Jordan Love

The goal

Integrate a new AI-powered search while rebuilding the experience with our new design system to improve accessibility, consistency and usability.

The aim was to make search smarter and simpler - powered by AI relevance, guided by design system principles and validated through eCommerce best practices.

The problem

The approach

Researched Baymard Institute’s eCommerce search best practices to guide UX decisions around auto-suggest, query handling and empty states.

Collaborated with stakeholders & engineers to align design constraints with the new AI model (semantic intent, typo-handling, synonyms).

Rebuilt layouts and patterns using design system components to ensure scalability and accessibility (WCAG 2.1 AA)

Introduced skeleton loading patterns for perceived performance and smoother user feedback.


The solution

Header

Rebuilt the header using design system components to optimise space utilisation, improve visual hierarchy and ensure accessibility across devices

solutionHeader.png

Auto-suggest

AI-driven predictive suggestions with product cards, trending & recent searches, and category/page shortcuts.

solutionAutosuggest.png