Case Study / 01
Timeline
March 14 – May 4, 2026
Role
Lead Product Designer
UX Researcher
Team
Victoria Hong (myself)
Briana Grant
Han Lee
Process
Needfinding (User Survey, Participant Observation, Heuristic Observation, Think-Aloud Studies, Product Review Analysis)
Individual & Group Brainstorming
Lo-Fi (Paper & Pencil Sketches), Mid-Fi & High-Fi Prototyping (Figma)
Evaluation
Final concept preview showing a collaborative apartment search experience for roommate groups.
Overview
Finding rental housing with roommates is a collaborative but fragmented process. While Zillow provides extensive property listings, the current experience primarily supports individual browsing and evaluation. Roommates still have to manually share links, track tradeoffs, compare listings, and align on preferences outside the product.
Our project explored a roommate-focused rental comparison feature for Zillow’s mobile interface. The goal was to help users save listings, compare properties side-by-side, evaluate shared criteria, and reduce the cognitive effort involved in choosing housing as a group.
Problem Overview
Zillow makes it easy to browse rental listings, but it does not fully support the collaborative work that happens after users find options. Roommates often share links through group chats, revisit multiple listings, compare attributes from memory, and use external tools like screenshots, notes, or spreadsheets to track tradeoffs.
How might we
Research Methods
To understand how renters search for housing with roommates, we conducted a one-month needfinding
study focused on comparison, sharing, and group decision-making. Because collaborative apartment
search involves both individual preferences and shared tradeoffs, we used a mixed-method approach
to capture broad patterns, real behaviors, and in-the-moment decision-making.
Our research strategy included three methods: a survey to understand common behaviors and pain points,
participant observation to study how users navigate Zillow during a real search workflow, and think-aloud
studies to uncover how users reason through tradeoffs while comparing listings. Together, these methods
helped us triangulate findings and identify design opportunities beyond what a single method could reveal.
We surveyed renters to understand their search habits, group size, tracking methods, and collaboration challenges. The results helped us identify common workarounds, such as sharing links through messaging apps, relying on screenshots, and manually comparison listings outside of Zillow.
View Survey Questions and Results.
We observed a mobile Zillow search workflow focused on browsing, sharing, and comparing listings with roommates. This helped us identify friction points in the current experience, including difficulty revisiting saved listings, comparing multiple options, and keeping roommate preferences visible during the decision process.
View Observation Results.
We asked participants to complete a rental search task while verbalizing their thoughts. This revealed how users prioritize factors such as price, location, amenities, commute, and roommate needs. The session also helped us uncover where users experienced confusion, cognitive overload, or uncertainty during comparison.
View Think-Aloud Guide and Results.
We synthesized findings across all methods into key themes and design opportunities. The strongest patterns centered on fragmented communication, limited comparison support, and the need for a shared decision-making space. These insights informed our final features, included shared folders, side-by-side comparison, renter preferences, and compatibility insights.
Research Findings
Across our research, we found that users were able to search for apartments on Zillow,
but struggled when they needed to compare listings with roommates and make a shared decision.
Because Zillow primarily supports individual browsing, users created their own workarounds
using memory, screenshots, group chats, saved links, notes, and spreadsheets.
These workarounds helped users continue their search, but they also made the process more
disorganized, time-consuming, and stressful.
Because listings are viewed one at a time, users often switch back and forth between properties while mentally tracking differences such as rent, commute, amenities, safety, and location. This increased cognitive load and made it difficult to confidently compare multiple options.
Users shared listings through group chats, shared documents, spreadsheets, and in-person discussion. As a result, links became buried, preferences were hard to track, and group feedback was spread across multiple platforms.
60% of participants reported losing tracking of a listing during comparison. This showed a clear need for better organization, shared tracking, and centralized comparison tools.
Users reported that making the final decision and aligning roommate preferences were the most stressful parts of the search process. Because each roommate may prioritize different factors, the final decision often required users to balance competing needs without structured support.
User Personas & Profiles
Based on our survey, interviews, and observation sessions, we identified three primary user personas that represent different rental search behaviors, priorities, and decision-making styles. These personas helped us design features that would support the diversity of roommate groups.
Our research revealed that successful collaborative search depends on group dynamics. Roommate groups typically include a mix of decision-making styles: some prioritize price, others focus on lifestyle, and most try to balance group consensus with personal preferences. Understanding these profiles helped us design features that support both individual priorities and collective decision-making.
