The Design System Nobody Uses

A top-down view of a crisp coral-edged paved walkway with a worn trail of footprints veering off and bypassing it across bare ground, representing a well-built design system that teams work around instead of using.

"We have a design system. Nobody uses it."

A design lead told me that last year, half proud and half defeated, the way you'd describe a gym membership. The system was real. It had components, tokens, a documentation site, the whole kit. It also had the quiet gravity of a building everyone walks past. Her team kept rebuilding the same button three different ways in three different files, next to a library that already had the right one.

The common story says this is a quality problem. Build a better system, document it harder, add more components, and adoption follows. I used to believe that. It is wrong, and the data is not subtle about it. In zeroheight's Design Systems Report 2025, a survey of 294 practitioners, only 10% of teams said their design system was fully adopted across the whole company. Half were not even "widely" adopted. The systems exist. The use does not.

Building is the easy half

Here is the uncomfortable pivot: making a design system is the part we are good at. It is design work, and designers like design work. Naming tokens, drawing components, setting up the Figma library, that is a satisfying, legible job with a clear finish line.

Earning adoption is a different job entirely, and it is not design. It is closer to internal product management and sales. You have to convince busy engineers to stop doing the thing that already works for them and depend on something you maintain. You have to keep it current while they ship. You have to make using the system genuinely faster than ignoring it, because the moment it is slower, people route around it, the same way a worn footpath cuts across a lawn the landscaper was proud of.

The report backs this up from the other side. When zeroheight asked teams their single biggest challenge, the top answer was not "our components aren't good enough." It was lack of resources and staffing, at 63%. More than one in five teams, 22%, had no dedicated design system people at all. We treat the system as a one-time build and then act surprised when an unstaffed, unmaintained artifact loses to the path of least resistance.

Adoption is a relationship, not a launch

A design system is not a deliverable you hand off. It is a product with internal users who can quit at any time, for free, with nobody noticing. That reframe changes what the work actually is.

It means you measure the right thing. Encouragingly, adoption is now the metric teams most want to track, with 57% in the zeroheight data naming it their key measure of success. The problem is the follow-through: the tooling for it is, in their words, fractured, so a lot of teams fall back to a vague gut feeling about whether anyone is using the thing. You cannot improve a number you are guessing at. Component coverage, detach rates, how often the "right" component is the one in the file, these are the vitals. If you are not watching them, you are flying blind on the one thing that determines whether any of the build mattered.

It also means the system has to pay rent continuously. This is where it connects to the designer-developer gap I've written about before: a system only gets adopted when it lives in both the designer's Figma and the developer's codebase as the same source of truth, not as a design artifact that engineers admire and then re-implement by hand. A library that is beautiful in Figma and absent in production is not a system. It is a mood board with governance.

And it means you treat neglect as a cost. An unused system does not sit still. It rots into the kind of design debt that compounds as a product scales, because now you are maintaining the official system and the three shadow versions teams built because the official one was too slow to reach for.

What I actually do now

On product work like the Monety fintech platform, where I was designing across dozens of connected screens, the system only earned its keep when using it was the fastest path to a shipped screen, not a tax on top of one. That is the whole test. Not "is the system comprehensive." Not "is the documentation thorough." Just: when a designer or an engineer sits down to build the next thing, is reaching for the system the easiest move in the room?

If the answer is no, no amount of new components will save it. People are rational. They take the quick path. Your job is not to shame them off the desire line. It is to pave the line they were always going to walk.

So before the next big system initiative, before the token audit and the component sprint, ask the harder question first. Not "what should we build?" but "why isn't the last thing we built being used?" The answer is almost never that it wasn't good enough. It's that nobody made it the easy choice.

What does your team rebuild from scratch that your design system already has?

Share on

Let's talk

I like to connect and see how we can work together

All trademarks, logos, and brand names are the property of their respective owners. All company, product, and service names used on this website are for identification purposes only. Use of these names, trademarks, and brands does not imply endorsement.

© 2026, Felipe Linares - flinbu. All rights reserved. | Terms and Conditions | Privacy Policy | Cookies Policy

Let's talk

I like to connect and see how we can work together

All trademarks, logos, and brand names are the property of their respective owners. All company, product, and service names used on this website are for identification purposes only. Use of these names, trademarks, and brands does not imply endorsement.

© 2026, Felipe Linares - flinbu. All rights reserved. | Terms and Conditions | Privacy Policy | Cookies Policy