A bank CSV from a European account can look perfectly tidy and still be rejected, because the file may be using semicolons between columns and a comma as the decimal mark. An importer that expects commas between columns and a full stop in amounts will see one giant column or garbage numbers. Tell it, or convert the file, and most rejections stop.
Why is my whole row crammed into one column?
A CSV is just text with a separator. In many European setups the separator is a semicolon, because the comma is already busy as a decimal mark. When a tool built for commas opens such a file, it finds no commas to split on and keeps each row together. You can confirm in a text editor: if a line looks like 15.03.2026;Supermarket;-84,30;1.915,70 then the separator is a semicolon.
Two repairs work. In Excel, use Data, then From Text or CSV, and choose semicolon as the delimiter in the preview window. In Google Sheets, open the file with File, Import, and set the separator type to Custom, entering a semicolon. Then save a new copy as a standard comma CSV.
What happens to amounts like 1.234,56?
This is the trap. In that format the full stop groups thousands and the comma marks cents. A comma-expecting importer might read 1.234,56 as the text 1.234 followed by junk, or as 1.234, a little over one euro. Both are wrong, and the second one is silent.
| As written in file | Meaning | Safe form for import |
|---|---|---|
| 84,30 | eighty-four euros thirty | 84.30 |
| 1.234,56 | one thousand two hundred thirty-four euros fifty-six | 1234.56 |
| -12,00 | a twelve euro payment | -12.00 |
| 2.500,00 | two thousand five hundred euros | 2500.00 |
To convert in a sheet that does not recognise the format, a helper column works: =VALUE(SUBSTITUTE(SUBSTITUTE(C2,".",""),",",".")). The inner SUBSTITUTE deletes thousands full stops, the outer turns the decimal comma into a full stop. Be aware that VALUE reads text according to your computer's regional settings; if your sheet is already set to a comma-decimal region, it may be simpler to leave the numbers alone and change the regional setting instead.
Dates: 15.03.2026, 15/03/2026 or 2026-03-15?
Day-first dates with dots or slashes are common in this region. A date with a day above 12, like 15.03.2026, cannot be read as month-first, so the importer errors. Something like 05.03.2026 is worse because it can be read two ways and the error is silent: spending from 5 March gets booked on 3 May.
- Find a date in your file with a day above 12 to confirm the order.
- Convert the column to YYYY-MM-DD, the order nobody confuses.
- For text dates in A2 written 15.03.2026, use =DATE(RIGHT(A2,4), MID(A2,4,2), LEFT(A2,2)).
- Paste as values and delete the helper.
- After importing, compare one month's total against the statement.
Why do accents turn into strange symbols?
Merchant names with accented letters or a euro sign can arrive as unreadable characters when the file is saved in an older encoding. Re-save from your spreadsheet as CSV UTF-8. If a name is mangled only in a few rows, it is cosmetic; if the whole description column is scrambled, redo the encoding step before importing, since categories depend on readable merchant names.
A short routine before each import
- Read the first lines in a text editor to learn the separator.
- Remove any bank headline lines above the column names.
- Convert amounts to plain numbers with a full stop for decimals.
- Convert dates to YYYY-MM-DD.
- Save as CSV UTF-8 and keep the original untouched.
SpendixAI accepts Excel and CSV exports and handles merchant, type, status and currency columns, with a review step before saving. It is built around India and INR, so it only fits readers who also hold Indian accounts, and the same formatting care applies whichever tool you use.
Useful ideas start with a clearer picture of your own spending.
Explore your reports