Task Analysis
We analyzed the key tasks involved in group apartment search, from initial browsing through final decision. Each task presented specific challenges and opportunities for improved support.
Users independently browse listings and save favorites to share with roommates.
Roommates exchange listings through various channels and discuss initial reactions.
Users examine multiple options side-by-side to understand tradeoffs across attributes.
The group reconciles different preferences and commits to a final choice.
User Journey Mapping
By mapping the current user journey, we identified key moments of friction, frustration, and opportunity. Participants reported highest stress during the comparison and decision phases, when cognitive load peaked and group coordination became most challenging.
Critical pain points included:
After sharing a link, roommates often forget why it was recommended. No context travels with the listing, making it hard to understand initial interest.
Comparing across multiple properties requires mentally tracking rent, location, amenities, and safety—often across inconsistent format variations.
Without visibility into which criteria matter most to each roommate, final decisions become prolonged negotiations with limited data to support agreement.
Ideation
Our brainstorming focused on reducing cognitive load, addressing user workarounds,
and treating roommates as multiple stakeholders with constraints that need to be
reconciled.
We generated up to 10 concepts individually for approximately 20 minutes,
converged as a group, and selected three design alternatives for initial prototyping.
Design alternatives
Each design alternative addressed a different part of the collaborative search problem: comparison, shared organization, and decision support.
This design introduced a central dashboard where roommates could collect favorited listings, view group preferences, and quickly understand which apartments were the strongest matches. Listings would be organized based on compatibility with shared criteria such as budget, commute, space, and amenities. A compatibility score would help users scan options quickly, while supporting details would explain why each listing was ranked highly. This concept helped move users from fragmented sharing toward a more transparent group decision-making process.
This design included shared folders, side-by-side comparison, renter preferences, notes, compatibility scoring, and conflict highlighting. The goal was to make both alignment and disagreement visible. By showing where roommate preferences matched or conflicted, the system could help users better understand tradeoffs and reach decisions more efficiently. This concept ultimately shaped the direction of our final prototype.
Instead of forcing users to switch between multiple listing pages or rely on memory, this design allowed users to select apartments and view them in a structured side-by-side comparison. Key listing details such as rent, location, commute, amenities, and policies would be organized in one view. Users could also attach notes directly to a listing, reducing the need to track feedback in separate apps. This concept addressed the need for a centralized comparison workspace where users could evaluate options more clearly.
Evaluation
We evaluated the mid-fidelity prototypes through semi-structured user interviews and comparative tasks. Participants interacted with the design alternatives and provided quantitative ratings and qualitative feedback on clarity, ease of use, cognitive offloading, and collaborative decision support.
Rated highly for clarity and reducing the burden of memorization.
Viewed as intuitive and useful for centralizing group decisions.
Helped users make faster decisions by aggregating preferences into one metric.
Removed lower-value features and merged the strongest concepts into one system.
Final solution
The final high-fidelity prototype integrates four core features: Shared Folder, Side-by-Side Comparison Chart, Compatibility Score & Insights, and Renter Preferences. Together, these features help roommates externalize information, reduce memory burden, and make more transparent housing decisions.
A centralized shared space where roommates can save and access the same set of listings, reducing the need to send links across group chats.
Users can compare selected listings across rent, bedrooms, bathrooms, availability, square footage, amenities, parking, safety, and neighborhood factors without switching between pages.
The system evaluates saved listings against each roommate’s preferences and provides a score, criteria breakdown, and explanation to support faster group decision-making.
Each roommate can define constraints such as budget, bedroom and bathroom count, parking, amenities, and safety preferences. These inputs power the comparison chart, dashboard, and compatibility insights.
Final evaluation results
In the final evaluation, participants rated all four core features highly. Shared Folder received the highest overall ratings, followed closely by Side-by-Side Comparison and Compatibility Score & Insights. Participants especially valued features that reduced mental tracking and made group preferences visible.
The evaluation also revealed an opportunity for improvement: users relied on Compatibility Score & Insights, but wanted clearer explanations of how the score was calculated.
My role
As a UX researcher and product designer, I played a leading role across the project’s research, design, and evaluation phases. I conducted participant observation, developed structured materials for the think-aloud study, synthesized needfinding insights, created personas and task flows, and shaped the team’s brainstorming process. I also designed and evaluated the side-by-side comparison concept, refined the high-fidelity prototype, and helped execute the final usability evaluation to validate how the redesigned Zillow experience supported collaborative decision-making.
Reflection