Tag, you're it
how to build a DAM tagging strategy
Before we get too far into the weeds, let’s discuss what tags are, and what they are not.
Tags are unique identifiers applied to assets that are organized in a nesting structure. They are pieces of data consisting of only 1 to a few words. They are part of your metadata strategy. Think of animals. I have children so I’m often learning animal facts against my will. All mammals are animals; not all animals are mammals. All koalas are mammals; not all mammals are koalas. The logic can go on (and on, and on) to include mammal subcategories of marsupials vs primates, with koalas in the former. Then Latin gets involved.
Quite a variety of animals under the top-level structure of Mammals, but a fairly straightforward tagging structure. In the koala’s case: animal - mammal - marsupial - koala. In a gorilla’s: animal - mammal - primate - gorilla. You would be able to create a tagging structure with clear levels:
animal
|
mammal bird fish reptile amphibian etc.
|
marsupial primate rodent etc.
|
koala kangaroo wombat etc.
Pretty straightforward, right? But maybe you also want to classify your animal assets by location, usage rights, photographer name, or style (color vs b/w, square/portrait/landscape). You will want to build a tag structure that can apply to all your assets uniformly. With these classifications as an example, your tag structure could look like this:
By using a sheet to build the structure, you can share the work and build on it as your asset library grows. Since some DAM systems are finicky when adding tags, I like to workshop any questions in the sheet and only build in the system once decisions are made.
Now we are mostly not dealing in animal classification - if you are, I have a child who would love to discuss turtles with you - so how does this apply to the digital assets we manage?
The good news is that the tools and rules for tag structures apply across subject matter. Breaking down whether a tag is categorized as subject, style, or usage (plus additional categories like location, campaign, asset type, etc.) allows you to expand the subcategories as new assets and asset classes are included. In the example above, I know where to add more mammals, photographers, or kinds of snakes as the library grows without having to restructure what has already been tagged.
So what is not a tag? I avoid adding subjective terms (creative, furry, brown) to any tag libraries I build. Those might be useful keywords but are not distinct or precise enough to tag. My tip? The tag should answer if something is definitely THIS or THAT. While AI can identify many characteristics of assets as well as art directors when it comes to image description, I prefer to have both inputs saved as either a secondary descriptive tag list or keywords to help with search while maintaining a tight primary tag structure.
Sharing the tag structure across your team will help you keep it current. If there are new campaigns or topics used in your organization or company, adding them to the tag library will help asset managers apply them to new sets of assets at ingest and to existing assets that are part of that campaign. Some DAM platforms have a “suggest a tag” option for all platform users to give feedback directly. Like other aspects of DAM success, communication and collaboration will be the best ways to keep the tag structure a valuable tool to enrich the assets powering your DAM.