Kaya

Transforming the roommate search into a compatibility-driven experience

Team & roles

Jesh Anies - UX Research & Design
Laura Trouiller - Mentor

Project Duration

November 2024 - April 2025

Responsibilities

User Research, Wireframing, Prototyping, Usability Testing

Compatibility is Key

Finding the right roommate takes more than shared rent. It takes shared values, trust, and understanding. According to the Pew Research Center, 31.9% of U.S. households are “doubled up,” meaning the space is shared with at least one adult who isn’t a spouse, romantic partner, or college student. That’s nearly 79 million adults living in a housemate-style arrangement. As housing costs continue to rise, sharing space has become common — yet finding someone compatible remains a difficult task.

I experienced this firsthand when I moved to New York. The roommate-search process felt impersonal and uncertain: profiles were shallow, interfaces outdated, and trust was hard to build. While I was fortunate to find a roommate, the journey made one thing clear: there should be a better, more human way to connect people who are looking to share a home.

The problem

Professionals moving to or within a city often struggle to find compatible roommates who share their lifestyle, habits, and budget expectations.

The solution

Design a user-friendly, compatibility-driven app that helps people find and connect with roommates who align with their values and living preferences.

Discovery & Research

Primary Research

To understand the challenges people face in finding compatible roommates, I conducted six user interviews with individuals actively searching for or currently sharing living spaces. Before doing so, I first used a screener survey to identify participants who would provide the most relevant insights, including those with or without a place but looking for roommates, or those currently sharing a space with non-romantic roommates.

Interviews were recorded and analyzed using an affinity diagram, helping me identify key patterns and common pain points.

After synthesizing my data, I boiled down the insights to 6 key findings:

    • Trustworthy

    • Cleanliness

    • Respectful

    • Good communication

    • Fiscally responsible

    “A lot of it is about trust.”

    - Participant 01

    “We want to make sure this space is a welcoming space and that is a clean space.”

    - Participant 04

    This confirms the secondary research from what I’ve gathered online about what people are seeking in a roommate.

  • This could be conducted either in-person or via video chat as an alternative. Some would grab a meal together to better understand their personality and to see if they're trustworthy. A great way to "check the vibe."

  • Though most things could be resolved with a conversation, participants wanted to level-set expectations ahead of time to mitigate any conflict that could build up in the future. During my conversations, some participants missed the opportunity to do so early on and would have liked to discuss them in the vetting process. Some things are going to be discussed on-going so it was important to be comfortable chatting proactively and early on in the relationship.

  • Before relying on apps to find a compatible roommate, participants asked around in their own circles first. One participant utilized significant word of mouth referrals (50%) and another was usually invited by friends to be a roommate as there is already built-in trust from pre-existing relationships

    “If our friend knows you, we trust that he's making a good judgment on you”

    - Participant 03

  • They could be shorter depending on the environment's sense of urgency and level of commitment, like in New York City.

  • The most interesting recommendation that was made during  the user interviews: Include a letter of recommendation as part of the vetting process. Two interviewees wanted to include a letter of recommendation as part of the vetting process for the next time they search for a roommate and one already utilizes it as a part of their process with 3 references (professional, pastoral, personal). Similarly, one interviewees requested bank statements to verify financial capability.

Persona

From the interview insights, I developed Emily, a persona representing a professional actively searching for a compatible roommate in a new city. Emily’s story guided design decisions, keeping compatibility, communication, and trust at the forefront of every feature. Designing with Emily in mind ensured the solution met both practical and emotional needs, from lifestyle alignment to fostering a harmonious living space.

Jobs to Be Done & How Might We Questions

Based on my research, Emily’s main goal was to find a compatible roommate while minimizing risk and stress. She needed tools to filter potential roommates by budget, lifestyle, and values; communicate effectively; and verify trustworthiness. This understanding of the persona, organized in a Jobs To Be Done framework, became the foundation for key design opportunities framed through these guiding questions:

  1. How might we increase access to reliable roommate profiles?

  2. How might we reduce the time spent searching for roommates?

  3. How might we enhance communication between potential roommates?

  4. How might we promote trust and safety when looking for potential roommates?

From Insights to Concepts

Initial Ideas & User Stories

Using the insights from research and the How Might We questions, I brainstormed ideas to help users find compatible roommates. I focused on features that foster trust, highlight shared preferences, and streamline the search process. Next, I prioritized the top 10 user stories that would be included in the Minimum Viable Product (MVP) and define the key moments in a user’s journey. For this case study, I will focus on the 3 key stories:

  • As someone

    Who’s building a profile

    I want to...

