Prepare a UTF-8 product CSV for WooCommerce

Use UTF-8 and the WooCommerce product schema

WooCommerce's official product CSV documentation recommends UTF-8 and its built-in schema. The importer uses commas as separators, expects values containing commas to be escaped, and maps known headers such as SKU, price, attributes, and images.

• Japanese or accented text is broken: convert the confirmed source encoding to UTF-8.
• A category or attribute contains commas: preserve the field boundary with the documented quoting or escaping rule.
• Products update unexpectedly: distinguish an existing id: value from a SKU before import.

• Convert the CSV encoding or delimiter →Choose the source and output format explicitly before the destination import.
• Diagnose CSV structure before changing values →Review column counts, quotes, empty rows, and duplicate rows in a separate workspace.
• Select the columns required by the import template →Remove export-only fields while keeping the original row count.

Start from the official sample or an exported product CSV

WooCommerce recommends using an exported product file or the sample CSV as the starting schema. Do not rename headers because a spreadsheet label looks clearer; map custom metadata explicitly with the meta: prefix described in the documentation.

• Confirm whether the file creates new products or updates existing products by ID or SKU.
• Review boolean values, product status, taxonomy hierarchy, and image URLs before converting anything.
• Keep the original export and a pre-import copy for rollback and comparison.

• Find CSV rows with inconsistent column counts →Diagnose malformed rows before uploading them to a platform.
• Select the columns required by the import template →Remove export-only fields while keeping the original row count.
• Check or convert the CSV character encoding →Use this when Japanese, Korean, or accented text is garbled after export.

Convert and review the product export in LocaCSV

If the diagnosis finds inconsistent columns or unclosed quotes, stop and fix the source structure deliberately. Encoding conversion cannot repair a broken product schema.

1. 1Open a copy of the WooCommerce export in LocaCSV's format conversion workspace.
2. 2Select the source encoding that displays product names correctly and UTF-8 as the output encoding.
3. 3Keep comma as the output delimiter unless the importer and file preview prove another delimiter is required.
4. 4Review the output header, SKU or id: values, quoted descriptions, and row count before starting the WooCommerce import.

Verify products before committing the import

• The output is UTF-8 and every row has the same columns as the WooCommerce schema.
• Comma-containing categories, descriptions, and attributes stay in one field.
• The local product count and the number of records in WooCommerce's preview match.
• Import a small set first and confirm product status, prices, variations, and images before the full file.

• Diagnose CSV structure before changing values →Review column counts, quotes, empty rows, and duplicate rows in a separate workspace.
• Learn the general CSV split workflow →Choose a row count or target size while keeping the header in every output file.
• Remove duplicate records before importing →Check duplicate IDs or email addresses before you create a second record.

Treat product mapping errors as a separate branch

A valid UTF-8 CSV can still fail because a header is not in the built-in schema, an image URL is inaccessible, a variation is missing a required value, or an ID/SKU points to the wrong product. Use WooCommerce's mapping preview and logs for those cases.

• Select the columns required by the import template →Remove export-only fields while keeping the original row count.
• Diagnose CSV structure before changing values →Review column counts, quotes, empty rows, and duplicate rows in a separate workspace.
• Fix duplicate emails before a Mailchimp import →Remove repeated addresses locally, then review Mailchimp's import history.

Hand WooCommerce a traceable product file

Keep the source, converted output, preview screenshot, and imported product count together so an update can be reversed or audited.