6 Most Common CSV Import Errors and How to Fix Them
Every data engineer has wasted hours debugging CSV imports. Here are the six errors we see most often, why they happen, and the fastest way to fix each one.
Error 1: Garbled characters (encoding mismatch)
Symptom: garbled sequences like é instead of readable characters. Cause: file saved in Windows-1252, opened as UTF-8. Fix: re-decode with the correct encoding, or use CSV First Aid's auto-detection.
Error 2: First header starts with garbage (BOM)
Symptom: header is '\ufeffid' instead of 'id'. Key lookups fail. Cause: UTF-8 BOM from Excel's 'Save as CSV UTF-8'. Fix: strip the 3-byte BOM prefix.
Error 3: Data in one column (wrong delimiter)
Symptom: entire row appears in a single cell. Cause: file uses semicolons/tabs but parser expects commas. Fix: detect the actual delimiter and re-export.
Error 4: Shifted columns (broken quotes)
Symptom: data appears in wrong columns starting from a specific row. Cause: unmatched or unescaped quote character. Fix: repair quoting per RFC 4180.
Error 5: 'Expected N columns, got M' (inconsistent column count)
Symptom: import tool rejects the file or truncates rows. Cause: some rows have extra or missing fields. Fix: pad short rows and flag long ones.
Error 6: Phantom blank rows or extra NULL records
Symptom: database has empty rows, row count is inflated. Cause: trailing newlines, blank lines between data blocks. Fix: strip empty rows.
Fix all 6 at once
CSV First Aid auto-diagnoses all of these problems and more (14 issue types total). Drop your file, review the diagnosis, apply the fixes you want, and download a clean CSV. Free, no signup, browser-only.