Prepare your product catalog before an ecommerce import

A catalog spreadsheet may look complete while leaving important questions unresolved. Is a row a product or a variant? Does a price refer to one item or a case? Will an empty cell preserve an existing value or clear it?

Resolve those questions before running an import. The file's ability to upload is only the first check.

Establish product identity

Agree a stable reference for the product and for each sellable variant. Document how those references relate to the stock system, supplier records, and any existing storefront.

For an illustrative clothing catalog, a style could have several size and color combinations. Ask whether each combination has its own stock reference and image. Do not assume that a shared display name uniquely identifies the item someone can buy.

List which references external systems already depend on. If an import creates new identifiers, the team needs to know which connections must be updated.

Make the fields understandable

Create a short catalog dictionary alongside the import file.

Assign an owner for missing or contradictory values. An importer should not quietly guess a pack size or resolve conflicting prices.

Read the receiving platform's overwrite rules

Shopify's CSV import guidance explains that, when overwriting matching products, a blank included non-required column can clear an existing value, while an omitted column preserves it. It also warns that changing option values can recreate variant IDs and disrupt dependencies on those IDs.

Those are Shopify-specific behaviors. Check the documentation for your target platform rather than transferring the rule to another importer.

Ask the implementation team to explain the match key, required columns, blank-value handling, variant behavior, and whether the operation can be reversed. Keep a copy of the input and an appropriate export of the starting state.

Test the awkward products first

Build a representative test set that includes a simple item, several variants, a discontinued product, a product with multiple images, and any pack or unit distinction that matters.

Use a suitable test environment. Inspect the result as both an administrator and a buyer. Check search, product selection, displayed price, image assignment, and the order record created for the chosen variant.

If another system receives the order, verify its product reference too. A correct-looking page can still send the wrong identifier.

Review the full import as a reconciliation

After the approved import, compare counts and investigate exceptions. A count match alone does not prove that values or variant relationships are correct. Sample important fields and follow any warnings through to a decision.

Record changes to product URLs and discuss the required redirect plan if an existing store is moving. Separate catalog cleanup from broader data migration work, but keep their owners coordinated.

Finally, agree who maintains the catalog after launch. A clean first import helps most when later edits follow the same definitions and approval rules.