XMP (Extensible Metadata Platform) is a metadata standard, created by Adobe, that stores structured information directly inside a file using an open format — and unlike fixed-field standards, it can be extended with custom fields a specific workflow needs, not just the built-in ones.
In plain English
Older embedded-metadata standards like IPTC define a fixed list of fields — Creator, Caption, Copyright — and that's it; if your workflow needs a field the standard didn't anticipate, there's nowhere standard to put it. XMP was built to solve exactly that limitation: it's based on an open, extensible format (built on XML/RDF), so a photo agency, publisher or software vendor can define their own custom fields — a project code, an internal approval status, a usage restriction specific to their business — and store them right alongside the standard ones, inside the file itself.
In practice, XMP and IPTC often coexist in the same file: many tools write IPTC's standard editorial fields and layer XMP's extended, custom fields on top for anything the fixed standard doesn't cover. Adobe's own products (Photoshop, Lightroom, Bridge) read and write XMP natively, which is part of why it's become a de facto standard across the wider photo and creative industry, not just Adobe's own ecosystem.
The tradeoff of that flexibility is less certainty: because anyone can define custom XMP fields, two different tools' "custom" fields won't necessarily mean the same thing to each other, unlike IPTC's fixed, universally-understood fields. Custom XMP fields work best within one organization's own consistent workflow, not as a way to communicate meaning to an outside party who's never seen your schema.
Why it matters in a DAM
Most libraries eventually need to track something the standard metadata fields don't cover — an internal project code, a licensing tier specific to the business, a custom approval stage. XMP is what lets that tracking live embedded in the file itself, surviving exports and transfers the same way IPTC's standard fields do, instead of living only inside one DAM's proprietary database where it's lost the moment the asset leaves that system.
Buyer’s test: ask whether the tool lets you define your own custom XMP fields, not just read the standard ones. Some DAM tools read existing XMP fine but only let you add custom metadata in their own proprietary database format — which means that custom data won't travel with the file if it's ever exported to another system.
Related terms
See it in action
See Adobe's own XMP documentation at adobe.com. For a DAM tool that reads and writes both IPTC and XMP fields directly in its metadata panel, see our Daminion review.