Tags, learning, and names make boards easier to search

The more a reference library grows, the more personal its structure becomes. Custom tags, learned suggestions, and face naming help Reference remember the words and people that matter to you, so a large board can stay searchable without asking every saved item to become a filing task.

Reference Board showing a visual board with item details, custom tags, suggested tags, and named faces
Custom tags and learned names give larger libraries a little more memory.

Structure in your own words

Reference has always tried to keep organization close to the work instead of turning it into a separate chore. Automatic descriptions, colors, and search help a board stay useful in the background, but there are still moments when the best label is the one you would have chosen yourself.

Custom tags are for those moments. You can add the language that already belongs to a project: a campaign name, a material, a client, a mood, a location, a product detail, or any small category that only makes sense inside your own library.

That matters because creative references rarely arrive with clean metadata. A photo might be useful because of the chair in the corner, the way light hits a wall, or the kind of person it reminds you of. A custom tag lets you keep that reason attached without forcing the whole board into a rigid system.

Suggestions that learn from you

The useful part of tags is not just that you can add them. It is that Reference can start to understand the structure you keep returning to. As you name things, tag items, and build up a library, suggestions can become more familiar and more specific to the way you work.

That kind of learning should feel quiet. The app is not trying to take over the board or decide what the project means. It is trying to reduce the small moments of friction that appear after a library gets large: remembering the exact tag you used last time, keeping similar ideas connected, or surfacing a label that already belongs to the material in front of you.

The goal is a softer kind of organization. Not a perfect taxonomy. Not a form you have to fill in every time you save something. Just a system that pays attention to the patterns you already create.

People are part of the reference

Some references are useful because of who appears in them. That might be a model, an artist, a collaborator, a founder, a public figure, or someone whose expression, styling, posture, or presence carries the feeling you are trying to remember.

Face recognition and naming make those references easier to return to. When a person appears across saved images, you can give that face a name and let Reference connect the related material. Later, the name becomes another way back into the board.

This is especially helpful when visual memory is stronger than verbal memory. You might remember a person from a campaign, a shoot, a presentation, or a collection of images without remembering the filename, source, or caption. Naming gives that memory a handle.

Easier to find later

Together, custom tags, learned suggestions, and face naming make Reference better at the second half of collecting: coming back.

Saving is the fast part. You see something, it matters, and you keep it. The harder part comes days or months later, when the board has grown and the reason something mattered has become less obvious. Search gets better when the app remembers more of your language, your patterns, and the people that appear across your material.

That does not mean every board needs to be carefully managed. The whole point is to make a growing library feel easier to revisit without making the act of collecting heavier.

Reference should still feel like a place to think visually first. These features just give that thinking more memory, so the things you save can find their way back to you when they become useful again.