The Local Design System

Most of the internet looks exactly the same—at least the parts that can be considered to be the commercial mainstream internet. We see the same buttons, same popovers, and, to some degree the same colors and typefaces. But why does this happen? Digital media is inherently omnipotent, so why don’t we see a richer expression of its capabilities? How did we end up converging towards a visual and functional monoculture? And most importantly, could we imagine a way out of this?

In large parts, I think design systems are to blame. Given the title of this website, this might come as a shock, so let me substantiate: I think there’s an industry-wide misconception about what a design system is, what it can achieve and how it should be approached. In short, the term “design system” has come to mean a collection of ready-to-use components like buttons, form inputs, etc, themed using the organization’s brand design tokens.

While this demotes “design” to mere styling, it more consequentially exposes a detrimental understanding of systems, by reducing the “system” aspect of “design system” to a checklist of components and their functional requirements. This reflects organizations’ desire for control and predictability in an inherently chaotic world, resulting in every company using some version of the same system. A system so generalized that some industry experts have started contemplating the idea of creating a “global design system” to solve design systems for good.

This begs the question, what actually is a system and how global can it be?

Cybernetics is the interdisciplinary study of what constitutes systems, how they operate and how complex behavior can emerge from the interaction of simple components. Systems are understood as sets of interconnected components working together toward a specific goal.

A system also encapsulates its components. What’s inside this boundary is part of the system, while what’s outside is not, just like cells shield their internals using a membrane. This mechanism also applies to non-biological systems.

The system-environment boundary is subject to iteration—through evolution in biological systems and through communication processes in non-biological systems—allowing them to adapt and evolve over time. What seems like a simple mechanism is what enables systems to become self-steering and emergent, producing outcomes greater than the sum of their individual parts.

Systems thinking culminates in Niklas Luhmann’s complex but epochal systems theory, where the sociologist focuses on understanding every aspect of society through the lens of self-steering systems, coming to the conclusion that in order for a system to maintain its ability to act, it needs to reduce complexity.

Systems simplify the complex world around them using system-specific symbols and meaning. Instead of dealing with every unique situation or detail, systems can focus on shared ideas that help them make decisions and take actions. Artists, for example, have a common understanding of what “art” means, while the political system revolves around the generalized idea of power.

A global system, in turn, would either be completely overwhelmed by the infinite diversity of its environment, unable to pick and generalize, sending it into stasis, ultimately rendering it useless. Or—in order to remain capable of action—the global system would reduce and generalize environmental complexity so much that all that’s left is standardized averages.

While standardization can certainly streamline processes and create efficiencies, this extreme reduction of complexity comes at the expense of ignoring anything that deviates from the average, overlooking the richness and complexity of local contexts.

James C. Scott’s book “Seeing Like a State” gives the famous example of the Prussian state treating their forests as a global system through the lens of maximizing timber yield in the 18th century. Local contexts, such as soil conditions, exposure to the elements or the presence of vermin were ignored completely, while uniformity was imposed instead. The result was literal monoculture, rapidly depleted soils and thousands of acres of forest destroyed in the process.

While design system monoculture might not directly destroy acres of land, it surely limits organizations from realizing their full potential, locking them into an averaged local maximum, instead of taking their own bespoke needs and local contexts into account.

Attempting to create a global design system is as futile as trying to establish a “global chair”, a “global house”, or a “global typeface”—a finite solution in a landscape of ever-changing needs, changing budgets, changing tech stacks, changing legal requirements, etc. For a design system to create true value, the system’s fluid nature, as well as the organization’s local contexts, need to be embraced, rather than hidden under the blanket of generalization.

There cannot be a definitive guide on how to create a local design system, as such a guide by its nature would have to claim to be “global” again, and as we’ve seen above, that would inevitably fall short. However, there are some principles that I think can guide thinking about design systems in a more local way.

1. Systems are about connections, not artifacts. Just like social systems operate using a generalized symbolic medium, design systems operate using shared knowledge to reduce complexity and remove friction between the organization’s different parts: from design, to marketing, to engineering.

2. Think of systems as living things. As we’ve seen, systems are subject to constant iteration. A robust system embraces this ephemeral nature by prioritizing workflows over artifacts and guidelines.

3. Resilient systems are self-steering. Therefore, design systems should be governed like open-source projects. The people who work with the system daily are much better equipped to steward its evolution than some checklist an industry expert compiled.

4. Design systems are more than the sum of design and systems. They begin long before any “traditional” design takes place, before any HTML or CSS is written, even before Figma artboards are created. At their core, design systems encapsulate rules to govern processes, workflows, visual appearance, language, interaction, and technical best practices. These rules acknowledge change as the only reliable constant, grasping what can be controlled, while embracing what cannot.

While favoring thinking through local contexts over generalized global solutions will increase the initial effort, the benefit of creating a more resilient and flexible domain-specific system will pay off rather quickly. Instead of forcing organizations to retrofit their idiosyncrasies to match some generalized pattern, a local system will truly put itself at the organization’s service, streamline workflows, and uncover new efficiencies. A local design system will also allow for a broader exploration of designing for the digital world, allowing organizations to set themselves apart from the visual and functional monoculture of today’s internet.