Product Variant Consistency Checker
Variant rows are easy to get subtly wrong: a SKU reused on two rows, an option value left blank, or "Red" written three different ways across one product. Paste or upload your variant data and this checker groups every problem by row so you can fix them before they reach your import.
Paste is processed locally in your browser.
Add data to check
Paste variant rows or upload a CSV/TSV file above. Include a header row with product ID, SKU, option and price columns — the mapping is detected automatically.
How It Works
- 1
Paste your variant rows or upload a CSV/TSV file — the first row is treated as the column headers.
- 2
Confirm the column mapping. Product ID, SKU, options and price are auto-detected from your headers, and you can remap any column at any time.
- 3
The checker runs six checks on every row: missing option values, duplicate variant combinations, duplicate SKUs, missing product identifiers, inconsistent option naming and inconsistent attributes within a product.
- 4
Review the results table, sorted by spreadsheet row number, so you can jump straight to the problem cells in your original file.
- 5
Download the issue report as CSV, or build a corrected file by selecting exactly which fixes to apply.
Example
- A sneaker has three rows with options "red / 9", "Red / 9" and "red. / 9" — after ignoring case, spaces and trailing punctuation they are the same combination.
- The checker flags the second and third rows as duplicate variant combinations of row 2, and flags the reused SKU on the row where it appears again.
- The same rows are flagged for inconsistent naming, each with the suggestion: Use "red".
- A fourth row leaves Option 2 empty — it is flagged both as a missing option value and as an attribute mismatch with the other variants of that product.
- You tick "Standardize option capitalization to the dominant form" and download variants-corrected.csv with only those flagged cells rewritten.
Accuracy & Limitations
- Checks cover the first 5,000 rows per run; larger files are truncated and a notice is shown.
- Only the columns you map are checked — unmapped columns are ignored, so a correct result depends on a correct mapping.
- The corrected file applies only the fixes you select; everything else in your data is left exactly as it was.
- This tool does not validate against a specific platform’s full import specification — a clean result here does not guarantee a platform will accept the file.
- Fully empty rows are skipped and reported separately as skipped rows rather than issues.
- Duplicate combinations and naming checks compare trimmed, lowercase values, so case and spacing differences are treated as formatting noise, not as different options.