News on Desktop
Role UX Lead
Company Google
Two years after launching Top Stories on mobile, we brought its AI-powered storylines, flexible content types, and organized layout to desktop. The redesigned grid handles 400M+ queries per day, surfaces more publishers, and helps users scan and understand the news faster.
Desktop Top Stories mixed articles from different storylines, buried context, and suffered low engagement beyond the first three results. The carousel UI also limited exploration and discoverability—most users never clicked past the first three results.
The Challenge
Users weren’t interacting much
Before redesign
Working iteratively, we compared two design candidates via A/B experiments and qualitative research, and landed on the version with stronger metrics and comprehension. It includes:
AI-powered storyline labels: Machine learning labels that provide context at a glance.
Dynamic grid layout: Surfaces more sources, creates clearer hierarchy, and adapts to a query’s news needs.
Scalable template: Extends the Search design framework for other teams to use in the future.
The Solution
A flexible, AI-powered grid
The MVP produced a net increase in overall clicks (rare for major Search redesigns as users resist change), more high-quality clicks, improved click distribution, and positive UXR feedback.
Designing for Search means balancing novelty, AI, and metrics at scale. Our solution has influenced Search long-term as teams began using asymmetrical grids and visually-forward layouts in their designs.
Launch and Impact 🚀
The new Top Stories on desktop
Following the Top Stories launch, subsequent Search launches featured aymmetrical grids and visually-forward layouts