CSV Special Characters Broken in Excel: How to Fix Encoding Problems
Why are CSV special characters broken in Excel, and how do you fix them?
CSV special characters broken in Excel means accented letters such as é, ñ and ü appear as odd pairs of symbols. The file is usually saved as UTF-8, but Excel reads it with an older encoding. The fix is to import the file and choose UTF-8 instead of double-clicking it.
Do not panic. Your data is almost always intact, because only the on-screen interpretation is wrong. This order works in most cases.
- Start Excel with a blank workbook instead of double-clicking the file.
- On the Data tab, choose the option to import from a text or CSV file, then pick your file.
- Set the file origin (encoding) to UTF-8 in the preview window.
- Check the delimiter. Try comma or semicolon until the columns split correctly.
- Load the data and check the accented letters in a few rows.
- If you save again, choose a CSV UTF-8 format.
What causes CSV special characters to break in Excel?
A computer stores numbers, not letters. Each letter becomes one or more bytes, and the encoding tells a program which letter those bytes mean. If the program that writes the file and the program that reads it use different encodings, the letters come out wrong.
For example, the letter "é" takes two bytes in UTF-8. If Excel reads those two bytes with an older single-byte encoding, you see two separate strange characters. So the data does not change; only the interpretation does.
That is why the same file can look perfect in a plain text editor and broken in Excel. It can also load correctly in a web browser or an online store dashboard. The problem sits in the reading application, not in the file itself.
This distinction saves time. Instead of rebuilding the file, you often only change how you open it. Open the file in a text editor first. If the letters look right there, the source is healthy and the fix belongs in Excel.
What do broken accented characters look like?
The shape of the damage tells you which way the encoding mismatch runs. The table below lists common pairs. However, the exact look depends on the font and software version, so treat it as a clue rather than a strict rule.
| What you see | Likely meaning | Where to look first |
|---|---|---|
| "é" instead of "é" | A UTF-8 file was read with an older encoding | Set the file origin to UTF-8 |
| "ñ" instead of "ñ" | The same mismatch on another letter | Use the import screen |
| "ü" instead of "ü" | The same mismatch again | Use the import screen |
| Question marks or black boxes | The letter may have been lost when the file was written | Export the file again from the source |
| Letters fine, all columns in one cell | A delimiter problem, not an encoding problem | Change the delimiter |
Be careful with question marks. In that case the letter was usually lost during export, and no Excel setting will bring it back. You need a fresh UTF-8 export from the source system.
What is the difference between UTF-8 and Windows-1252?
UTF-8 is a modern encoding that can write almost every language in one system. Windows-1252 is an older code page used by many English and Western European Windows systems, and it covers only a limited set of letters. Your computer region decides which one Excel assumes.
As a result, the same CSV can look different on two computers. A file from an online store is usually UTF-8, while an old accounting program may still write the older code page. This depends on each system, so it is not identical everywhere.
- UTF-8: multilingual, the usual choice of current systems and online stores.
- Windows-1252: older code page for Western European languages, found in legacy exports.
- Other code pages: some regions use different ones, so check your source system.
In practice, the safest route is to create files as UTF-8 and to choose UTF-8 when you read them. If a legacy program produced the file, you can also try the older option in the preview.
What is a BOM, and why does it matter for Excel?
A BOM (Byte Order Mark) is a small marker at the very start of a file. It tells a program that the file is UTF-8. According to Microsoft, you can open a UTF-8 CSV file normally if it was saved with a BOM. Source: Microsoft Support guide to opening CSV UTF-8 files in Excel.
A UTF-8 file without a BOM often triggers the older default when you double-click it. Then the accented letters break. Also, the same file may open fine in another program, because that program does not look for a BOM.
If another system creates your product file, look for a BOM option in its export settings. However, the opposite problem exists too. Some import tools treat the BOM as part of the first column name. So learn what the target platform expects before you decide.
How do you fix CSV special characters broken in Excel with Get Data?
The most reliable method is to use the import feature instead of double-clicking. Microsoft describes the path as Data, Get Data, From File, From Text/CSV. Menu names can vary by version and language, so do not rely on exact button text.
- Open a blank workbook in Excel.
- On the Data tab, find the command to import a text or CSV file.
- Select your CSV and wait for the preview.
- Choose UTF-8 as the file origin and confirm the letters look right.
- Check the delimiter, then load the data or transform it first.
The preview is your best friend. If the letters look right there, they will look right in the sheet. If they look wrong, then pick another encoding and refresh the preview until the text is clean.
How do you open a CSV with the older text import wizard?
Microsoft also notes that the "Get Data From Text" option opens the Text Import Wizard and does not create external links to the workbook. You can reach it from the ribbon or the Tell me box. Source: Microsoft Support guide to opening CSV UTF-8 files in Excel.
The wizard runs in three steps. First, you pick the file type and the file origin, where UTF-8 or an older code page appears. Second, you set the delimiter. Third, you set the data format of each column.
