One team would be responsible with implementation, aggregation, and adoption. We planned to start with mandatory components, then add to the system on an ad hoc basis.
Switching to Figma
I was splitting my time between Sketch, InVision, and Abstract for prototyping and versioning and found it time-consuming to monitor 3 different tools. I made the switch to Figma as an all-in-one tool to not only support my workflow, but the structure of the design system.
Audit and roadmap
I took screenshots of elements from SketchyMedical, SketchyNursing, SketchyLaw, and SketchyUndergrad and created folders to categorize them. The audit showed concrete examples of where our developers created different instances of the same component, and where design could've been more unified.
Alongside a brief, I outlined the needs, stakeholders, feature list, requirements, and estimated timeline for the design system. We prioritized the MVP and broke the feature list into manageable chunks for our product roadmap, aiming to release around 3 months.
Defining the abstract
I created a survey that would collect existing beliefs on what our company mission and values were. I then uploaded their responses to SurveyMonkey, analyzed and presented the results to our co-founders so that we could finalize our thoughts.
I chose to collect handwritten responses about our company mission and values because it felt more brainstorm-like and participatory.
After suggested a few architectures to our front-end engineers, we decided to move forward with BEM, "Block Element Modifier". We were most familiar with its framework and agreed it was the best choice because it offered modularity, legibility, and massive scope.
While the design system was being built, I used Notion as the project's reference and handoff tool for content. It served as an easily accessible source of truth that rendered snippets of code and showed interactive examples of components.
Components and mockups lived in Figma to help engineers visualize the end product and further inspect CSS. I also familiarized myself with project management and created tickets in Jira detailing functional requirements for development hand-off.
I considered GitBook and Pattern Lab in initial conversations about hosting. We didn't have the manpower to set up Storybook in the time desired, so we decided to copy a repo and internally host the system as a proof of concept before migrating to Storybook.