What garbled text means
Characters like é, ’ and € take more than one byte in UTF-8. When software reads UTF-8 as if it were Windows-1252, each byte becomes its own character, so “é” turns into “é” and a curly apostrophe into “’”. This is called mojibake. The original bytes are still there, so it can be reversed exactly, which is what Repair garbled characters does.
Safe by design
The repair only changes a value when the result is valid UTF-8, so genuine accented text like “Zoë” or “à la carte” is left alone. Text that was garbled twice (“é”) is repaired too.
Opening the result in Excel
Excel on Windows assumes a CSV file uses the computer's legacy encoding unless it starts with a byte order mark (BOM). Keep Excel-friendly UTF-8 ticked when you download and accents display correctly when you double-click the file. If the file goes to a developer or a database instead, you can untick it.
Read the guide to garbled characters in CSV files for more background and fixes in other tools.