Seizure Safe Profile
Reduces risks for users prone to seizures by eliminating flashing lights, rapid animations, and certain color contrasts.
01 / Overview
Role — original study
UX Researcher
Responsibilities
User research, UX/UI design, usability testing, interaction design
Timeline
14 weeks
Collaborators
2 designers
How might we make the NPS website more accessible and inclusive?
NPS supports about 318 million visitors every year. Under the Americans with Disabilities Act (ADA), any web product or service needs to be fully accessible.
Upon talking to a few users with special needs, Aditya found their pain points and concluded that the NPS website was not A11Y-friendly. This project aims to come up with A11Y recommendations.
02 / The solution — four accessibility profiles
Try them on this page — they actually work here:
Vision Impaired Profile
Makes websites readable by screen readers and offers visual adjustments for those with vision impairments, simplifying navigation through keyboard shortcuts and enhancing readability with contrast and font changes.
ADHD Friendly Profile
Minimizes distractions with a clean layout, reduces motion, and highlights essential information to aid users with ADHD in focusing on key content.
Cognitive Disability Profile
Designs for ease of use with predictable navigation, simple language, and supportive visuals to help users with cognitive impairments understand and interact with digital content more easily.
03 / Investigating the problem
To address the objective of enhancing the accessibility and inclusivity of the NPS website, the investigation involved:
- Heuristic evaluation
- Conducting user surveys
- Conducting an accessibility audit of the NPS website under Web Content Accessibility Guidelines (WCAG) 2.1
The website remained non-compliant with the accessibility standards, with a score of 76% — highlighting crucial areas needing improvement to achieve full accessibility for all users.
User research
Research and design, from initial concept to final execution, organized the results into a few categories: heuristic evaluations, user surveys, user interviews, prototypes, and usability test results.
Heuristic evaluation
Jakob Nielsen's 10 usability guidelines were examined, with issues measured on Molich and Jeffries' severity scale — minor, serious, and catastrophic. This revealed areas for improvement, particularly in enhancing the UI and alternative text for images, and ensuring all functionality is accessible via keyboard navigation for screen readers.
Surveys & results
Surveys collected feedback from a wide array of users, including those with special needs, focusing on challenges and desired improvements in website navigation, content clarity, and overall satisfaction. Responses indicated a critical need for enhanced accessibility options for users with visual and cognitive disabilities, along with improvements to the user interface for a better visual experience.
Usability test & results
After thorough user research, a usability test plan led to testing 10 specific tasks with 5 participants in a remote, moderated setup. The outcomes revealed a range of serious issues, from significant obstacles to major problems. Navigating the media gallery section stood out as particularly challenging, and overall, tasks took longer to complete than expected.
04 / Scope for redesign
- Accessibility enhancements
- Navigation and readability improvements
- Content clarity and simplification
- Enhance UI to make users connected
Designs
Prototypes were developed to address issues identified in earlier evaluations, then pushed for a round of testing with 5 participants, with in-depth positive outcomes. The average time to finish a task reduced to 5.1 seconds, with an increase of 80% efficiency (previous: 10 seconds).
05 / User impact
76% → 92%Accessibility score, post-redesign
−15%Bounce rate
+30%User satisfaction
+20%Engagement — avg. time on site
Post-redesign, the website's accessibility score increased from 76% to 92%. The bounce rate decreased by 15%. User satisfaction surveys post-redesign showed a 30% increase in positive feedback. Metrics revealed a 20% increase in the average time spent on the site per visit.
06 / How I'd engineer it — Akarsh's take
A good accessibility profile is not a theme toggle — it is a user preference that has to survive navigation, work on every page, and never regress. This is how I would build it:
- Profiles as persisted preferences: store the chosen profile server-side / in local storage, apply it before first paint, and respect
prefers-reduced-motionandprefers-contrastby default. - One design-token system: contrast, font size, spacing and motion driven by tokens, so a profile changes the whole site — not a handful of pages.
- Keyboard-first, always: visible focus, logical tab order, skip links, and full screen-reader coverage — tested, not assumed.
- Accessibility in CI: automated WCAG / axe checks on every pull request, so a new feature cannot ship a 76% regression.
- Built for scale: the same patterns I use in production — .NET / Azure, feature flags for a gradual rollout, and telemetry on profile usage and task completion, so the 76% → 92% claim is measured, not guessed.
07 / Key takeaways
- User feedback is invaluable: engaging with real users, especially those with disabilities, provided insights no automated tool could. Their feedback was crucial in identifying areas for improvement and validating the redesign.
- Continuous improvement is essential: continuous monitoring and iteration based on user feedback and evolving standards are crucial to maintaining and improving accessibility.
- Principle of inclusivity: this project set a precedent for future work, emphasizing that at the heart of great design is the principle of inclusivity, ensuring digital experiences are accessible and enjoyable for everyone.
- Accessibility enhances overall user experience: the improvements that help users with disabilities — clearer navigation, simpler language, less motion — make the site better for everyone.
Source: Aditya's NPS UX case study, adtiyaux.com/nps. Figures quoted from that study.