Name
eBay
My Role
UX Researcher
Timeline
March 2025 - May 2025
Tools
Figma, FigJam, Google Forms, Silktide
Past Studies
Relevant Research
KEY LEARNINGS
One of the most valuable lessons from this project was the importance of conducting in-depth user research to uncover trust and usability concerns in AI-assisted navigation. By studying how users interact with eBay’s platform, I learned that clarity, transparency, and efficiency are critical for building confidence. This experience reinforced my understanding that successful navigation design is not just about functionality, but about creating an intuitive and trustworthy user journey.
Reflection
Journey Maps
During testing with a non-eBay user, we mapped out their goals, expectations, and challenges while navigating the platform. At first, the user felt that their expectations were met by the wide range of products displayed on eBay’s main menu, which gave them confidence in the platform’s offerings. However, as they moved deeper into the experience, frustration set in due to poor UI and confusing navigation. A lack of transparency throughout the process also created a sense of wariness, highlighting key areas for improvement in the overall user journey.
Usability Testing
We conducted in-person usability testing with five participants, all between 18–21 years old, to evaluate how younger users interacted with the platform and to identify friction points in the experience. The study included both interview questions and guided tasks. Participants were asked about their past experiences with similar platforms, their online shopping habits, and their first impressions when navigating e-commerce apps. We also explored their expectations, what the platform signaled it could help them accomplish, and whether any surprises emerged during task completion.
For the task portion, participants were first asked to visit ebay.com and spend up to five minutes scrolling through the home screen while practicing the Thinking Out Loud Protocol. They then searched for an affordable (under $1000) used LG TV in good condition verified by eBay, and repeated the same task on the mobile application to compare desktop and mobile navigation. Additional tasks focused on applying filters and using the Brand Outlets feature to complete specific goals, while we observed ease of navigation and overall perceptions of the flow.
eBay Survey
To complement usability testing, we also conducted a survey to capture broader insights into user habits and preferences. The survey focused on three main areas: shopping habits and preferences, common pain points while shopping, and navigation behaviors.
Participants were asked questions such as whether they prefer shopping on their phone or laptop, which sites or apps they typically use, and which tools they would rely on to navigate eBay. We also gathered feedback on how they would rate their overall shopping and navigation experience on eBay, helping us identify opportunities to improve usability and align with user expectations.
Final Study
E.B. Concept Testing Objectives
Prototype Introduction: Test E.B., an AI chatbot prototype for eBay
Research Goals: Evaluate how AI could be integrated into the platform
Risks & Issues: Understand user trust and concerns around AI usage
E.B. Concept: Competitor Landscape
In evaluating E.B., we also looked at competitors leveraging AI in the e-commerce space. Meta’s Facebook Marketplace integrates shopping within a social platform, creating a peer-to-peer marketplace experience. Amazon, through AWS AI, uses advanced artificial intelligence to power personalization, recommendations, and shopping efficiency. Shopify Magic brings AI assistance directly into Shopify’s ecosystem, helping both merchants and buyers streamline their processes. These comparisons provided context for how E.B. could differentiate itself within the market.
E.B. Concept: AI Navigation Prototype
E.B. is an AI-powered navigation tool designed to assist users with filters and search results, creating a faster and more intuitive shopping experience.
Provides easy and fast search results
Takes in user commands (e.g., price range, item, shipping preferences)
Attaches links and images to match product requests
Automatically applies relevant filters
Delivers follow-up responses and personalized recommendations
Findings
Security and Trustworthiness
Remembering It’s Just A Computer
Users are cautious of information from the bot and get frustrated if it doesn’t understand their needs on the first try.
Withholding Navigation Results
Some worry the bot might withhold navigation results or push certain items, raising fears of being “scammed.”
AI Trust With Sensitive Information
While hesitant with sensitive data, users are open to letting the bot handle practical tasks like tracking shipping when it feels useful.
AI Integration
Simple Commands
Users prefer short, direct prompts (e.g., “pokemon card”) instead of holding full conversations with the AI. This keeps interactions quick and to the point.
Image Search
Participants valued when the AI asked clarifying questions, since it made results feel more accurate and personalized.
