
This may be contentious to some but hear me out. What I’ve learned over the last few years is that we have a choice about what we consider the source of truth, we all have differences in our teams, skills, and time; there’s more than one way to make a design system. Part of that is how we scope what we consider our design system to be.
To me, the most successful scenario is when code and design are as tightly intertwined as they can be so that each component is the same between them (tools like Framer are interesting in exploring this space); there’s as little hand-over as possible and communication is clear and responsive to feedback. A clear workflow would get us from conception to release of code to the live site in a collaborative way and we’d all be talking the same language. We all have different circumstances for design system projects and so that might sound a little utopian!
If we go so far as having our design library in great shape and well documented, we should be really proud of that but I’d challenge that this is still an aspiration; this is what we hope will end up live but there’s an inherent disconnect whether we have a multidisciplinary team or work with disparate product teams. It’s not yet real as our end users, our audience(s) aren’t yet using it and other factors we might come across while getting our ideas realized can impact our designs such as technical constraints or results of user testing.
When there’s handover of some kind, there’s the ability for intent or context to be lost, no matter how well documented, so bringing engineering and design together early on is really important. In a previous team, I’d encourage people to leave their specialism at the door a little and just be ‘design system people’ so we could blur the lines more freely. That may also be a luxury not everyone has but in seeing that pathway from a problem to be solved through to something people out there actually use as the scope of the design system, it was a vehicle to have more joined-up conversations with less of a handover process.
There’s nothing inherently wrong with an immaculate design system that ends with a library of components and awesome documentation but that does come with it an overhead that is worth recognizing. The job doesn’t end there as spending time with engineering colleagues would help to get the best result, or taking on feedback from them and reworking some of your designs.
Seeing work on a live site that you’ve designed but feel hasn’t been implemented as well as you’d like can feel quite demoralizing so one of the questions is, how can you support people you might not directly work with? The scope of your design system might end with a library in a design tool so even at that stage, is it possible to communicate not just how it looks but a clear idea of what success looks like when it’s built?
Because of the differences in team structures or organizations, we work within, and whether the design system is ‘side of desk’ or has a well-supported team around it; this feels like one of the bigger hurdles to overcome. In that previous team of mine, we said that the live code was our source of truth because that’s what was used by people and so in that sense, ensured design as a specialism was involved all the way through the process. It may well be that you nominate your design library as the source of truth for good reason. It may also be that design tokens become the bridge between disciplines.
Design tokens can be incredibly powerful because they are storage for design decisions. Depending on how much you store there and whether comments or guidance around their use of them is present, they can form a part of the picture, however valuable they may be. If your tokens stored your core set of values (colors, typography, etc), then a more elaborate or detailed component will have many decisions that aren’t yet in token form so here we rely on documentation and communication.
So after my ramblings here, what I hope is that with your team of designers, developers, testing, product people…whatever form that might take, you:
A design without build is an aspiration but depending on how the environment is around us, that might be all we feel we can affect. The closer we’re able to bring worlds together the more powerful the design system, whether that’s through great documentation, design tokens, solid communication or actually bringing people together in some way. Design systems can be a fun and complex beast but ultimately in some way, they’re about people and how we can get good results for other people - our end users.
