Data Vizualisation Accessibility Components for Coinbase Design System
GOAL: Understand the behaviors and pain points of those who use assistive technology for data visualization to inform product design decisions
SITUATION: Cross-functional collaboration with other researchers, product designers, and engineers after an accessibility audit to improve accessibility components of Coinbase Design System
IMPACT: Our 25+ insight-based recommendations addressed multiple severity 1 issues identified in the audit and were added to Coinbase Accessibility main page
Company: Coinbase
Methods: Industry & literature review (best practices), Interviews
Role: User Researcher (intern)
Team: 3 Researchers, Accessibility Lead
Duration: 6 weeks
Summary
BACKGROUND AND PROBLEM
As a trading platform, Coinbase relies heavily on real-time statistics and data visualizations to convey complicated information while enabling users to complete small, but consequential, actions. Coinbase has over 100 million verified users. However, not all of these users experience the visualizations in the same way. Approximately 23% of the global population has a visual impairment, some of whom require assistive technologies (ATs) to engage with the online world.
ROLE AND CONTRIBUTION
I was one of four user researcher interns working on this project. Having the most product experience, I took the lead on analyzing the insights keeping in mind that we wanted them to be action-oriented and mapped to specific product strategies. I also articulated our recommendations in the final Coinbase Design System documentation.
Research Question
How do users of assistive technology (AT) best interpret, navigate, and interact with data visualizations?
Research goals:
- Understand what an accessible user experience in the context of data visualization is
- Identify major pain points and workarounds that AT users have when interacting with data visualizations
- Establish which data visualizations are most appropriate for specific ATs, and how they can be adapted to be more appropriate
PROCESS
Phase 1 (Foundational Research)
In order to conduct effective and empathetic interviews, we wanted to build our foundational knowledge in two main areas: understanding ATs and what accessibility means in the Coinbase context.
Understanding Assistive Technology
Literature review, expert interviews (e.g., the accessibility lead at Coinbase), conversations with users of ATs through the Coinbase network
- What are the different kinds of AT?
- What does each entail?
- How does each influence the user’s experience?
Accessibility in the Coinbase context
Internal stakeholder interviews with the Coinbase Staff Accessibility Lead, Staff User Researcher who’d conducted initial Coinbase accessibility research, and Coinbase Design System designers
- What does an accessible experience mean for Coinbase?
- What are the Coinbase’s urgent and/or important needs in this context?
- Are there limitations Coinbase has that we should know?
- How can we ensure our product recommendations are as helpful and actionable as possible for the team? (i.e., supporting information like examples, their structure, and format of presentation)
Phase 2 (Interviews)
We leveraged all the information we’d collected during Phase 1 to build our in-depth interview discussion guides, anchored to our research questions and goals. To conduct our interviews, we used Fable: a platform built to connect to people with disabilities for user research and accessibility testing.
Challenge: Given the lack of experience we had with ATs, it wasn’t easy to visualize the experiences the participants were describing. We addressed this by requesting the participants to share their screens and show us while talking through their experiences. This allowed crucial contextual nuances, like what made them hesitate or zoom in, to emerge and provided us with tangible examples (and video clips!) to present with our final recommendations.
Phase 3 (Data Analysis and Synthesis)
We built an affinity map to identify and extract themes and insights from our interviews. Leveraging tags, we incorporated tone and user sentiment (e.g., positive or negative experience) into our insight sticky notes. The themes reflected our goal of informing product and design decisions – e.g., displaying evolving information, receiving feedback on system change of state, and types or format of data visualizations.
INSIGHTS AND RECOMMENDATIONS
We had five key insights with multiple product and design recommendations addressing different aspects of each.
Presenting the Recommendations
We were lucky to have multiple opportunities to present our findings and recommendations to the research team and the entire Coinbase Design System team (researchers, product designers, illustrators, product managers, and engineers).
🚀 Learning: Our lack of context made it tricky to relate to the experiences of our participants and we knew our audience would experience something similar. Therefore, we incorporated video clips of users showing what they go through to support our insights and final recommendations. This enabled our audience to share the participants’ experiences and highlighted the importance of implementing our recommendations.
IMPACT
Our work gave the crucial task of making the Coinbase experience more inclusive to users of assistive technology the eyes it deserved.
- In the short term, our recommendations addressed 8 severity 1 issues that the accessibility audit of Coinbase’s products identified. Our recommendations were also added to the Coinbase Design System accessibility main page – putting it front and center for all data visualizations going forward.
- In the long term, we’re hoping that this sets the ball rolling to make Coinbase, and other visualization-heavy digital experiences, more mindful of and cater to its assistive technology users.
REFLECTION
What went well
- Navigating sensitive issues. Interviewing the participants was initially nerve wracking. However, by making the extra effort to educate ourselves and speak to experts in the field, the team and I were able to probe deeper during interviews while being mindful and empathetic towards our users so they never felt uncomfortable or offended.
Next time…
- Performing a task analysis. In the future, or if we’d had more time, it would be great to perform a task analysis. Although users shared their screens with us to walk us through certain examples, providing more structure to it would have been helpful in validating our recommendations!