An empty cell and an absent column are different instructions.
Shopify’s overwrite behavior depends on the columns supplied and the import settings. Do not assume that deleting a cell is the same as leaving a field out. Consult the current Shopify documentation for the exact fields and operation you intend to update.
For a bounded specification repair, a useful default is to preserve the supplied snapshot and change only the explicitly accepted cells. A small replacement file can unintentionally omit dependencies or leave the merchant unable to inspect what was preserved.
Use a change ledger with a clear boundary.
This fictional review ledger separates an allowed supporting-content change from fields that should remain byte-for-byte equivalent in meaning. The actual export format may normalize CSV quoting or line endings; that is different from changing a field’s value.
| Field | Review decision | Reason |
|---|---|---|
| Product specification metafield | Accept supported correction | Exact-item source backs the change |
| SKU and handle | Preserve | Identity is outside the repair scope |
| Options | Preserve | Variant structure is outside the repair scope |
| Price and inventory | Preserve | Commerce values are outside the repair scope |
Keep a fresh original and compare after import.
Create a fresh export close to the planned import. The corrected catalog is a full supplied snapshot: changes made by staff, apps or another import after that export may not be represented in it. Importing an old snapshot can restore stale prices, inventory or other preserved values. Compare the repair with the current record before importing.
After importing, export the affected products again and compare the intended changes with the actual record. A file-level recheck can show agreement or differences in the supplied snapshots. It does not prove what shoppers see, validate all app behavior, or establish that a product’s physical specifications are correct.
Check the header version.
Shopify now documents names such as ‘URL handle’ and ‘SKU’ and says older column names remain compatible. Keep the header mapping explicit. A tool should tell you which headers it recognizes rather than quietly treating an unknown column as something else.
Inspect an example, then check your own family.
Download the fictional catalogs, exact-item records and expected reviews. The preliminary fit checker helps separate a supported repair from a task that needs another workflow.
The free preflight checks whether your CSV fits Specstead’s supported scope. It does not change your file or store. Source-backed proposals and customer decisions follow in a private workspace.
Sources and method.
Examples on this page are invented to explain review decisions. They are not customer results or measured platform tests. Guidance checked 2026-10-11; Specstead’s implemented scope is described in the file guide.