Reference Glossary
Faceted search
A search interface where users narrow results by combining independent metadata filters — file type, usage rights, colour, date — with result counts updating live as each filter is applied.
Why it matters in a DAM
Facets are only as good as the structured metadata behind them. A 'usage rights' facet lets a marketing team exclude expired-licence assets before they ever show up in results, but if editors log rights status inconsistently, the same facet either hides usable assets or lets expired ones through. Facet reliability is really a live read-out of how disciplined the tagging process has been, which is why it breaks first when governance is weak.
A worked example
Common mistake
Teams build facets on free-text fields instead of controlled vocabularies, so one attribute splits into near-duplicate facet values like 'Red', 'red', and 'RED' — fragmenting what should be a single filter into three that each undercount the real total.
Faceted search lets someone drill into a library by stacking filters — file type, then usage rights, then colour — with the system recalculating how many assets match at each step. It’s the interface pattern behind the filter sidebar on most DAM search pages, and it’s what turns a search of 12,000 assets into a usable shortlist of a few hundred without anyone typing a keyword.
The catch is that every facet is a direct reflection of a metadata field, and it only works as well as that field is populated and normalised. A facet built on a controlled-vocabulary field (a dropdown with fixed values) behaves predictably. A facet built on a free-text field inherits every inconsistency in how people typed into it — different capitalisation, plurals, regional spelling — and each variant becomes its own facet value instead of collapsing into one.
Because of that dependency, faceted search is often the first place metadata governance problems become visible to end users, well before anyone runs a formal audit. If a ‘department’ facet shows fifteen near-identical values instead of five, that’s a tagging discipline problem surfacing through the UI, not a search bug.
Frequently asked
What is faceted search, mechanically?
A search interface where users narrow results by stacking independent metadata filters — file type, usage rights, colour, date — with the system recalculating how many assets match at each step, turning a search of thousands into a shortlist of a few hundred.
Why is a 'usage rights' facet only as reliable as the underlying tagging?
It lets a marketing team exclude expired-licence assets before they show up in results, but if editors log rights status inconsistently, the same facet either hides usable assets or lets expired ones through — the facet just reflects the metadata behind it.
What happens when facets are built on free-text fields instead of controlled vocabularies?
One attribute splits into near-duplicate facet values like 'Red', 'red', and 'RED', fragmenting what should be a single filter into three that each undercount the real total.
Why does faceted search often surface metadata governance problems before a formal audit does?
If a 'department' facet shows fifteen near-identical values instead of five, that's a tagging discipline problem becoming visible to end users directly through the UI, not a search bug.
How does a facet built on a controlled-vocabulary field behave differently from a free-text one?
A controlled-vocabulary field, like a dropdown with fixed values, behaves predictably because it inherits no typing inconsistencies, unlike a free-text field where every capitalization or spelling variant becomes its own separate facet value.
What's a typical example of narrowing results with faceted search?
Starting at 12,400 assets, filtering by file type = Image narrows to 8,100, adding usage rights = Royalty-free narrows to 3,250, and adding colour = Blue-dominant narrows to 410.