A CSV with merchant names that include French accents can look scrambled in Excel even though the file is perfectly fine. The problem is how Excel reads the text, not the data. Re-import the file with UTF-8 selected and confirm the delimiter, rather than opening it with a double-click.
Why do accented merchants break first?
Plain English letters look identical in nearly every encoding, so a file full of ordinary names opens fine. Accented letters are where encodings differ. A name such as Marche Jean-Talon written with its proper accent can turn into two odd characters when Excel assumes the wrong encoding. If you track spending across both English and French merchants, you will notice it only on those rows.
- Plain names look correct, which hides the problem
- Accented names show extra symbols
- Sorting and searching then fail on those rows
- Matching a merchant against a rule or list can silently miss them
Step-by-step repair
- Do not save the garbled view. Close it.
- Open a new workbook and use Data, then From Text/CSV.
- Pick UTF-8 as the file origin and look at the accented rows in the preview.
- Confirm the delimiter so each column is separate.
- Load the data into a new sheet and save it as a workbook.
What about regional settings that use a decimal comma?
Some Canadian computers are set to French regional options, where the decimal mark is a comma. A file written with points can then load as text, and a total shows zero. Test with a quick formula next to the first amount: =ISNUMBER(C2). If it returns FALSE, convert the column. Select the column, use Find and Replace to swap the point for the mark your system uses, or import with the correct locale.
| Merchant row | Amount | Check |
|---|---|---|
| Grocery order | C$64.80 | Is the amount a number? |
| Pharmacy | C$23.15 | Does the name read correctly? |
| Pub night | C$41.00 | Is the date in the right month? |
These three rows total 64.80 + 23.15 + 41.00 = 128.95. If your sheet says 0, something is still stored as text.
When is a different tool the right move?
If you do this every month, move cleaning into a routine rather than repeating manual repair. A tracker with consistent CSV output helps. SpendixAI is built around India and INR, so it is not aimed at Canadian dollar records, but its CSV download is one transaction per row with plain headers, which is the standard to look for.
Does it matter which program made the file?
It matters a little. A file written by a web service is usually UTF-8, while one written by an older desktop program may use an older encoding. You cannot always tell by looking, so let the import preview decide. Try UTF-8 first, since it is the most common, and if accented letters still look wrong, try the next option on the list until a known merchant name reads properly. The preview costs nothing, because you can close the window without loading anything.
Is Google Sheets any better?
It can be a useful second opinion. Importing the same file into Google Sheets uses its own detection, and if accents and columns look right there, you know the file is sound and only Excel's guess was wrong. Use File, then Import, and pick the separator if it asks. You can then download the sheet again as a CSV or as a workbook. Treat it as a diagnostic step rather than a permanent home for your records.
Short checklist
- Use Data, From Text/CSV
- Choose UTF-8
- Confirm delimiter and decimal mark
- Check one accented row and one amount
- Save the fixed version as a new workbook
Useful ideas start with a clearer picture of your own spending.
Explore your reports


