Shopify CSV redirect: should I use the full URL or just the path?

The two columns follow different rules

Shopify's redirect CSV has exactly two columns: Redirect from and Redirect to. "Redirect from" identifies an incoming request on your own store, so it must be written as a path starting with a forward slash — /products/old-mug, not https://yourstore.com/products/old-mug. Shopify is matching against the part of the URL after your domain; including the domain again either gets rejected during import or never matches a real incoming request, since no browser actually requests a URL with your domain baked in twice.

"Redirect to" is allowed to be a full URL

The destination column is more permissive on purpose, because redirects sometimes need to point off-site — to a new domain after a rebrand, a partner site, or a page that no longer lives on Shopify at all. A relative path (/collections/mugs) sends the visitor elsewhere on the same store; a full URL (https://example.com/mugs) sends them to a different domain entirely. Both are valid for this column specifically.

What goes wrong when the two get swapped

Putting a full URL in "Redirect from" is the single most common mistake in bulk redirect imports, usually because the list was built from an exported sitemap or crawl report that naturally includes the full domain. Depending on the exact format, Shopify either flags the row during import or accepts it and the rule simply never fires, because nothing a browser sends will ever match a "from" value with a domain embedded in it.

How to check your file

Upload your CSV above — the checker inspects both columns separately and flags any "Redirect from" value that includes a scheme or domain, so you can strip it down to a path before uploading to Shopify.