Overstock Search
Overstock Search
Overstock Search
Desktop Website
Role
Sole Designer
Overstock
Year
2019
The Result
✅ 4.67% increase in conversion.
✅ Improved usability through better filter design.
✅ Validated design decisions through rigorous user testing and A/B testing.
Overstock analytics showed that customers who engage with search filters are 2x more likely to convert.
The Problem
Overstock's existing search filter UI had a couple major flaws that I identified:
Only the top two filters were visible when the page loaded.
As users scrolled down, they often bypassed the filters and had to scroll back up to interact with them.
These usability issues led to friction in the shopping experience and missed opportunities for conversions.
Project Summary
My Role
As the sole UX Designer on this project, I:
Conducted a competitive audit to analyze industry best practices.
Presented the idea to the product manager and development lead.
Led user research & testing, gathering feedback from over 300 participants.
Performed data analysis on key user behavior metrics.
Creating wireframes, prototypes, and final designs in Figma.
Working with developers to ensure the final build matched design specifications.
Tools Used
Figma (Design & Prototyping)
UserZoom (User Testing & Analytics)
Design & Prototype
To validate our hypothesis, I built three interactive prototypes in Figma:
Control (Current Website) – Opened, scrollable filter sections.
Variant 1 – Filters partially opened with a "See More" button.
Variant 2 – All filter sections collapsed by default.
This allowed us to test in a controlled environment and eliminate the variable differences between a prototype and a live site.
User Testing
I conducted three rounds of unmoderated user tests with 100 participants in each round via UserZoom. Participants completed two tasks:
Applying multiple filters (e.g., "Beige, Wingback Chairs").
Selecting a brand filter from the bottom of the list.
Key Metrics Tracked:
Time on task
Number of clicks
User-rated ease of use
Findings:
The 'Default Collapsed' version was the clear winner across all metrics except number of clicks.
Users found this version easier to use despite requiring an extra click.
Findings
The 'Default Collapsed' version was the clear winner across all metrics except number of clicks.
Users found this version easier to use despite requiring an extra click.
Final Implementation:
After validating our approach, we built all three variants and ran an A/B/C test on the live website. The 'Default Collapsed' version emerged as the best performer, improving conversion rates by 4.67%.
Key Learnings
Collapsing filters by default simplified the UI and improved visibility without requiring excessive scrolling.
While this version increased the number of clicks, the addition of a sticky filter section ensured users could easily engage with filters at any point in their shopping journey.
The improved design provided a frictionless filtering experience, leading to higher engagement and a meaningful lift in conversions.
Key Results & Outcomes
✅ 4.67% increase in converting visitors.
✅ Improved usability through better filter discoverability and accessibility.
✅ Validated design decisions through rigorous user testing and A/B testing.
Defaulting to collapsed filters lead to 4.67% lift in conversion.
This project demonstrated the impact of thoughtful UX design on business metrics and reinforced the value of data-driven decision-making in e-commerce optimization.