    Enter details such as moving location, habits, interests, work schedules, lifestyle preferences, preferred living arrangements (e.g., cleanliness, noise levels, pets), etc.

    so that...

    Others understand who I am and to find someone with similar preferences and move-in timeframe

  • As someone

    Who’s in the middle of their search

    I want to...

    Look at user profiles

    so that...

    I can find a compatible roommate

  • As someone

    Who’s in the middle of their search

    I want to...

    Start a conversation

    so that...

    I can gauge their personality

User Flow

To visualize how users would accomplish tasks from start to finish, I created user flows for each prioritized story. These flows ensured that navigating through the app felt seamless and intuitive, reinforcing the theme of compatibility at every touchpoint. Users could easily move from discovering compatible profiles, initiating conversations, scheduling video calls, to verifying trust and making decisions — all without friction.

Sitemap

I designed a final sitemap that emphasized action-oriented interactions and familiar navigation patterns. The architecture included:

  1. Explore: Finding and reviewing roommate profiles.

  2. Messages: Communicating with matches securely.

  3. Notifications: Tracking updates and alerts.

  4. Account: Managing personal information, verification, and preferences.

Bringing the Solution to Life

Sketches

After defining user stories, flows, and the sitemap, I sketched potential interfaces to visualize how users could find compatible roommates. While the sketches didn’t need to be perfect, they needed to clearly communicate ideas for feedback with a fresh set of eyes.

Guerilla Usability Testing

I conducted guerilla testing via Zoom, asking participants to explore the scanned sketches in Figma while sharing their screens. This allowed me to observe interactions and gather feedback on how intuitive the app was for finding a compatible roommate. From this exercise, two key findings emerged:

  1. Ease of understanding: Participants quickly grasped the app’s purpose, aided by familiar UI patterns inspired by dating, travel, and social media apps.

  2. Onboarding refinements: Users suggested adding a birthday field, search bars for state/city pages, adjusted copy, categorized tags, and stops on sliders to better personalize profiles and highlight compatibility.

Wireframes

With insights from testing, I built wireframes in Figma, creating the digital framework for helping users find compatible roommates. Live copy, grid systems, and type hierarchy were incorporated to improve clarity and establish spatial relationships between components.

Building the Brand

Prior to testing, I created an extensive brand and design system that would be applied to my concept, thus transforming the wireframe into high-fidelity prototypes. This would further flesh out my ideas so that the solution closely resembles an experience that exists in the real world.

Once the direction of the brand was set, I designed a library of components that would frequently be used in the prototype. In Figma, these components were created so that adjustments to the master element would apply to all the matching elements within the prototype, streamlining the design process.

Hi-fidelity Prototype

With branding applied, Kaya became an interactive high-fidelity prototype. During this phase, I managed to improve the user experience including: 

  • Improved onboarding for users offering a room, allowing extra photos.

  • Clearer toggles for ‘looking for a room’ vs. ‘offering a room.’

  • Expanded notifications to differentiate between ‘all’ and ‘unread.’

  • Flexible account settings to switch between user types seamlessly.

Testing the Solution

Usability Test

After completing the initial high-fidelity prototypes for Kaya, I conducted usability testing with 5 participants to evaluate how easily users could navigate the app, discover compatible roommates, and complete key tasks. Overall, participants found the app “intuitive and engaging”, quickly understanding how to explore profiles and assess compatibility. Their feedback highlighted areas to refine the experience and ensure trust and compatibility remain central to the app. Here’s how the prototype was updated based on user feedback…

Critical Feedback

One critical piece of feedback was the need for users to track their connection invitations. Participants valued transparency when connecting with potential roommates, and unlike dating apps where invitations are often hidden in the back end, Kaya required a more personable front-end solution. To address this, a page was added to track invitations, and users could include a short message when sending an invitation adding a human touch to the connection process.