Quick Text Responses
Many expressed interest in uploading photos for product searches, seeing it as a faster and more reliable way to find specific items.
Recommendations
Security and Trustworthiness
Transparency Disclaimer
Users are cautious of information from the bot and get frustrated if it doesn’t understand their needs on the first try.
Control of Search Results
Some worry the bot might withhold navigation results or push certain items, raising fears of being “scammed.”
Personal Info Security
While hesitant with sensitive data, users are open to letting the bot handle practical tasks like tracking shipping when it feels useful.
AI Integration
Support Short Prompts
Short prompts work best, aided by auto-complete suggestions (e.g., “pokemon card 1st edition”).
Seamless Image Search Integration
Seamless integration is key, with the ability to upload photos, combine text + image (e.g., “Find this jacket in blue”), and get visual similarity results. Camera access for in-app scanning was also desired.
Optimize for short concise responses
Users want short, direct responses with key details (price, condition, shipping) surfaced first and the option to expand for more.
Quick action buttons
One-click buttons for common tasks (e.g., “Track order,” “Start return”) streamline the experience.
Smart follow-up questions
Users appreciate smart, contextual clarifying questions when searches are broad, but want follow-ups kept to a minimum.
Overwhelming Navigation
PROBLEM
eBay’s massive marketplace makes it difficult for users to quickly find what they need. Cluttered results and unclear pathways often lead to frustration, repeat searches, or missed opportunities.
Smarter, Streamlined Navigation
SOLUTION
Research uncovered opportunities for AI to simplify the experience by surfacing key details first, supporting short prompts, enabling image + text search, and adding quick action buttons. These improvements create a faster, more intuitive, and trustworthy shopping journey.
EBAY’S NAVIGATION
This project focuses on improving eBay’s navigation through in-depth user research. The goal was to understand where users struggled in finding and purchasing items, then reimagine how navigation could feel clearer, faster, and more trustworthy. The design explores AI-assisted search features like image + text input, auto-complete prompts, and quick action buttons, creating a streamlined experience that reduces friction and builds user confidence.
Past Studies
Relevant Research
KEY LEARNINGS
One of the most valuable lessons from this project was the importance of conducting in-depth user research to uncover trust and usability concerns in AI-assisted navigation. By studying how users interact with eBay’s platform, I learned that clarity, transparency, and efficiency are critical for building confidence. This experience reinforced my understanding that successful navigation design is not just about functionality, but about creating an intuitive and trustworthy user journey.
Reflection
Journey Maps
During testing with a non-eBay user, we mapped out their goals, expectations, and challenges while navigating the platform. At first, the user felt that their expectations were met by the wide range of products displayed on eBay’s main menu, which gave them confidence in the platform’s offerings. However, as they moved deeper into the experience, frustration set in due to poor UI and confusing navigation. A lack of transparency throughout the process also created a sense of wariness, highlighting key areas for improvement in the overall user journey.
Usability Testing
We conducted in-person usability testing with five participants, all between 18–21 years old, to evaluate how younger users interacted with the platform and to identify friction points in the experience. The study included both interview questions and guided tasks. Participants were asked about their past experiences with similar platforms, their online shopping habits, and their first impressions when navigating e-commerce apps. We also explored their expectations, what the platform signaled it could help them accomplish, and whether any surprises emerged during task completion.
For the task portion, participants were first asked to visit ebay.com and spend up to five minutes scrolling through the home screen while practicing the Thinking Out Loud Protocol. They then searched for an affordable (under $1000) used LG TV in good condition verified by eBay, and repeated the same task on the mobile application to compare desktop and mobile navigation. Additional tasks focused on applying filters and using the Brand Outlets feature to complete specific goals, while we observed ease of navigation and overall perceptions of the flow.
eBay Survey
To complement usability testing, we also conducted a survey to capture broader insights into user habits and preferences. The survey focused on three main areas: shopping habits and preferences, common pain points while shopping, and navigation behaviors.
Participants were asked questions such as whether they prefer shopping on their phone or laptop, which sites or apps they typically use, and which tools they would rely on to navigate eBay. We also gathered feedback on how they would rate their overall shopping and navigation experience on eBay, helping us identify opportunities to improve usability and align with user expectations.
