What to check before reimporting product data
A careful review process for spreadsheet updates, missing rows and product images.
Last updated
An updated spreadsheet can contain several kinds of change at once: a new product, a corrected description, a higher base price or an absent row. Treating all of them as a blind replacement makes it difficult to understand what will happen. A review step should separate the changes before anything is saved.
Check the source file first
- Use CSV or a plain XLSX workbook without macros or formulas.
- Keep SKUs as text and retain their leading zeros.
- Use a single row per SKU, with a product name and an explicit base price.
- Keep the currency consistent with the customer versions that use the data.
- Remove accidental blank columns and check the intended product count.
If a workbook stores a SKU as a number, displaying leading zeros with a number format is not equivalent to storing the identifier as text. Convert the cell to a real text value in the source file. Do not rely on the import tool to reconstruct zeros that may already have been lost.
Separate the five review categories
- New products: SKUs that are not in the current product library.
- Missing products: existing SKUs absent from the new file; keep them unless explicitly archived.
- Price changes: different base wholesale prices and their customer-specific impact.
- Field changes: updated names, descriptions, categories or ordering details.
- Image conflicts: a different image reference or a newly supplied file that needs review.
Match images explicitly
Start with exact filenames or exact SKU stems. When two files could match the same product, ask for a choice. A similar-looking filename is not enough evidence to replace an existing product photo. Keep previously assigned images when an update contains no replacement image.
Make partial imports deliberate
A file may contain valid rows and rows with missing prices, duplicate SKUs or malformed values. Show the row numbers and reasons. When partial imports are supported, require an explicit decision to skip invalid rows. Skipped rows should not erase existing products or become zero-priced replacements.
Confirm against the version you actually reviewed
Another browser tab may change the product list while an import preview is open. Compare the current revision with the revision used for the preview. When they differ, ask for a fresh review instead of overwriting newer work. Repeating the same confirmation after a network retry should not create a second revision.
After confirmation, inspect the updated product list and create new customer catalog snapshots. Older exports remain records of the earlier data. That separation makes ordinary maintenance predictable: review the source, approve the differences and send a clearly identified new version.
Put the guide into practice.
Explore the fictional collection, check buyer prices and see how the layouts handle the same product data.