Building a mobile app solution based on behavioral patterns observed through an autoethnography of my own lived experiences.
ROLE
UX Researcher
April 2026 (2 weeks)
3 UX Researchers
+ Expedia Research Team
Outcomes
major findings with actionable recommendations for product improvement
found the compare tool to be a helpful, pleasant surprise
rating for overall experience ease of use (median)
AI Integration
(3/6) users requested an AI chatbot to guide their booking experience without prompting
AI tailored property suggestions based on user preferences for easier browsing
Users expect an AI assistant within the comparison tool to help weigh their booking options
These features are currently being implemented by the Vrbo team!
Tool Personalization
Algorithm oriented system that remembers previous search filters and user preferences
A vibe-coded implementation example.
Context
A vacation rental booking platform under Expedia that connects property owners with travelers looking for their next stay.
Using qualitative and quantitative research methods I designed study task flows and facilitated user interviews in collaboration with my team. I translated research findings intro recommendations presented to major product stakeholders.
The problem space.
If a data point impacted 2+ participants and/or had a severity level of 2 or lower, it was considered a finding.
Research Design
To make the exploratory nature of the first usability study for the mobile web experience more digestible we broke the experience into four key areas.
Part 1
Part 2 - 1 hour
Part 3
Trip planning habits
Experience withVrbo
Their upcoming trip
Task observation
Product usability
Feature discovery
Overall product perception
Overall product perception
Additional thoughts
To make the exploratory nature of the first usability study for the mobile web experience more digestible we broke the experience into four key areas.
PROS
CONS
Multiple Angles
Screen Visibility
Less Intrusive
Additional Set-up
Moderator Multi-Tasking
Findings
Participants could fully complete tasks despite discoverability issues. The experience was familiar to the web experience (for existing users) and other booking sites (for new users).
Ease of Use Likert Scale for quantitative data measurement
What is a finding?
If a data point impacted 2+ participants and/or had a severity level of 2 or lower, it was considered a finding.
01 Visibility Issues
Participants were unable to discover the comparison tool without moderator guidance, indicating an issue of discoverability.
—
Participant 1
5 / 6 participants
Needed tool navigation guidance
3 / 6 participants
Looked in saved properties page
3 / 6 participants
Missed the tool entry point (button)
3 / 6 participants switched between browser tabs to compare properties
02 compare tool usability
Most participant frustration stemmed from a lack of their personal prioritized information when deciding on a property.
—
Participant 2
—
Participant 4
—
Participant 4
03 Location Context
participants explicitly define
participants look at
—
Participant 6
participants look for
—
Participant 5
However
Some users cannot easily evaluate personal preferences using just the primary tool provided, forcing them to piece together clues from photos or reviews.
—
Participant 1
—
Participant 1
—
Participant 3
Result
Some users cannot easily evaluate personal preferences using just the primary tool provided, forcing them to piece together clues from photos or reviews.
Recommendations
Recommendations were prioritized by estimated experience impact and team resource costs.
Tool Personalization
A compare tool more responsive to user's specific needs supports them to confidently select their ideal property.
Algorithm oriented system that remembers previous search filters and user preferences
A vibe-coded implementation example.
AI Integration
(3/6) users requested an AI chatbot to guide their booking experience without prompting
AI tailored property suggestions based on user preferences for easier browsing
—
Participant 4
Expected an AI assistant within the comparison tool to help weigh their booking options
These features are currently being implemented by the Vrbo team!
Enhanced Visibililty
Navigating to saved properties was difficult, due to its hidden location and misaligned identifiers. The "compare stays" entry point lacked visual weight and users frequently scrolled past.
The text of the button could also be an alternative color, or increased in size. to increase visual weight
Many participants struggled with finding the trips page and associated the "heart" icon with favorites.
Next Steps
Our study was largely explorative. In order to further develop the product based on the research insights further research in user trust and discovery is recommended.
Lack of quantitative data due to open-ended nature of study
Small number, and unrepresentative set of participants due to time constraints
Test phone usage limited data to IOS user flows only
A/B testing of the compare tool discoverability and graphical changes
User research on expectations of AI integration
Some of these steps are already being taken!
What I Learned
Open-ended nature makes it difficulty to interject with further questions
A large scope makes it easier to miss important quantitative data collection
Ran into time crunches at the end of sessions, forcing us to guide participants to decide on properties sooner
My wonderful team!