The column format matters a lot. For example, barcodes and phone numbers with leading zeros lose those zeros unless you mark the column as text. So set those columns to text inside the wizard.
How do you stop CSV special characters broken in Excel when saving as UTF-8?
If you must send the corrected table back, the save format matters too. In the Save As dialog, the file type list contains an option similar to "CSV UTF-8 (Comma delimited)". The exact name may differ slightly by version. The plain "CSV" option often uses the older system encoding.
- Go to File, then Save As, and open the file type list.
- Choose the CSV option that mentions UTF-8.
- Read the warning that only the active sheet will be saved.
- After saving, open the file in a text editor and check the letters.
Microsoft also reminds you that saving as CSV keeps only the current worksheet and that some features can be lost in text formats. Source: Microsoft Support guide to importing and exporting text and CSV files.
What if all columns appear in one cell?
If the letters look fine but each row sits in one cell, you are dealing with a delimiter problem. Some files separate columns with commas, others with semicolons. Excel then checks the list separator in your regional settings.
In regions where the comma is the decimal mark, files often use semicolons. On the other hand, a system with English settings expects commas. So the same file can look different on two computers.
Pick the delimiter by hand on the import screen, and the preview updates at once. Microsoft says you can choose a delimiter from a list when importing, or change it in step two of the wizard. Source: Microsoft Support guide to importing and exporting text and CSV files.
How do regional settings and the list separator change the result?
When Excel saves a CSV, it uses the default list separator from your Windows region settings. Microsoft warns that changing the Windows setting affects every application on the computer. If you only want to change Excel, adjust the decimal and thousands separators instead.
The practical result is simple: the person who receives your file may use different regional settings. So if you exchange files with a supplier abroad, agree on the delimiter first.
| Situation | Likely symptom | Fix |
|---|---|---|
| File uses semicolons, system expects commas | One long column | Choose semicolon during import |
| File uses commas, system expects semicolons | One long column | Choose comma during import |
| Prices mix 12.50 and 12,50 | Numbers look like text | Set the decimal mark during import |
| Address field contains a comma | Rows shift | Make sure the field is in quotes |
Why does opening and saving a broken CSV damage it for good?
If you double-click a file, see broken letters and press Ctrl+S, you write the damage into the file. Excel treats the broken characters as real data and saves them all with the old encoding. Restoring the original letters then becomes very hard.
So close a file that looks broken before you save, and get the original again. Read any warning about lost characters. Also keep a copy with a version name, which keeps your way back open.
- Never overwrite the original. Always work on a copy.
- Open the file with the import method before editing.
- Choose a UTF-8 format when you save.
- Check the letters in a text editor before you send the file on.
Do special characters look right when you open the CSV in Google Sheets?
Often yes, because Google Sheets usually reads UTF-8 files well. According to Google, you go to File, then Import, and for plain text files you pick a separator such as detect automatically, tab, comma or a custom character. Source: Google Sheets help on importing CSV files.
The page does not mention an encoding choice. So if the file arrives in an older encoding, it can look broken there too. However, UTF-8 files that break in Excel often open cleanly in Sheets, so always check the preview.
A practical path: open the file in Sheets, check the letters, and download it as CSV from there. That gives you a UTF-8 file. If you automate with Sheets, see our guide to connecting ChatGPT and Google Sheets with Make.
How do you prevent encoding problems in online store product imports?
Product names, descriptions and category paths carry many accented letters, so store files hit encoding problems often. If you export a product list, edit it in Excel and upload it again, every link in that chain must use the same encoding.
We suggest a simple routine: export, import into Excel with UTF-8, edit, save as UTF-8 CSV, upload. Then open a few products in the store and check the letters. Also, a small test file of ten rows catches errors before you upload everything.
For URL slugs with accents, our slug generator and the guide to ecommerce SEO for product and category pages help. If you want help with the whole process, see our ecommerce consulting page.
What should you check for ad catalogs, newsletter lists and catalog files?
Every platform expects a slightly different file, but the encoding rule stays the same: UTF-8 is the safest choice. Newsletter lists carry first and last names with accents, and a broken import can greet a person with nonsense characters.
For ad catalogs, read our guide to Meta Commerce Manager and catalog management. There we explain the feed logic; here we only cover file encoding.
- Newsletter list: verify name and city columns with sample rows before you upload.
- Product catalog: check accented letters in titles and descriptions.
- Price column: make sure the decimal mark matches the platform.
- Barcode and phone: keep values with leading zeros as text.
How do you choose the encoding when you upload to a platform?
Many dashboards offer no encoding option on the upload screen. They simply expect UTF-8. Others give you a list. If an option exists, tick UTF-8. If not, producing the file as UTF-8 is your job.
Menu and setting names differ by platform, so check the help page of the import section. If the help page does not state an encoding requirement, run a small test upload. However, never test with live customer data; use made-up sample rows.
If the platform gives an error report, read the row numbers. The problem usually clusters in a few rows. For instance, a single curly quote or hidden character in one product name can make the whole file fail.
Which fix method should you choose?
