Why encoding matters
Computers represent text using character encodings. The most common encoding today is UTF-8, which supports characters from every language in the world. But older systems and software sometimes use different encodings like Latin-1 (Windows-1252) or ASCII, which only support limited character sets.
When a CSV is saved in one encoding and imported into a system expecting a different encoding, special characters break.
Characters become garbled
José becomes José, Müller becomes Müller, 你好 becomes mojibake.
Emojis and symbols fail
Emoji characters (😀), currency (€, £, ¥), and special symbols (™, ©) can get stripped or replaced.
Search and matching breaks
Searching for 'José' won't find 'José'. Deduplication treats encoded and decoded versions as different people.
International data is lost
Names in non-Latin scripts (Cyrillic, Arabic, Chinese, Korean) can be completely mangled or deleted.
Real example:
Character encoding problems:
Name (UTF-8) | Name (garbled) ----------------------|--------------------- José García | José GarcÃa François Müller | François Müller 李明 (Li Ming) | [mojibake or blank] Søren Kierkegaard | Søren Kierkegaard ✓ System saved in UTF-8 ❌ Imported as Latin-1 ❌ Characters mangled
The solution: Always save CSV files in UTF-8 encoding and make sure your import system expects UTF-8.
Common special character problems
Accented letters (é, ñ, ü, ø)
Cause:
File saved in Latin-1 or ASCII, but imported as UTF-8 (or vice versa).
Example:
Café becomes Café, Señor becomes Señor
Currency symbols (€, £, ¥, ₹)
Cause:
System doesn't recognize Unicode currency characters.
Example:
€100 becomes €100, £50 becomes £50
Emoji (😀, 🚀, ✅)
Cause:
Older systems or Excel can't handle emoji characters.
Example:
Status: ✅ becomes Status: [garbage] or blank
Non-Latin scripts (中文, العربية, Кириллица)
Cause:
System expects ASCII or Latin-1, can't read other scripts.
Example:
Chinese, Arabic, Russian names become blank or corrupted
Trademark/Copyright symbols (™, ©, ®)
Cause:
Limited encoding support for special symbols.
Example:
Acme™ becomes Acme or Acme™
Dashes and quotes (" ", – , — , ' , ')
Cause:
Excel and Word use special typographic characters, standard CSV expects straight characters.
Example:
"Smart quotes" become straight quotes with different characters
Understanding character encodings
Here's what you need to know about the most common encodings:
UTF-8
Supports:
All characters in all languages
Use:
✅ Always use this. It's the modern standard.
ASCII
Supports:
Only English letters, numbers, and basic symbols
Use:
❌ Avoid. Loses accents and special characters.
Latin-1 (ISO-8859-1)
Supports:
Western European characters (accents, but not € symbol)
Use:
⚠️ Only if your system requires it. UTF-8 is better.
Windows-1252
Supports:
Similar to Latin-1, with some Microsoft additions
Use:
❌ Legacy encoding. Use UTF-8.
How to fix encoding problems
Step 1: Save your file in UTF-8
In Excel:
- Click File → Save As
- Choose "CSV UTF-8 (.csv)" or equivalent
- Do NOT save as "CSV (.csv)" — that uses the wrong encoding
In Google Sheets:
- Click File → Download → Comma Separated Values (.csv)
- Google Sheets always exports in UTF-8
Step 2: Don't rely on smart quotes or special dashes
Word and Google Docs convert straight quotes and dashes to fancy typographic characters:
If pasting from Word, use "Paste Special → Unformatted Text" to keep characters simple.
Step 3: Avoid copying from web pages
Copying text from web pages, PDFs, or emails can include hidden formatting or unusual characters.
- Copy from the raw source (plain text, database)
- Paste into a plain-text editor first to strip formatting
- Use Find & Replace to clean up unusual characters
Step 4: Test with sample international data
Before importing thousands of records, test with a small file containing accents, symbols, and special characters:
Special character best practices
Handle international data with confidence
Import Builder fully supports UTF-8 encoded data, including accents, symbols, and characters from all languages.
Start building free