Canonical vs 301 for duplicate URL paths — different tools

Canonical vs 301 for duplicate URL paths — different tools
Site had /shoes/nike and /nike/shoes both live, both indexed, splitting rank.
— 301 redirect: removes the duplicate entirely, forces one URL, passes ~all equity. Use when nobody needs the second path alive.
— rel=canonical: keeps both URLs reachable (think faceted/filtered versions you want users to land on), consolidates ranking signals to one.
The call: genuine duplicate nobody should visit? 301, kill it. A filtered or tracking variant that needs to stay clickable? Canonical it.
Don't canonical across truly different content — Google treats canonical as a hint and ignores it when the pages don't match. Then you've got two indexed dupes and a false sense of safety.