The right method depends on your situation. The table below compares the options side by side, so you can see quickly which path fits.
| Method | When it fits | Watch out for |
|---|---|---|
| Import with UTF-8 selected | Almost always, first choice | Menu names vary by version |
| Text Import Wizard | When you want to set column formats one by one | Considered a legacy tool, but step by step |
| Save as CSV UTF-8 | When you upload the file back to a platform | Only the active sheet is saved |
| Open in Google Sheets | When Excel looks broken but Sheets looks fine | No encoding choice offered |
| Export again from the source | When you see question marks or lost letters | Depends on the source system settings |
For most teams, the first two methods are enough. However, if you handle files regularly, turn the steps into a written checklist. That protects your team time.
Is XLSX safer than CSV for keeping accented characters?
Yes, there is a big difference. An XLSX file carries its encoding inside, so accented letters usually look the same on every computer. CSV is plain text and brings no encoding information with it. So the reading program has to guess.
A safe approach is to edit products in XLSX and create the CSV only at upload time. That way your letters stay safe during daily work, and you make the encoding decision once, at the end. Remember, though, that formulas and formatting disappear when you convert to CSV.
- Working copy: keep it as XLSX so letters and formats stay intact.
- Upload copy: export as UTF-8 CSV and keep it in a separate folder.
- Archive copy: keep the first file from the source system, untouched.
This three-copy routine also makes it easier to find where something went wrong. Besides, a colleague can take over the process without confusion.
Can you repair broken text with find and replace?
Sometimes only a few cells are broken, and a fresh export is not possible. In that case, Find and Replace can swap the common broken pairs for the correct letters. For example, replacing every "é" with "é" is a reasonable first step.
However, this is a patch. Broken pairs are numerous, and one missed pair leaves the text half fixed. Also, some broken characters hide an invisible byte, so your search may not match. It works on small files, but it is risky on large catalogs.
The lasting fix is always to go back to the source and reopen the file with the right encoding. Use the patch only as a last resort, then check ten random records by eye. Keep a backup copy either way.
What sample text can you use to test the encoding quickly?
Before you touch real customer data, build a tiny test file. Write one short line that holds the special letters of your languages. A made-up phrase such as "Café Niño Über Straße" is enough.
Save it as CSV, then load it into Excel or the target platform. If every letter returns correctly, the chain is sound. If a letter breaks, you also learn where the problem starts. Because the file is tiny, you find the cause in minutes.
- Accented vowels such as é, è, ê, á and ö.
- Letters such as ñ, ç and ü.
- Typographic marks such as curly quotes, apostrophes and long dashes.
- A field with a comma and a semicolon to see the delimiter behavior.
Typographic marks cause trouble often. Curly quotes pasted from a word processor can break under some older encodings. So include them in your test line.
How should your team agree on one CSV file rule?
Most problems come from different people opening and saving the same file in different ways. One double-clicks, another imports, a third uses Sheets. In the end, the file carries a different encoding in every hand. One written rule is the most effective fix.
Keep it short: "Files are saved as UTF-8 CSV and opened in Excel with the import feature." Put this sentence as a note at the top of the shared folder, so a new colleague starts from the same page.
- Add a version and date to the file name so everyone knows the current copy.
- Write down the delimiter decision: comma or semicolon.
- Let everyone test a small sample file the same way before uploading.
- Make sure everybody knows how to return to the original copy.
A rule feels slow at first. However, repairing a broken catalog can take hours. Besides, a mistake in a file that reaches customers, like a newsletter, affects your brand directly.
How do you report a broken CSV to the supplier who made it?
If another company creates the file, a short and clear message saves time. First, check whether the file also looks broken in a plain text editor. If it looks fine there, the issue is your Excel setting. If it looks broken there, the issue is at the source.
In your message, include a few sample rows and the result you expect instead of a screenshot. Also say which program you used. Then the other side knows right away whether to change the encoding or the delimiter.
- Open the file in a text editor and note the result.
- Copy three or four sample rows that hold no personal data.
- State the encoding (UTF-8) and the delimiter you expect.
- Ask for a new export and repeat the same test.
Never send rows with customer or subscriber details. Use made-up examples instead. If privacy rules like the GDPR apply to your files, this article is not legal advice.
What if the special characters stay broken after all these steps?
If you tried everything, it is time to question the source. The file may have been created with a mixed encoding by another program. Sometimes the text was converted wrongly more than once, so the damage comes in layers.
- Open the same file in a text editor; if it is broken there, the problem starts at the source.
- Look for a UTF-8 option in the export settings of the source system.
- Open the file on another computer and compare the result.
- Contact the platform support with sample rows, but never share real customer data.
If you cannot recreate the file, you may need to fix the broken text by hand. Work in small batches. If your setup is complex, our team can review your process, but we cannot promise to recover a file whose data was lost.
Note: this article is general information. Menu names and behavior change with Excel and platform versions, so check the current steps in the official help pages. Our guides on QR codes that do not scan and a favicon missing in Google search cover similar small technical issues. You can also review workflow automation for repeated file tasks.
