If the reader cannot decode the export’s text, it asks you to export the file as UTF-8 and try again. This prevents unreadable bytes from silently changing merchant descriptions or references.
- Keep the original file, then obtain a fresh text export from the bank or a supported UTF-8 working copy.
- Check that names, descriptions and references display correctly before importing.
- Read the file again and compare several rows with the source, especially names containing accents or other non-ASCII characters.
The reader recognizes common byte-order marks and certain declared OFX encodings. A malformed or incorrectly declared file can still fail, so changing its extension will not repair the text.
Distinguish binary files
A spreadsheet workbook or PDF is not a text CSV simply because its filename ends in .csv. Use the appropriate statement workflow or export actual text rows.
After conversion, compare dates, amounts and references as well as visible text. Preserve leading zeros in identifiers where they belong. If you cannot create a faithful text representation, report the original format and exact error rather than replacing unknown characters or reconstructing the source from memory.
Was this article helpful?
Tell us whether this answered the question you came with.