Rather than arriving with a diagnosis, I ran an anonymous survey across all 15 designer.

The survey served three purposes:

To signal that the process was changing and their input would shape on what is to come.

To understand where the real gaps were between what the system offered, and what designers actually needed.

To gather data I could bring to management that framed the problem as a systemic one and not the failure of the designers who built it.

All 15 designers responded. The results gave me something more useful than a list of problems. They gave me a mandate.

Helps me work more efficiently

Meets their project requirements

Confidence using the system

The open text responses were more specific. Tabs styled like segmented controls, a tertiary button in light red that read as an error state, chips indistinguishable from buttons with no differentiation between single and multi-select. Designers weren't avoiding the system out of preference. Using it correctly meant compromising the work.

We introduced async check-ins with design teams before anything went to a development . Designers could try new components in their work and give feedback before anything was locked. The pipeline became transparent. Designers knew what was in design, what was in development, what was published. The process change mattered as much as the components themselves.

That shift from a system imposed on designers to one built with them is what I'd point to as the real work. The components and documentations were the output. The collaboration structure was the fix.