Research

Owned by Adam Butterworth (Deactivated)
Last updated: Jun 15, 2021

[TODO] Add core metrics and measures of success

Stakeholder Interviews – Fall 2020

Half hour interviews conducted with six internal staff including: marketing manager, product manager, engineer, engineering manager, accessibility specialist, and ux designer.

Key Takeaways

1. What aspects of the current workflow do you think could be improved?

  • A more methodical approach to meeting accessibility standards

  • QA is hard to do. We have an undependable staging environment and lack sandboxes for UI testing.

  • Engineers working transactionally with designers, and not having direct engagement with the design team in general.

2. If you had a magic wand and could change one thing about how things work here, what would that thing be?

  • Dedicated engineering time to build accessibility tooling or features.

  • Clearer tie between roadmap and product strategy.

  • Would abandon Bootstrap and use CSS and JS instead to simplify and change the tech stacks related to Paragon. Bootstrap requires one to have a big monolithic stylesheet that has a lot of subtleties that need to be worked out and is confusing tech. CSS and JS might be a better model.

  • Stakeholder management is murky and burdensome

3. How do you hope the design system will help?

  • Reduce the number of new a11y bugs to less than a quarter of what we generate today

  • New bugs should only happen where there is new innovation

  • Produce software that is optimized for everyone with more sophisticated accessibility features

  • Easy to know what components exist, making it easier to design things based on available patterns. (I would like section of latest updates to the system that I could regularly reference)

  • Let Design lead Engineering to do things right; to have a unified voice, theme and user mode of interaction

  • Less back and forth between functional areas (design, product, engineering)

4. What about the design system makes you nervous, apprehensive, or skeptical?

  • Worried that we'll create yet another pattern on top of everything else and fail to unify our interfaces.

  • Design System takes a lot of maintenance - worried about our ability to keep the Design System fresh and evergreen

  • Divestment because of “shiny object” distraction

5. What does success look like?

  • Our design system and the products we make with it address the full range of human experience, not just named disabilities.

  • Going to edx.org, HubSpot, or WordPress, or the LMS and seeing consistent look and feel

  • Paragon working group remains strong and healthy, and fully attended by both Engineering and Design. Having two frontend-oriented engineers dedicate a good chunk of time consistently, and prioritize important problems.

  • Design system verbal language is common language at edX (component titles, what they do, how they’re used)

6. What do you think a design system should not do, or not attempt to do?

  • Should not stymie innovation, though shouldn't encourage reinventing things for no reason +2

  • Should not be all things to all people (at first). UI designers should be the the first audience.

  • Should not try to be perfect before launching. 80% is good to launch, then iterate after that

7. What’s going to sink this ship? Fast forward a year from now and the design system initiative is a total failure. Why did it fail?

  • Squads decide it's faster to build things on their own than with the design system.

  • Lack of cross-functional agreement and prioritization to follow through.

  • Lack of investment

8. What questions should we be asking you that we aren’t?

  • If there are components with no accessible versions available, what do we do?

  • How will we plan and time engineering effort needed to bring this system to life?

  • How can I help drum up the support?

  • What is it going to take for some teams to realize the full benefit and be totally bought in?

  • What is a design system?