dubizzle
2026
AI-Powered Search Experience
Redesigned the exisiting search with AI-driven interactions and animation to create a more intuitive and engaging discovery experience.
Overview
Introduced a new AI search experience within dubizzle’s existing search interface.
The experience included a transition between traditional and AI search, a distinct visual treatment for the AI state using glow and frosted-glass effects, and animated suggestions designed to give users ideas of what they could search for.
The Impact
Significant improvement in user engagement and session time.
Problem
Search often required users to do too much of the work.
Traditional keyword search relied on users knowing what to type and manually applying the right filters to narrow down results. More complex or conversational queries could also fail to return relevant matches, sometimes leading users to a no-results page rather than helping them continue their search.
01
High reliance on filters
Users often had to manually apply multiple filters to reach relevant results
02
Limited understanding of complex queries
Traditional keyword search struggled when users searched using more descriptive or conversational language.
03
Dead-end experiences
Queries that couldn't be interpreted could lead to no results, interrupting discovery instead of helping users refine their intent.
Ideation
Exploring of new ways to search:
Text, voice, image and more
What we learned
Good AI search isn't just about understanding more, it needs to make searching feel easier.
The opportunity was to move some of the effort away from the user. Instead of expecting people to translate what they wanted into the right keywords and filters, AI could interpret more natural, descriptive queries and help users get closer to relevant results.
We also recognised that introducing AI created a new interaction challenge: users needed to understand that they could search differently. The interface therefore needed to demonstrate the possibilities of AI search through examples and prompts rather than expecting users to discover them on their own.
Designs
Making AI search discoverable through interaction
The final experience introduced a toggle between traditional and AI search, with a visual transition into a distinct AI state using glow and frosted-glass effects.
Animated search suggestions rotated through examples of what users could search for, helping expose the possibilities of the AI-powered experience through the interface itself.
The interaction work included:
Traditional → AI transition
switching between the two search experiences.
Prompt inspiration
animated suggestions showing examples of searches.
AI visual language
glow and frosted/glass treatments distinguishing the AI state.
Lessons
Designing for AI meant designing for a new behaviour, not just a new feature.
One of the key takeaways was that adding AI capability alone wasn't enough. The experience also needed to build users' understanding of what the technology could do. Using prompt suggestions, motion and a distinct AI state became part of communicating that shift—helping users move from thinking in keywords and filters toward expressing what they were actually looking for.
Key outcomes
Shifting the pattern
From “Which filters should I select?” → to “What am I actually looking for?”
