CSV Troubleshooting Guide
By Online CSV Editor · Last updated: 2026-03-24
The short answer is: when a CSV breaks, the cause is usually one of six things — wrong delimiter, broken quotes, malformed rows, encoding mismatch, hidden null bytes, or schema problems. If you check those in order, most CSV import errors become fixable in minutes instead of guesswork.
Use this page as the main diagnostic hub for broken rows, shifted columns, unreadable characters, lost leading zeros, and import failures in spreadsheets, CRMs, ecommerce tools, support systems, and internal apps.
Fast diagnosis: match the symptom to the likely cause
| Symptom | Most likely cause | Best next step |
|---|---|---|
| Everything appears in one column | Wrong delimiter | Change CSV delimiter safely |
| Rows break in the middle of text | Bad quoting or embedded line breaks | Review quoted field rules |
| Characters look like Ã, ’, or � | Encoding mismatch | Fix garbled characters |
| IDs lose leading zeros | Numeric coercion by another tool | Preserve text-like IDs |
| Parser mentions invalid byte sequence | Corrupt or mixed encoding | Fix invalid UTF-8 byte sequence issues |
| Import still fails after edits | Header/schema mismatch or hidden row issues | Run the import checklist |
The troubleshooting order that saves the most time
- Check delimiter first. If the parser uses commas but the file is semicolon-delimited, nothing else will look right.
- Check quote handling next. One unclosed quote can make every later row look broken.
- Check encoding. If text is garbled, column names and values may be unreadable to the importer.
- Check headers and required columns. The file can be structurally valid but still rejected by the destination schema.
- Check text identifiers. ZIP codes, SKUs, and account IDs often fail quietly after spreadsheet roundtrips.
- Test the cleaned output in the destination system. That final validation catches hidden edge cases the preview may miss.
Common CSV problems and the right fixes
Wrong delimiter: commas, semicolons, and tabs are not interchangeable
If the file opens as a single column or columns shift across the table, the reader is likely using the wrong separator. This is especially common when moving files across regional spreadsheet defaults. Start with how to change CSV delimiter safely and the broader delimiter and encoding guide.
Quoted field problems: embedded commas and line breaks
A value such as "Paris, France" is valid CSV data, not two columns. Multi-line cells can also be valid when the whole field is quoted correctly. If quotes are missing or unclosed, parsers will misread later fields. Review quoted fields, the dedicated guide for line breaks inside cells, and the specific repair flow for unclosed quoted field errors. If your parser specifically reports extra fields in a row, go straight to the too many columns fix guide or use the broader uneven rows repair workflow when the file has both extra and missing fields.
Encoding issues: weird symbols, replacement characters, and failed imports
Garbled text like é or � usually means the file was saved with one encoding and opened with another. For the common fixes, start with CSV UTF-8 encoding explained, then review why CSV characters become garbled or how to fix invalid byte sequences.
Import-specific failures: headers, required fields, and hidden junk
A CSV can open fine and still fail in Shopify, HubSpot, or another importer because header names are wrong, required columns are missing, or hidden junk values remain in the file. Use the broken header repair guide when the first row itself is wrong, then run the CSV import checklist before re-uploading.
Quick tips before you touch a broken CSV
- Keep the original file unchanged so you can compare outputs.
- Do not run blind find-and-replace on commas or quotes in raw CSV text.
- Check a few rows at the start, middle, and end of the file after each fix.
- Test with a small import batch before uploading the entire cleaned dataset.
FAQ
Why does my CSV look fine in one tool and broken in another?
Different tools assume different defaults for delimiter, encoding, quote handling, and type coercion. One tool may open the file “helpfully” while another enforces stricter CSV rules.
What is the most common CSV import failure?
Wrong delimiter and header mismatch are the two most common practical causes, followed closely by encoding problems and bad quotes.
What should I check before sending a fixed CSV to a teammate or client?
Confirm delimiter, encoding, required headers, representative row integrity, and whether IDs still keep leading zeros. Then send the cleaned file plus one sentence describing what changed.
Related troubleshooting guides
Canonical: https://csveditoronline.com/docs/troubleshooting