The design system vision locked this week around a single measure, and the component count lost the vote.
Velocity is the word we locked as the anchor for our design system vision this week. It’s a word that needs defending, because it is easy to hear as a plea to hurry when it isn’t one. The measure we chose has nothing to do with components shipped: nobody outside a design system team has ever cared how many components shipped, in the same way nobody judges a library by its shelving. The measure, instead, is whether the product managers and designers prototyping against the system can build something realistic when their work taps the actual system. And whether they get faster at it week over week, month over month. If they do, the system is working, and if the component count grows while those people slow down, we have shipped furniture.
“Can they do that faster and faster, week over week, month over month?”
Committing to one measure settled arguments we’d been going back and forth on for months. Take the question of what to build next: we’re mapping every piece of design system work through a jobs-to-be-done lens and sorting which jobs an agent can carry and which need a human’s judgment, and that map sets the priorities. (An agent can propose the accessible color ramp… a human decides whether the brand can live with it.) This retires the older method, which was prioritizing by whoever argued longest, a method with deep organizational roots and a poor track record.
The ownership model was the other anchor, and it’s the part I expect to find personally difficult. I write the high-level vision, and that is where my drafting ends. The design system lead takes the vision on a listening tour, hears where teams actually catch on it, and returns with an action plan, and my job at that point is to approve. I said it in the meeting almost exactly this way: I’m the approver of it versus the drafter of it. It sounds like less work, and it’s actually the demotion of my favorite hobby, which is drafting other people’s plans for them… a hobby I’ve been assured nobody else enjoys.
There’s a longer game here too. If velocity is the product, the design system stops being a catalog we maintain and becomes a performance we can chart. A chart is honest with you in a way a catalog never is. A catalog always looks complete, while a chart pointing the wrong direction ruins your morning early enough for the fix to be cheap. My job from here is mostly to keep the question from softening into a slogan.
The teachable part
Pick the one measure your design system exists to move, write it at the top of the vision, and let it veto the roadmap. Ours is the velocity of the people building on it.