Visual Search with Suggestions
THE PROBLEM
Basic text based searches returned too broad of results that didn’t help users navigate to their products with haste, increasing friction and slowing purchasing decisions.
MY ROLE
I redesigned the search typeahead to include product thumbnail images, category labels, and top item results to give users immediate visual confirmation before committing to a full search results page. I worked with and directed engineering, junior designers, and the Bloomreach AI to define the new experience from initiation through final QA.
Resulting Improved KPIs
36%
Increase in search revenue metrics
25%
Growth in average cart size
10%
Increase in click-through rate
Design Process
All projects were led and executed by me in close collaboration with product managers, IT depts, marketing content creation, customer service, and analytics.
What I did:
- Directed methods of surfacing child items in searches when product parent families are the sole entities CMS recognize
- Analyzed behavioral data in Google Analytics, Qlik, and Glassbox to map users’ behavior in the search-to-purchase funnel
- Audited competitor and best-in-class B2B e-commerce sites to benchmark expectations
Constraints
- Integrated within the existing global header and search infrastructure
- Required support for 40,000+ SKUs with complex specifications and variants
- Limited development bandwidth required focused, high-impact improvements.
Goals
- Surface the most relevant and purchased products
- Enable direct add-to-cart functionality from search results
- Increase conversion efficiency within the search experience
- Improve clarity for search results
- Reduce cognitive overload in high-level searches
Results
The visual search redesign achieved all target goals by surfacing relevant products and the ability to add them to cart and minimized unnecessary page transitions.
The updated search experience contributed to a 10% increase in search-driven revenue, demonstrating the impact of improved clarity.