long column and complains about missing headers; other importers need the separator set by hand, and 24,90 still isn't a number to them.
There's no switch for this in the Save dialog: it comes from the regional settings in Windows, not from Excel. For a single file it's easier to swap the separator and the decimal comma afterwards.
3. Garbled characters that stay garbled
The opposite happens too. Double-click a UTF-8 file that has no byte order mark and Excel reads it in your local code page: “café” turns into “café”. If you save at that point, the broken text is written into the file for good. It's now valid UTF-8 that literally says “café”.
If you see this, close the file without saving and open it through Data > From Text/CSV instead (see below). If it's already saved, the damage can usually be reversed, because it follows a fixed pattern.
4. Barcodes, EANs and other long numbers
Excel treats anything that looks like a number as a number. A 13-digit EAN such as 8580000001001 is displayed as 8.58E+12, and if you save while it looks like that, “8.58E+12” is what ends up in the CSV. A UPC or SKU that starts with 0 loses the zero. Anything longer than 15 digits has its last digits replaced with zeros, because 15 digits is all the precision Excel keeps.
None of this can be undone from the saved file: the digits are gone. Format those columns as Text before you edit, or export the products from the shop again.
5. Dates, and values that look like dates
A sale date written as 2026-11-01 comes back in your local format, such as 11/1/2026, and platforms that expect YYYY-MM-DD reject it. Sizes and codes suffer too: values like 1-2 or 3/4 can be turned into dates, and the original text is lost.
How to edit a CSV in Excel without breaking it
If you have to use Excel, import the file instead of double-clicking it, and save it in the right format:
- Open a blank workbook and choose Data > From Text/CSV.
- Set File Origin to “65001: Unicode (UTF-8)” and choose the right delimiter. Check that accented letters look right in the preview.
- Click Transform Data and set the barcode, EAN, SKU and date columns to Text, so Excel leaves them alone.
- Make your changes.
- Use File > Save As with a type that starts with “CSV UTF-8”. If your Windows uses semicolons, the file will still have them, so swap them before importing.
- Don't double-click the saved CSV and save it again: that undoes steps 2 and 3.
Check the file before you import it
Even a careful edit can leave something behind: a semicolon file, a non-breaking space, a header with a trailing space. The ReportWise CSV Fixer reads the file in your browser and lists every problem by row and column before your shop does:
- It detects the encoding and saves the file as UTF-8.
- It swaps semicolons for commas, and decimal commas for points in price columns.
- It repairs garbled text such as “café” back to “café”.
- It flags Amazon product IDs and Shoptet EANs that Excel turned into E+ numbers, so you know which ones to export again.
- It checks headers and values against Shopify, WooCommerce, Shoptet and Amazon rules.
- Checking is free, and files up to 50 rows are fixed for free.