Tags are separate from prefixes and act as key-value labels you can attach to objects in a bucket. They help you apply policies to only the data you need. For example, you might tag objects with department=finance and set a rule that automatically deletes older finance data after a certain period without affecting other departments’ files.