Shopify CSV updates: why an empty cell and a missing column are different

A product CSV update can look harmless until it reaches an existing catalog. The important question is not only what a cell contains. It is whether the column exists, whether the handle resolves to the intended product, and whether a blank value means “leave this alone” or “replace it.”

Shopify’s own product CSV documentation is the source of truth for supported headers and import behavior. Before changing a live store, test a small exported copy and keep a recoverable export of the original catalog.

Start from a fresh Shopify export

Do not begin with a generic spreadsheet template if the goal is an update. Export the exact products you intend to change from Shopify first. That preserves the handles, variant rows, and header conventions that identify the existing catalog.

Then make only the columns and rows needed for the update. Avoid reformatting the entire file “for tidiness”; an accidental edit to a field that is not part of the intended change can become an unintended catalog change at import time.

Check the structure before changing values

There are three common sources of trouble:

1. Missing identifying data. A row without the product handle or another expected identifier can be hard to interpret during review.
2. Unexpected blank cells. A blank in a column that was meant to be preserved deserves a manual decision before upload.
3. Image metadata drift. Image alt text is easy to overlook when several variants share an image workflow.

The free Catalog Preflight checker runs locally in the browser. It does not upload the CSV or connect to a Shopify store. It highlights structural conditions to inspect before importing; it cannot predict every consequence of an import, so a small test and backup remain necessary.

A safer update sequence

1. Export the target products from Shopify.
2. Duplicate the export and make the intended edits in the duplicate.
3. Compare row counts and headers against the export.
4. Run the duplicate through a local preflight check.
5. Import a small, noncritical sample first when the change is unfamiliar.
6. Confirm product and variant pages in the admin and storefront before applying the same pattern to the rest of the catalog.

Treat image alt text as data, not decoration

Alt text is often added late, after the product data work is “finished.” That is why it gets lost during bulk changes. Include it in the CSV review: make sure the intended image rows have clear, product-specific descriptions and that rows that should not carry an image description are not accidentally populated.

The goal is not to make an import look clean. It is to make every intended change understandable before it reaches a live catalog.