Short answer
A controlled vocabulary is a fixed list of approved values for a metadata field. Controlled values prevent synonyms and typos that break search and filters. A controlled vocabulary belongs on every field used as a search filter, because one stray spelling hides every asset tagged with it from that filter.
How a controlled vocabulary works in a DAM
A controlled vocabulary works by limiting a metadata field to values that an owner approves in advance. Users pick values from a list, and the DAM rejects anything else on upload, bulk edit and API import.
Each value usually carries three parts:
- A stable code that integrations and the API read.
- A label per language, such as Dutch, German and English.
- Optional synonyms that point users to the approved value.
The ANSI/NISO Z39.19 guideline describes several forms of controlled vocabulary. DAMs use four of them.
| Form | Structure | DAM use |
|---|---|---|
| List | Flat set of values | Channel, asset type, region |
| Synonym ring | Equivalent terms grouped together | Search expansion |
| Taxonomy | Values in a parent and child hierarchy | Brand and product line |
| Thesaurus | Hierarchy plus related and preferred terms | Subject keywords in large libraries |
Example of a controlled vocabulary
Worked example: a 'Channel' field limited to Web, App, Print, Marketplace and Social. Before the change, the same field was free text and held 'Web', 'website', 'online' and 'eCom' for one meaning. A filter on Web returned only the first group of assets.
After the switch, a librarian maps the old spellings to the five approved values in one bulk edit. The Channel filter then shows five entries, each with a complete result set.
Controlled vocabulary in a DAM selection
A controlled vocabulary is designed at Checkpoint 2: Define and proven at Checkpoint 5: Test. The buyer lists every field that needs a fixed list and names an owner for each list.
Ask each vendor to show these behaviours on your own lists:
- Import a list of values with codes and labels from a CSV file.
- Show labels in two languages while the stored code stays the same.
- Rename one value and confirm every tagged asset updates.
- Restrict who may add new values, and route suggestions to the list owner.
- Constrain auto-tagging output to the approved values.
The metadata model builder records each list with its owner. The DAM taxonomy guide explains when a flat list should become a hierarchy.
Common controlled vocabulary mistakes
The most common controlled vocabulary mistake is letting every user add values freely. That setting turns the list back into free text within months. Other recurring errors:
- One list mixing several concepts, such as a Channel list that also holds campaign names.
- Lists with hundreds of values in one dropdown, which users stop reading.
- Values without codes, so translations or renames break integrations.
- No owner, so outdated values stay in the list after products or brands end.
Related terms: Metadata schema, Taxonomy.
Frequently asked
What is the difference between a controlled vocabulary and a taxonomy?
A controlled vocabulary is any fixed list of approved values for a field. A taxonomy is a controlled vocabulary whose values sit in a parent and child hierarchy. Every taxonomy is a controlled vocabulary, while many DAM fields such as Channel need only a flat list.
Should users be allowed to add values to a controlled vocabulary?
Users should suggest new values, and a named list owner should approve them. DAMSelection rule of thumb: route suggestions through a short request workflow and review the list once per quarter. Open lists without approval drift back into synonyms and typos.