The ROI of a design system nobody asked for

Conventional wisdom says wait until you have three products before investing in a shared design system — otherwise you're guessing at abstractions you don't need yet.

We built ours after our first product, and it slowed that first launch down by roughly two weeks. Every launch since has been faster, because the components, tokens, and motion rules already exist and are already accessible by default.

The break-even was product number three. Everything after that has been pure time saved — and, just as important, pure consistency saved, because nobody's re-deriving what a card's border radius should be for the fourth time.