Final Study
E.B. Concept Testing Objectives
Prototype Introduction: Test E.B., an AI chatbot prototype for eBay
Research Goals: Evaluate how AI could be integrated into the platform
Risks & Issues: Understand user trust and concerns around AI usage
E.B. Concept: Competitor Landscape
In evaluating E.B., we also looked at competitors leveraging AI in the e-commerce space. Meta’s Facebook Marketplace integrates shopping within a social platform, creating a peer-to-peer marketplace experience. Amazon, through AWS AI, uses advanced artificial intelligence to power personalization, recommendations, and shopping efficiency. Shopify Magic brings AI assistance directly into Shopify’s ecosystem, helping both merchants and buyers streamline their processes. These comparisons provided context for how E.B. could differentiate itself within the market.
E.B. Concept: AI Navigation Prototype
E.B. is an AI-powered navigation tool designed to assist users with filters and search results, creating a faster and more intuitive shopping experience.
Provides easy and fast search results
Takes in user commands (e.g., price range, item, shipping preferences)
Attaches links and images to match product requests
Automatically applies relevant filters
Delivers follow-up responses and personalized recommendations
Findings
Security and Trustworthiness
Remembering It’s Just A Computer
Users are cautious of information from the bot and get frustrated if it doesn’t understand their needs on the first try.
Withholding Navigation Results
Some worry the bot might withhold navigation results or push certain items, raising fears of being “scammed.”
AI Trust With Sensitive Information
While hesitant with sensitive data, users are open to letting the bot handle practical tasks like tracking shipping when it feels useful.
AI Integration
Simple Commands
Users prefer short, direct prompts (e.g., “pokemon card”) instead of holding full conversations with the AI. This keeps interactions quick and to the point.
Image Search
Participants valued when the AI asked clarifying questions, since it made results feel more accurate and personalized.
Quick Text Responses
Many expressed interest in uploading photos for product searches, seeing it as a faster and more reliable way to find specific items.
Recommendations
Security and Trustworthiness
Transparency Disclaimer
Users are cautious of information from the bot and get frustrated if it doesn’t understand their needs on the first try.
Control of Search Results
Some worry the bot might withhold navigation results or push certain items, raising fears of being “scammed.”
Personal Info Security
While hesitant with sensitive data, users are open to letting the bot handle practical tasks like tracking shipping when it feels useful.
AI Integration
Support Short Prompts
Short prompts work best, aided by auto-complete suggestions (e.g., “pokemon card 1st edition”).
Seamless Image Search Integration
Seamless integration is key, with the ability to upload photos, combine text + image (e.g., “Find this jacket in blue”), and get visual similarity results. Camera access for in-app scanning was also desired.
Optimize for short concise responses
Users want short, direct responses with key details (price, condition, shipping) surfaced first and the option to expand for more.
Quick action buttons
One-click buttons for common tasks (e.g., “Track order,” “Start return”) streamline the experience.
Smart follow-up questions
Users appreciate smart, contextual clarifying questions when searches are broad, but want follow-ups kept to a minimum.
EBAY’S NAVIGATION
This project focuses on improving eBay’s navigation through in-depth user research. The goal was to understand where users struggled in finding and purchasing items, then reimagine how navigation could feel clearer, faster, and more trustworthy. The design explores AI-assisted search features like image + text input, auto-complete prompts, and quick action buttons, creating a streamlined experience that reduces friction and builds user confidence.
Name
eBay
My Role
UX Researcher
Timeline
March 2025 - May 2025
Tools
Figma, FigJam, Google Forms, Silktide
Overwhelming Navigation
PROBLEM
eBay’s massive marketplace makes it difficult for users to quickly find what they need. Cluttered results and unclear pathways often lead to frustration, repeat searches, or missed opportunities.
Smarter, Streamlined Navigation
SOLUTION
Research uncovered opportunities for AI to simplify the experience by surfacing key details first, supporting short prompts, enabling image + text search, and adding quick action buttons. These improvements create a faster, more intuitive, and trustworthy shopping journey.
