As part of an interview process, I conducted an independent accessibility audit of the Agoda web and mobile experience. Rather than treating this as a theoretical exercise, I approached it as a real product scenario, focusing on how accessibility improvements could be applied within an existing system without disrupting business or design constraints.

My goal was to demonstrate not only the ability to identify issues, but also to translate them into practical, system-level solutions and communicate them clearly in real time.

I structured the work around three key layers: analysis, user perspective, and rapid prototyping.

First, I conducted a WCAG-informed visual audit, reviewing color contrast, typography, icon clarity, spacing, and touch targets. I also paid special attention to dynamic type behavior, evaluating how the interface responds to font scaling and whether readability is preserved across different settings.

To move beyond guidelines, I simulated how the interface is perceived by users with visual impairments. This included protanopia (color blindness) and low vision scenarios, which helped reveal issues that are often overlooked in standard design reviews. This step was critical in shifting the analysis from compliance to real user experience.

To make the discussion actionable, I recreated parts of the interface in Figma and built a lightweight system including color tokens, typography rules, and key UI components. During the interview, I demonstrated how the interface could be improved in real time by introducing better contrast relationships, scalable typography, and clearer interaction patterns.

The audit revealed several systemic issues that impact accessibility and usability.

One of the main problems was that the interface relied heavily on color differences rather than contrast, making key elements indistinguishable for users with color blindness. In several cases, red, green, and gray values had similar contrast levels, which flattened visual hierarchy and made important information harder to perceive.

Another issue was related to ambiguous actions. Different buttons shared the same visual style despite representing different functions, which forced users to rely on reading instead of scanning. This creates friction, especially for users with impaired vision or cognitive load.

The language selection pattern also lacked clarity. It relied on flags and subtle visual cues that become difficult to interpret under color distortion or low vision conditions, making it harder for users to confidently understand their current selection.

Additionally, I identified limitations in typography and scaling behavior. Font sizes, spacing, and hierarchy were not fully optimized for readability, and the interface did not fully support dynamic type adjustments in a way that preserves usability.

Rather than proposing a full redesign, I focused on incremental, scalable improvements that could realistically be implemented.

I suggested shifting the design approach from color-based differentiation to contrast-driven hierarchy, ensuring that elements remain distinguishable regardless of color perception. I also introduced redundant visual cues, such as combining icons with labels and strengthening selection states, to reduce reliance on color alone.

To improve readability, I proposed implementing dynamic typography, allowing the interface to adapt to system font settings without breaking layout or structure.

I also demonstrated how introducing a dark mode as an additional layer could improve accessibility for users with low vision, while remaining compatible with the existing design system.

Finally, I recommended increasing spacing and touch target sizes, improving both accessibility and overall usability.

The case was well received and I progressed to the final stage of the interview, demonstrating my ability to quickly analyze a complex product and translate accessibility insights into actionable, system-level improvements.

Although I was not selected, the project reinforced my belief that accessibility is a product opportunity, not just a compliance requirement. The most effective changes were small, thoughtful adjustments to contrast, hierarchy, and interaction patterns, which significantly improve usability for a wide range of users.

This work reflects my approach to design: balancing user needs, system thinking, and practical implementation, with a focus on solutions that scale.