Why I sometimes let a product ship without a design system or tokens

Design grew out of art

Before design acquired grids, components and handoff files, it was drawing, painting and printing. It emerged from art, and art has never cared much for fences. Rules arrived later, because products grew, teams grew, and someone had to prevent a hundred screens from becoming a hundred opinions. That was a sound reason, and it remains one.

Somewhere along the way, however, many of us began to mistake the rules for the work itself. Every project acquires a system before it has found its voice, and every new idea is checked first against the library rather than against the problem.

The sketchbook in the park

Some of the best things I have made began far from any screen, in the manner of a sketchbook. You sit in a park and draw whatever is before you: a bird, a branch, people passing by and the way they carry their day. Nobody asks which token that line uses, because it is not that kind of drawing. There are no ugly drawings in a sketchbook. Nothing in it has to be perfect, because it is not made to be judged. It is a mental exercise; it fills you up, and it trains you to look.

In time, that habit finds its way into the work without being asked. Noticing how someone hesitates before crossing a street is not so different from noticing where a person hesitates on a page. The hundred and one loose drafts, notes and scribbles never become anything deliverable, nor do they need to. They are where the eye stays sharp, and there are no bad ones among them, only things you have learnt from.

Some projects are good precisely because they were permitted to retain a little of that looseness. They found their own shape first, and only afterwards something worth repeating.

We can’t all be robots

A design system is a tool for people who work together over a long period. It is not a moral code. Doing everything by the book, every time, on every project, reduces designers to operators of their own libraries, and the work begins to look as though the library made it rather than a person.

Being loose by intention keeps the hand moving and the eye alert. As it happens, it is also where most of the ideas that later enter a proper system come from.

When I let it go

There are moments when I let it go on purpose. When a product is still discovering what it is, tokens would only freeze decisions nobody has made yet. When the work lives for six weeks, as a campaign or an event does, a system costs more than it saves. When the purpose is to learn, experiments and side projects are precisely where I attempt what a system would never permit, and that is the sketchbook once again.

And when the team consists of one designer and one developer who speak every day, they already constitute the system. Writing it down can wait until there are more of them.

What makes it a decision, not a mess

The difference between freedom and chaos lies in knowing why you are doing it. I can leave a product without tokens because I know exactly what a system would look like, which parts would become components and how I would reconcile everything when the time came. That assurance does not come from rules. It comes from years of building systems, breaking them and mending them again.

Experience gives you the right to make that call. Not because you stand above the rules, but because you know them well enough to see when they would only get in the way.

When I would never do it

Large platforms, many teams, public services, long lifespans. There, forgoing the system is not freedom; it is a bill someone else will settle later, usually a new designer on a Monday morning. On those projects I am the most by-the-book person in the room, and happily so.

So yes, I believe in design systems. I also believe in the park, the sketchbook and the hundred and one drawings nobody requested. Knowing which one a project needs is most of the job.