Outcome
Faster results for the user
The new layout would bring the user to the results faster with less clicks (13 clicks for the revised design compared to 22 clicks in the current design), saving the customer 53% in time, giving them the opportunity to discover products.
Key Results:
-
Navigation time to product selection decreased by 50 percent
-
Average clicks required to refine a search dropped from 16 to 8
-
Search filter engagement increased by 35 percent
-
Conversion rate improved by 18 percent post-launch
-
Projected annual revenue impact of up to $300 million from a 0.5 percent conversion increase
Redesigning Search Filters
The Product Listing Page (PLP) is one of Zappos’ most trafficked customer touchpoints, playing a critical role in product discovery and online conversion. However, ineffective filtering led to user frustration, high bounce rates, and missed revenue opportunities.
I led a full redesign of the search filters to make the experience faster, clearer, and more intuitive. Our goal was to improve product findability, increase customer engagement, and drive measurable growth.
Project Overview (TL;DR)
Problem
-
Ineffective filters were leading to high bounce rates, user frustration, and missed revenue opportunities on the Product Listing Page (PLP).
Goal
-
Improve filter usability, reduce friction, and drive higher engagement and conversion.
Conversion Rate Results
-
53% faster navigation time (reduced from 22 clicks to 13)
-
35% increase in search engagement
-
18% lift in conversion rate after filter optimizations
My Role
Lead UX/UI Designer
Tools
Figma
Time
1 month
Deliverables
Search filters, Search recommendations
Stakeholders
Project Managers, Engineers, Tech Leadership, Marketing Leadership
Scope
MVP for filters for desktop, mobile, and iOS (not on android roadmap)
Methods
Competitive Analysis, Usertesting, Wireframe, Designing, Prototyping
Debbie, a busy mother of two and an aspiring runner, needed supportive shoes but struggled with the filtering experience. She faced unclear sizing options, overwhelming product results, and filters that reset with each refinement. Each misstep created frustration and slowed her purchase journey.
Debbie’s challenges represented a broader pattern across our customer base: time-sensitive users needing a faster, more reliable way to narrow results. Improving the filter experience meant not just helping Debbie, but empowering thousands of customers like her to find the right products quickly and confidently.
Customer Problem
Process
01
Filters are confusing and clunky
Users often felt uncertain about which filters applied to their goals, leading to confusion and abandoned searches.
"the filters are awful - I'm unable to keep filtering on type of shoes at some point and not to look at 10k results where maybe 10% are what would be returned if I could continue to filter ."
-Zappos Customer
02
Dynamic menu often confuses customers
Customers struggled to navigate the filter system and select multiple options efficiently.
"Why did you take away filters for styles...I am looking through 3k boots to find something I like. I will use a different vendor."
-Zappos Customer
Understanding the Problem
Usertest
I conducted extensive user testing through UserTesting.com across both desktop and mobile experiences. Our testing targeted two core groups: parents shopping for kids’ shoes and runners seeking performance footwear.
Research Goals:
-
Understand how users initiate searches (navigation versus search bar)
-
Observe how users interact with filters
-
Identify user behavior when a search fails
-
Analyze recovery patterns when users cannot find what they need
Results of Research
I received a lot of useful information on how customers use search and the filters. These were collected and determined to be main pain points to tackle on the design.
Key Findings:
-
42% of customers actively used filters during their shopping journey.
-
Users applying filters early in their search journey achieved better product matches.
-
Filters for kids’ sizes and shoe widths were major points of confusion, often causing users to restart their filtering process from the beginning.
-
Search bar users tended to enter overly detailed queries, leading to poor results when compared to navigation-first users.
Pain Points Identified:
-
Difficulty locating relevant filters
-
Overwhelming options for kids’ sizing
-
Lack of clarity in shoe width options
-
Filter persistence issues after adjustments
Design
Prioritizing Gender Selection to Improve Search Relevance
First Filter Users See
In the original design, the Gender filter appeared as the tenth option, even though users overwhelmingly prioritized it first when narrowing their results. I moved Gender to the top of the filter stack and made it a required early selection, improving the relevance of product matches and reducing customer backtracking.
Current Design
Proposed Design
Streamlining Kids’ Size Selection
Truncating Filters
Children’s footwear sizing presented an overwhelming challenge, with more than 45 options across infant, toddler, little kid, and big kid categories. To simplify the experience, I reorganized sizes into logical age groups, allowing customers to focus quickly on the most relevant options without needing prior knowledge of sizing nuances.
Current Design
Proposed Design
Clarifying Shoe Width Options
Truncating Filter
The original interface presented 22 different width options without clear explanations, leaving many customers confused about their selections. I reduced width selections to 6 clearly labeled categories, making the decision process faster and improving customer confidence during filtering.
Current Design
Proposed Design
Enhancing Search Bar Functionality with Smart Filtering
Give the power to the user while searching
The legacy search experience allowed the underlying page to remain visible while typing, which distracted users from completing their search. I proposed a redesigned search overlay that minimized background distractions and introduced keyword-based filtering directly into the search process. Users could now refine results by gender, size, or width while typing, aligning the search experience more closely with user intent and minimizing manual adjustments later.
Current Design
Proposed Design
Outcome & Reflection
Final Design
Brining it All Together
The final design incorporates the updated search bar and filter changes, addressing all the pain points identified during user testing. The result is a seamless experience with reduced friction, enabling users to easily find the products they need while also encouraging exploration.
Final screens for search
Reflection & Future Opportunities
Follow ups
While the new filtering experience significantly improved usability and conversion, further personalization opportunities emerged during research and testing.
-
Soft Save Persona
-
For users not logged into an account, we would save information from their first search session. If the user searches again during the session or returns to the site later, we would display their last-used filters, allowing them to quickly apply those preferences and align the results with their needs.
-
-
Hard Save Persona
-
For logged-in users, we would display their saved preferences, such as gender, shoe size, shoe width, and favorite brands. These personas could adapt based on purchase history, enabling users to switch between saved filters—for example, adults could have one set of saved filters for themselves, while their kids could have completely different filters saved.
-
-
AI assistance for 4+ keyword searches
-
Our research showed that users entering 2–3 keywords (17% of our customers) received accurate recommendations from our current algorithm. However, users entering 4 or more keywords (7% of our customers) often encountered mismatched results. Because the algorithm struggles to confidently interpret longer queries, we propose selectively applying AI assistance when users enter 4+ keywords. This approach would improve relevance without placing significant cost burdens on the company.
-