Design systems, how do we make them most useful for designers?

over 5 years ago from , Senior Mobile UX - realestate.com.au

If anyone has worked at a large company, chances are there is a design system in place, colours, fonts, type scales and so on.

But as designers we're constantly creating new systems, our practice is creating systems, I'm interested in how we create not only a living system, but one that feels useful.

I've some questions for the community:

  1. What do you love/hate about using a design system?

  2. Has a design system made a finished product a failure?

  3. Have you ever seen anything in a design system that does not belong?

  4. How do you add to a design system?

Some examples of designs systems:






If you've got anything to add, or know of any great systems. I'm sure we would all love to hear about it.


  • Marc EdwardsMarc Edwards, over 5 years ago

    What do you love/hate about using a design system?

    Love the focus on consistency and easier long term maintainability. It is extra work for smaller projects, but ultimately less work for bigger projects.

    I think if you consider a design system simply a collection of reusable elements and a grid to align them to, that covers most of the uses and explains it well. It should be seen as a way to streamline the design process and keep everything consistent.

    Another advantage is that it can help ensure each “blessed” component can go through whatever testing is deemed necessary, which can help with performance, quality, accessability etc.

    I’m definitely a huge fan of sticker sheets containing reusable elements.

    6 points
    • Daniel Whyte, over 5 years ago (edited over 5 years ago )

      Ha, a sticker sheet is an awesome way to look at it. It's the investment up front that pays it forward, Although i'd like to avoid having too many stickers up front, and i'd also like to figure out how to get new stickers going forward.

      Thanks for the thoughts.

      0 points
      • Marc EdwardsMarc Edwards, over 5 years ago

        Ha, a sticker sheet is an awesome way to look at it.

        I definitely didn’t invent the idea or name, but I like the workflow and concept. Another advantage is that your sticker sheet can also be a slice sheet — a place where all the parts are exported. That means there’s one central truth for your assets.

        It's the investment up front that pays it forward, Although i'd like to avoid having too many stickers up front, and i'd also like to figure out how to get new stickers going forward.

        I treat them as similar to CSS. Elements that are unique to one page or screen should not be part of the sticker sheet or your CSS. Elements that are used in multiple places definitely should be.

        That means the best time to promote an element to your sticker sheet or CSS or design system is when you decide to use an element in more than one place. I also like flagging deprecated or stand-in assets.

        Thanks for the thoughts.

        No probs. :)

        1 point
  • Caitlin G, over 5 years ago

    Overall, I think design systems are a great idea - my company is in the process of putting one in place right now. I've heard some horror stories, though. One of our product managers used to work at a company that had implemented a design system while he was there, but didn't properly maintain it when it was "done." They ended up with over a dozen examples of data tables alone because one team would improve on the table example from the design system, another team would improve on the first team's example, and etc. There was no one in a position to decide to update the original pattern. It ended up taking way more time to use the design system than to not use it, so they gave it up.

    My company's entire product team only has 2 designers, working across 4 products. Implementing a design system will make the visual design aspect of our work much easier for us in the long run, assuming we have the time to spend maintaining it in the future. We plan to have a system in place for reviewing changes to the system, making decisions on new patterns to add and when to add them, and designing new patterns as they're needed. I'm not sure if we'll have time to make a good system out of it, though, given that there are only two of us. I hope we do.

    3 points
    • Daniel Whyte, over 5 years ago

      Sounds like you're in for a bit of a hard time, I recently read this article around how design systems could be treated https://medium.com/eightshapes-llc/a-design-system-isn-t-a-project-it-s-a-product-serving-products-74dcfffef935 I think it's about selling the fact that it is a product in itself.

      Interesting point you have around legacy in design by having a dozen data tables, Perhaps having a custodian or gatekeeper for the design system who's role was to make sure it was applied and updated would help that.

      I wish you luck with the creation of your design system, perhaps you could try add a small amount of the system to your product design delivery work?

      1 point
  • Emanuel S.Emanuel S., over 5 years ago

    Some awesome guys spoke about these design systems: http://spec.fm/podcasts/design-details/40840

    3 points
  • Simone Simone , over 5 years ago

    I think you should first evaluate IF and WHY you'd want to develop a design system for what you are working on. Scale and product lifespan will inform that decision. Then you'd you would probably be able to answer those questions yourself.

    1. Its a very subjective question. Most likely you'll hate a system which doesn't provide you with enough flexibility or sets up rules that are overly granular (e.g.: having a 4px baseline equals to having no baseline at all). The only systemt you'll love is the one you created yourself. But only until you realize its not perfect as you thought it was.

    2. No. A design system per sè provides structure and consistency across a product platform. Of course the design decisions you make while building the system will affect the product, but that would happen with our without a system in mind (e.g. shitty design is shitty design no matter wether its consistent / scalable or not).

    3. All the time. There's no perfect system. There's always an exception you have not thought about and when you have multiple designers working on it they'll interpret things differently. There's a balance between flexibility and rules you have to be able to master. And also a system has to be able to evolve as the product evolves.

    4. You can do it from the bottom up (redesigning individual components first, e.g. a new module for comments) or from the top down (e.g. define a new grid which affects to all components). A system is a set of variables, some global, some local: you decide which ones you want to manipulate in order to achieve what you need.

    0 points