
Konstantin Semenenko
July 21, 2026
4
minutes read
Only one of these is a design system tool. Figma holds the actual system: components with variants and states, auto-layout, design tokens or variables, published libraries with versioning, and developer handoff. Mural holds the thinking around the system: the audit of what exists, the naming debates, the governance rules, the adoption roadmap, and the workshops where teams agree on principles. Trying to run a design system in Mural produces pictures of components rather than components. Use Mural to decide and align, Figma to build and publish, and keep the two linked.




The Mural versus Figma question for design systems has a short answer: Figma is where a design system lives, and Mural is where the decisions about it get made. Figma provides the machinery a system actually requires, reusable components with variants and states, auto-layout for responsive behavior, variables and tokens for color, spacing, and type, published libraries with versioning and update propagation, and structured developer handoff. Mural has none of that, and it is not attempting to: it is an infinite collaborative canvas for the work that surrounds a system, auditing the inconsistency you currently have, arguing about naming, defining governance and contribution rules, and running the workshops that get a distributed team to agree. Both jobs are real, and confusing them is why some design system efforts produce beautiful boards and no adopted system. This explains the split.
We build and consume design systems on client projects, so this is a practitioner's view of which tool does which part and how the handoff between them works.
A design system is not a collection of screenshots, it is a living, versioned, consumable set of components and tokens that designers pull into files and engineers implement against. That imposes concrete technical requirements. Components must be reusable instances that update everywhere when the source changes, with variants for states (default, hover, disabled, error) and properties for configuration. Layout behavior must be encoded, so components respond correctly rather than looking right at one width. Tokens must be named and referenced, so a color change propagates rather than being hunted down. Libraries must be published and versioned, so teams adopt updates deliberately instead of drifting apart. And handoff must expose measurements, tokens, and structure to engineers.
Figma provides all of that natively, which is why it became the default home for design systems. Mural provides none of it, because it is a thinking canvas: shapes and text on an infinite board with real-time collaboration. A rectangle in Mural labeled "primary button" is a picture of a button. It cannot be instanced, it has no states, it carries no token, and changing it changes nothing else. That is not a shortcoming to work around; it is the difference between a system and a diagram of one, and it settles the question of where the system itself lives.
The part teams underrate is how much design system work is not design work. Before components get built, someone has to answer: what do we actually have today, how many blues are in production, which patterns repeat with slight differences, what should the naming convention be, who is allowed to contribute, how do teams request changes, and in what order do we roll this out. Those are alignment and governance questions involving designers, engineers, product managers, and often leadership, and they are exactly what Mural is built for.
Concretely, Mural works well for the inventory and audit phase (pulling screenshots of every existing pattern onto a wall and clustering the duplicates, where its AI clustering genuinely saves time), the naming and taxonomy debates, the governance model (contribution process, review gates, ownership), the adoption roadmap across teams, and the workshops that build buy-in from people who will not open Figma. Design systems fail more often from lack of adoption than from lack of components, so the alignment work Mural supports is not a preamble to the real work, it is a substantial part of whether the system succeeds, the same collaboration-versus-execution split we lay out in Mural vs FigJam vs Figma.
Two failure modes recur. The first is building the "system" in Mural: creating a board of every component as static shapes, treating it as the source of truth, and discovering later that nothing is reusable, nothing propagates, and engineers have no spec to build from. The board looks like a design system and functions as a poster. The second is skipping Mural entirely: a designer builds a technically excellent Figma library alone, publishes it, and nobody adopts it because the teams who have to use it were never part of deciding what it should be or how contribution works.
Both failures come from the same misunderstanding, that a design system is either a set of components or a set of agreements, when it is both. The components have to exist as real, versioned, consumable artifacts, and the agreements have to exist as shared understanding across the people who will use them. Figma cannot create the agreements and Mural cannot create the components. Teams that succeed run both tracks and connect them.
The sequence that holds up in practice:
The link matters in both directions: the Mural board explains why the system is the way it is, which is what stops a new team from relitigating settled decisions, and the Figma library is what they actually consume. Engineers, meanwhile, need the tokens and structure that make implementation faithful, the readiness question we cover in design-to-code in 2026.
Figma owns the design system; Mural owns the decisions around it. A system requires reusable components with variants and states, auto-layout, tokens, published versioned libraries, and developer handoff, and Figma provides all of it while Mural provides none, because Mural is a thinking canvas rather than a design tool. What Mural does provide is the audit, the naming and taxonomy debates, the governance model, the adoption roadmap, and the workshops that create buy-in, and since design systems fail more often from non-adoption than from missing components, that work is not optional. Build the system in Figma, make the decisions in Mural, link them so the reasoning survives, and run both tracks rather than choosing between them.
If you want a design system that gets built properly and actually adopted, that is where our AI Design Team work starts.
Can you build a design system in Mural? No. Mural has no reusable components, variants, states, auto-layout, design tokens, published libraries, or developer handoff. A shape labeled "button" in Mural is a picture of a button, not an instance that propagates changes. Design systems are built in a design tool like Figma.
What is Mural useful for in design system work? The work around the system: auditing existing patterns and clustering duplicates, agreeing naming and taxonomy, defining governance and contribution rules, planning the adoption roadmap, and running the workshops that build buy-in with people who will never open Figma.
Why do design systems fail? More often from lack of adoption than lack of components. A technically excellent library nobody uses is a failed system, which is why the alignment, governance, and buy-in work is substantial rather than a preamble. Teams that only build components frequently hit this.
Should the design system documentation live in Mural or Figma? Usage guidance belongs alongside the components in Figma, where designers actually work, so it stays in view. Keep the decision history and governance discussions in Mural, so new teams can see why the system is the way it is instead of relitigating settled choices.
Do we need both Mural and Figma for a design system? Not strictly, but the two kinds of work both exist. Figma is required for the system itself. The alignment and governance work can happen in Mural, FigJam, or another collaborative space, and the right choice depends on whether your organization runs structured facilitated sessions across non-designers.


