A metadata schema is the blueprint for an asset record: which fields exist, what type of value each holds, which are required, and how they relate. It is closely tied to a taxonomy, the schema defines the fields, and the taxonomy and controlled vocabularies govern the values that go in them.
Why it matters
Without a schema, every asset is described differently and search collapses. A clear schema is what makes metadata consistent, governable, and portable between systems, especially when it maps to a standard like IPTC or Dublin Core.
How it shows up in practice
A team designs a schema with descriptive fields like title and keywords, administrative fields like rights and creation date, and structural fields like version, then sets the required ones so a baseline is always captured. The schema is mapped into the DAM as custom fields tied to IPTC where possible.
Common mistakes
- Adding every conceivable field, so most go unfilled.
- Designing the schema without mapping it to a recognized standard, hurting portability.
- Never marking fields required, so the schema is optional in practice.
Stacks covers field design in using custom metadata and keywords.