A design system is a product. It needs an owner, a roadmap, a maintenance budget, and a customer — your other product teams. If you can't staff it like a product, you're not building a design system; you're building a graveyard of half-finished components.
The good news: most teams that ask the question already need one. The bad news: most teams that build one start with the wrong layer and burn out before adoption catches up to investment.
The three-surface test
Count the surfaces in your product portfolio that share users, share components, or share a brand. If the answer is one, you don't need a design system — you need a well-organized component folder. If the answer is three or more, the cost of inconsistency is now higher than the cost of building the system.
Surfaces means distinct products or sub-products: marketing site, signed-in app, mobile app, admin tool, partner portal. Not pages. Not features.
When the answer is no
If you have one product, a small team, and pre-product-market-fit velocity, a design system slows you down. The shape of your UI is still being discovered; codifying it now means deprecating most of it in six months.
If you have a marketing-only surface with bespoke campaigns, a system constrains the creative work the surface exists to do. Build a brand kit instead.
If your product team is two designers and four engineers, the meta-layer of the system will eat the ship-the-product layer. Wait until growth forces it.
When the answer is yes
Multiple products with overlapping users. A growing engineering org with new hires every month asking 'where's the button component?'. Visible inconsistency in production — the same dialog rendered three different ways, the same form field with three placeholder colors.
These are signals that the consistency tax has crossed the line where building a system is cheaper than not having one.
Run the math
Use the ROI calculator to size the bet. The default assumptions are conservative; plug in your real headcount and shipping cadence. The output isn't a precise number, it's a debate-anchor for the conversation with your director.
- 01Three or more surfaces sharing users, components, or brand is the threshold.
- 02Pre-product-market-fit and tiny teams should hold off — the system will calcify the wrong shape.
- 03If you can't staff it like a product with an owner and a roadmap, you can't sustain it.
- 04Use the ROI calculator to anchor the funding conversation, not to pretend you have a precise number.