CSV Import Errors: A Troubleshooting Checklist That Finds the Actual Problem
Quick answer: CSVUndo recommends: when a CSV upload fails, check the file's separator, header names, text encoding, row structure, and destination-specific field requirements before editing values. Excel may also rewrite identifiers, dates, or long numbers. Keep an untouched copy, inspect the file as plain text, and fix one verified cause at a time.
This checklist is for people moving a CSV between Excel, Google Sheets, Shopify, eBay, WooCommerce, QuickBooks, Salesforce or another importer. Different products validate different columns, so a CSV can be technically well-formed yet still fail the destination's rules. Use the symptom map below to isolate the cause without guessing.
Match the symptom to the likely cause
| What you see | First thing to check | Detailed fix |
|---|---|---|
| Everything opens in column A | Comma versus semicolon or tab separator | Import with the right delimiter |
| ZIP code 00501 becomes 501 | Excel inferred a numeric field | Preserve IDs as Text |
| 885909560622 becomes 8.85909E+11 | Scientific-notation conversion | Check which digits survive |
| Names show café or strange symbols | UTF-8 decoded using wrong encoding | Diagnose mojibake |
| A product name splits into extra columns | Unquoted commas or malformed quotes | Fix CSV quoting |
| Some records have unexpected extra rows | Line breaks inside unquoted fields | Check embedded newlines |
| Upload says missing or unknown column | Destination template, spelling, BOM, whitespace | Compare with platform-specific schema |
| Dates reverse month and day | Locale or automatic date inference | Protect date formats |
Eight checks before you import a CSV
- Keep the original. Save a copy from the export system before opening it in a spreadsheet app. Do not repeatedly save the only copy in Excel.
- Check the template. Download the current sample/template from the destination platform; confirm required headers, case-sensitive names where applicable, and expected field order.
- Check separators. A comma, semicolon or tab may separate fields. A CSV viewer showing one giant column is often a delimiter mismatch, not corruption.
- Check quoting. Fields containing a separator, a quote or a line break need correct CSV escaping; a literal quote inside a quoted field is normally written as a pair of quotes.
- Check encoding. If non-English names break, inspect UTF-8 versus legacy encodings. A UTF-8 BOM can help some Excel workflows but may create an unexpected prefix for a simplistic importer.
- Preserve identifiers as text. ZIP codes, barcodes, SKUs and account identifiers are not quantities. Prevent spreadsheet auto-conversion before it happens.
- Inspect row consistency. Each logical record should contain the expected number of fields, accounting for quoted commas and embedded newlines; do not count commas naively.
- Test a small batch. Try a few representative, non-sensitive rows and inspect the destination result before running a full import.
For a local, privacy-first inspection, open the CSV Doctor and CSV Validator. Tool pages process the file in the browser rather than uploading it to CSVUndo.
Worked example: why a comma inside a product name breaks an import
This is illustrative synthetic data, not a real customer export. A product name such as "Mug, large" must be quoted so it remains one field.
sku,title,price A-101,"Mug, large",12.50 A-102,"Plate ""classic""",9.00
The second record has three fields even though the title contains a comma. The last title shows how a quotation mark is escaped inside a quoted field. By contrast, A-101,Mug, large,12.50 has four fields and may trigger an unexpected-column error.
To reproduce: paste the example into a plain-text file with a .csv extension, and check how the destination importer parses its columns. For more on quoting and separators, see the detailed explanation.
What to check for your destination
File structure is only half the job. Each application can require its own template, allowed values, variants, formats, matching identifiers or permissions. Never use a generic CSV fixer as a substitute for the destination's current documentation.
Sell products online
Shopify import errors, eBay CSV errors, Amazon flat-file uploads, WooCommerce imports and Etsy CSV errors.
Business and contacts
QuickBooks uploads, Salesforce Data Loader, Mailchimp lists and Outlook or Google Contacts.
When a repair tool cannot fix the original values
Lost information is different from changed formatting. If Excel already rounded a 17-digit identifier or saved a missing leading zero without any record of its original width, the exact original value may no longer be present in the saved CSV. Re-export from the source system or use an authoritative source-of-truth rather than guessing. The long-number precision guide explains why.
Questions people ask before importing
Is a CSV import error always a broken CSV?
No. A perfectly valid CSV can still be rejected for missing required columns, invalid values, incorrect account permissions, unsupported features or the wrong import template.
Should I open a CSV by double-clicking it in Excel?
For data with identifiers or date-like strings, prefer the import workflow and explicitly set sensitive columns to Text. Double-clicking may cause Excel to infer types before you can inspect the original text.
Can I upload a private customer export to a random online converter?
Avoid uploading sensitive files unless you understand that service's data-handling practices and have permission. CSVUndo's actual repair tools run locally in the browser; this guide contains only public text.