Most CSV imports do not fail because of the data — they fail because of the file. Here are the errors that account for the vast majority of broken imports, and how to fix each one.
1. Shifted columns (values in the wrong fields)
Symptom: prices appear in the description, categories in the price column, or the importer reports "unexpected value" on random rows.
Cause: a value contains a comma that is not wrapped in quotes, so the parser splits it into two cells and every following value shifts left.
Fix: quote any value containing a comma, quote or newline:
wrong: Cotton Tee,soft, breathable cotton,499
right: Cotton Tee,"soft, breathable cotton",4992. Encoding problems (garbled characters)
Symptom: café becomes café, ₹ becomes ₹, or non-Latin scripts turn into question marks.
Cause: the file was saved in an encoding other than UTF-8, or read with the wrong one.
Fix: re-save the file as CSV UTF-8 (in Excel: "CSV UTF-8 (Comma delimited)"). If you must support a legacy system that only accepts a specific encoding, match that encoding exactly.
3. Number formatting broken
Symptom: the importer rejects prices, or reads ₹1,299 as 1 (stopping at the comma) or 1299 as text.
Cause: currency symbols, thousand separators or locale decimal commas inside number fields.
Fix: number columns should contain plain numbers only — 1299.00, not ₹1,299.00. Currency belongs in the store's settings, not in the data.
4. Leading zeros lost
Symptom: SKU 00123 becomes 123 after the file passed through a spreadsheet.
Cause: spreadsheets interpret 00123 as the number 123.
Fix: format the column as text in the spreadsheet before entering data, or avoid opening and re-saving the CSV in a spreadsheet at all.
5. Formula injection
Symptom: a cell starting with =, +,- or @ executes as a formula when the exported file is opened in a spreadsheet — a real security risk (CSV injection), not just a formatting quirk.
Fix: sanitize exported values that start with those characters (prefix with a single quote or space). Good CSV tools do this automatically — CSV Pilot's export guards against formula injection.
6. Wrong or missing headers
Symptom: the importer maps fields incorrectly or reports "missing required column".
Cause: header names don't match what the target system expects, or the header row is missing entirely.
Fix: download the target platform's template and match its header names exactly — including capitalization. Never ship a CSV without a header row.
7. Inconsistent row lengths
Symptom: importer reports a parse error on specific lines, usually near the end of the file.
Cause: some rows have more or fewer values than the header — often from manual edits or a broken merge.
Fix: validate that every row has exactly as many values as the header. A structured preview makes misaligned rows obvious.
8. Duplicate identifiers
Symptom: products overwrite each other on import, or the importer reports duplicate SKUs.
Cause: the same SKU or handle appears on multiple rows.
Fix: make identifiers unique before importing — either deduplicate or add distinguishing suffixes.
Preventing errors before import
- Validate the file before importing, not after it fails.
- Use a tool that previews the parsed result — if the preview looks right, the import will too.
- Test with a small file (5–10 rows) before running the full catalog.
CSV Pilot combines parsing, validation and preview: upload a file, see exactly how it was interpreted, fix flagged issues, then export a clean file.