Major Feedback

Major feedback improved usability and clarity, helping users navigate, filter, and engage with potential roommates in ways that removed confusion, supported informed decisions, and reinforced the app’s focus on fostering harmonious, compatible living arrangements. Examples included:

Nesting the ‘Log In’ and ‘Create an Account’ pages under a splash page to reduce clutter while maintaining equal hierarchy for both CTAs.

Implementing neighborhood selection during profile creation after participants expressed confusion about location/range settings.

Switching from a heart icon to a bookmark icon for favoriting, preventing misinterpretation as a romantic “match.”

Minor Feedback

Minor feedback addressed smaller usability and clarity details, including:

  • Adding a pet option in filters and profile creation to represent roommates with pets.

  • Incorporating a calendar view for move-in dates instead of a text field and an “ASAP” option for flexible move-in dates.

  • Implementing a budget range slider for uncertain budgets.

  • Including an option to add calls to users’ calendars.

  • Adding headers under the search bar to clarify search types.

  • Adding visual indicators in message lists to show user decisions.

  • Renaming actions like “Archive” to “Available Again” and “Checklist” to “Suggested Topics” for clarity.

Second Usability Test

After applying updates from the first round, a second round of usability testing was conducted with 5 participants to evaluate the refined high-fidelity prototype. Insights were logged to further refine usability and ensure users could navigate key features intuitively. Two participants found the interface “intuitive and familiar,” especially for users accustomed to social media and dating apps. The warm color palette and human-centric design elements also resonated, reinforcing the app’s focus on compatibility and harmonious roommate connections. Here’s how the prototype was updated based on user feedback.

Minor Feedback

This round produced less feedback, but the edits helped further improve clarity and user experience:

  • Allowing users to adjust their “no longer interested” message when archiving conversations, considering some users may prefer manually sending a note rather than letting the app notify the other user.

  • Adding stops to profile sliders, helping users understand the range of options without feeling overwhelmed.

  • Including a media gallery for uploaded documents, ensuring easy access to shared images and files.

  • Clarifying room pricing terminology (e.g., “monthly rent” vs. “room offer”).

  • Rewording the “Send” button to “Send Message” for clearer action labeling.

Other feedback to consider

Some valuable suggestions were noted but were not implemented:

  • Nesting message requests under “Messages” instead of “Explore”. Two participants instinctively looked for pending messages after matching, rather than in the pre-match “Explore” section. Moving these requests would better align with natural user expectations and reduce confusion when tracking potential roommate interactions.

  • Managing multiple rooms in the same apartment. One participant pointed out that someone offering more than one room could create duplicate listings or overlapping conversations. Developing a workflow for this scenario would help maintain clarity and ensure users can make informed decisions.

  • Implementing group chat features. In shared living spaces with multiple roommates, a group chat would allow users to communicate collectively, promoting transparency and compatibility among all potential roommates.

  • Exploring document verification. Users may submit references, letters, or ID verification. A reliable verification process (e.g., email-based confirmation) would ensure authenticity, fostering trust and safety in roommate selection.

  • Adding a feature to clear or delete notifications. Users expressed concern about notification overload. Giving them control to manage notifications would enhance usability while maintaining a calm and organized experience.

Closing Thoughts

Project Highlights

A user-centered solution rooted in compatibility

Every design decision focused on helping users find roommates whose personalities, lifestyles, and values aligned.

Tested for enhanced clarity and usability

Usability tests surfaced critical and major feedback with enhancements that ensured the app was intuitive, easy to navigate, and human-centered.

Harmony between visual identity and interactions

The visual identity, tone, and interactions fostered warmth, trust, and harmony, making the roommate-matching experience feel personal.

What I learned

  • I realized that diving into design without deeply understanding the problem could have resulted in solutions no different from existing roommate apps. Research and insights informed a solution that truly helped users find compatible roommates.

  • Testing sketches and prototypes uncovered important adjustments, such as adding birthday inputs, adjusting onboarding language, and displaying only the first initial of a last name for security.

  • Organizing feedback from critical to normal allowed me to focus on the most impactful changes, improving usability while meeting user needs efficiently.

View Kaya

Selected Works