Taxonomy · Metadata
One vocabulary, 25+ teams
Shared metadata taxonomies adopted across 3D, video, photography, graphics, AI-generated and stock-imagery teams on a 2M+ asset DAM.
- 25+
- teams on the shared taxonomy
- 2M+
- assets in the DAM
- ~50K
- new assets a week
The problem
Every creative discipline names things differently. A 3D team, a product photographer and a video editor will describe the same product three ways, and search fails all three of them.
Approach
- Start from the questions people actually ask: which product, which campaign, can I use it here?
- Keep facets separate: project, product, rights, usage and content credentials (C2PA) each get their own field, not one overloaded tag list.
- Let AI fill what it's good at (smart tags, BAM-generated fields) and keep people on the fields that need judgment.
- Roll out team by team, with onboarding and one-on-one support.
Migrations
I moved several teams' libraries into AEM, hundreds of gigabytes each, renaming and restructuring them to automation-ready conventions so downstream tools could rely on the names. I also contributed to a metadata enrichment project that makes legacy assets reusable instead of re-shot.