Building a Design Function
From Zero

Building a Design Function From Zero

There was no design system, no shared process, no seat at the table. A year later, design had a roadmap tied to business metrics, not just “make things prettier.”

Year

Year

2025

Contributors

Contributors

2 Designers, 5 PMs

My role

My role

Project owner

Challenge

When I first joined Fairmatic, design operated largely as a service function. We received requirements from PMs and developers and focused on shipping within timelines, often without challenging or digging deeper into the underlying “why” behind what we were building.

Over time, this approach resulted in a chaotic product - a growing collection of features with little consistency or cohesion, as designers largely worked independently. The lack of a shared design vision and system ultimately made the product confusing and difficult to use for customers. By the time these issues became apparent, they were deeply embedded in the product, making them difficult and costly to fix quickly.

Building the Operating System

The team was shipping constantly, but nobody could say whether any of it was working no usability baseline, no shared vision beyond the current sprint. Instead of picking one more feature to design, I built a framework for how design would work as a function going forward: six initiatives - a design system, a product vision, benchmarking, writing guidelines, an insights repository, an innovation hub. Each with an owner, a metric, and a priority.

Winning Individual Teams Over

The hardest resistance was “too much effort, too little return.” I couldn't win that with a slide deck. I had to win it by showing them and debating each and every detail about the initiative. One small comparison for each initiative made the case for the whole system better than any argument about the initiative as a concept.


Example :- How a simple plus/minus control on number inputs as a component in the design system could save users a whole lot of trouble.


Even simple things like that could make a difference for our customers and that could only be shown by illustrating it.

Scaling It, and Growing Others

Once the system had proof, it scaled across five product teams - not just the components and initiatives, but the reasoning behind them. I worked directly with the designers on each pod: turning casual walkthroughs into stakeholder-ready presentations, and pushing them to own component decisions themselves instead of asking permission for every one. That ownership is what made five teams stick with it, not just mine.

It stayed consistent through a simple rule, not constant policing: any new component had to be contributed back when a team shipped a new project, and I approved every addition from the design side. Everything else stuck to what already existed with regular office hours for anyone stuck between the two.

Within a few months, the quality of the design had already increased with the principles and components being adopted to our new products :

Once the system had proof, it scaled across five product teams - not just the components and initiatives, but the reasoning behind them. I worked directly with the designers on each pod: turning casual walkthroughs into stakeholder-ready presentations, and pushing them to own component decisions themselves instead of asking permission for every one. That ownership is what made five teams stick with it, not just mine.

It stayed consistent through a simple rule, not constant policing: any new component had to be contributed back when a team shipped a new project, and I approved every addition from the design side. Everything else stuck to what already existed with regular office hours for anyone stuck between the two.

Within a few months, the quality of the design had already increased with the principles and components being adopted to our new products :

What happened

65%

of new projects shipped using the design system in year one

8%

Increase in SUS score for overall product

200+

insights logged in a shared repository

Design strategy isn't better opinions on visual details. It's building the infrastructure that lets a team's opinions be tested, measured, and trusted.