




Design Systems for global SaaS product.
Context. I inherited a landscape of more than 20 products with very little consistency, commonality or maturity in their user experience. A legacy design system offered some building blocks, but not the modern structure we needed for the next phase of the product's evolution.
Rather than simply redesign the design system, I used the moment to rethink how the platform was built, aligning product, design and engineering goals behind one approach. We called it Signal.
One sketch became our north star for years. It captured the approach I believed we needed, and a lot of effort went into cementing it in leadership's minds as the strategy.
Modernising the end-user experience couldn't be done by design alone. It needed product support and engineering buy-in. Breaking the platform into layers gave each group clear accountability, and let us move forward without everyone working on the same thing at the same time.
What design owned. Within design, Signal came down to three things:
A clear design language
A documented set of decisions and guidelines
A new set of tools and workflows
This was a multi-year effort built on incremental value. Teams kept delivering, while every release moved them closer to a common standard.
Crucially, the three areas gave designers on the ground room to raise, propose and often deliver improvements within their own products, contributing to the wider platform as they went. The power was in that local confidence: change was demonstrated team by team, not pushed through a central governance function competing with everyone's priorities.
We took back control of the experience. Before Signal, the most senior person in the room decided how the product looked and behaved, usually a technical or sales leader. The design language gave us a basis for consensus, and design became the recognised voice of how the platform should look and feel. Once a feature was agreed, design held formal sign-off on its experience.
We delivered confidence and consistency. Documented standards and evidence, open to the whole business, became the go-to place to back up decisions and answer questions about the platform. Being able to explain why a decision was right cut through opinion and ambiguity, which meant faster decisions and a more usable platform. Over the course of the strategy, the platform's ease-of-use rating rose from 3 to 8 out of 10. Queries and "suggestions" coming through product feedback dropped, and small backlog projects fell by 50%.
We worked smarter and faster. It's what every business wants to hear, but we did it. Clear design operations, consolidated tools, new ones to cut manual work, and a defined way of working with engineering. Designers went on to create approved prototypes five times faster, on average, than before Signal. Designers and engineers both became more confident about what was needed, which freed up space for innovation and exposed gaps in the platform.
Reflection. Signal worked because we never treated it as just a design system. It was a platform strategy, and design was one layer of it, whilst being so much more than a component library. That distinction is the thing I'd pass on to anyone starting the same journey.
"Design system" gets offered as the perfect fix for efficiency and consistency. But on its own it can't solve deep-rooted complexity, especially in a business built from acquisitions, legacy platforms and fragmented tech stacks. There, the real challenge isn't design. It's business and technical architecture.
Looking back, the better question wasn't "do we need a design system?" It was "do we need experience guidelines and a platform convergence plan first?" In practice, that means four moves:
Experience principles - shared guidelines that create behavioural consistency without forcing pixel-perfect alignment (Supernova).
Design tokens only - harmonise the brand feel, colour and typography, without building full component libraries (Tokens Studio).
Tech architecture alignment - before spending energy scaling components across incompatible platforms.
Product rationalisation - to decide whether every product needs to coexist long term.
Only once these are moving does a full design system make sense, and only if it earns its return.
Without a unified business, product and tech strategy, you end up forcing a design system on top of fragmentation, and it breeds frustration rather than efficiency. It's like building beautiful roads on separate islands with no bridges between them. Each island looks great, but nobody gets across the ecosystem any faster.
