Building the Platform I Wished Existed
Every college student faces the same problem: finding off-campus housing is chaos. I didn't just design a solutionâI founded a company to build it. Purple Panda Rentals is the rental platform I wished existed when I was a student.
| Dimension | Details |
|---|---|
| My Role | Founder & Product Designer |
| Timeline | 2 months (mobile app design) |
| Research | 4 interviews, 270 survey respondents |
| Outcome | Validated mobile app design for launched platform |
The Problem Was Personal
Finding off-campus housing as a college student is a mess. You're competing with other students for limited listings, negotiating with landlords who don't take you seriously, and trying to coordinate move-in dates around academic calendars that don't align with standard lease terms.
I lived this. So did everyone I knew. The options were Zillow (not built for students), Craigslist (sketchy), or word-of-mouth (limited). None of them understood the specific needs of college renters.
So I started Purple Panda Rentalsâa rental listing platform built specifically for college students. The website launched in December 2023. The next step was mobile.
"Along with saving money, students benefit from the independence and housing flexibility that off-campus housing provides."
Validating the Problem at Scale
Before designing the mobile app, I needed to confirm my assumptions weren't just my own experience. I ran a nationwide survey with 270 college students and conducted interviews with four former students about their leasing and searching habits.
What I Expected
The survey confirmed what I suspected:
- 76% currently live off-campus
- 53% use Zillow to find housing (a tool not built for them)
- 57% said proximity to campus was the most important factor
- 51% said rental price was the deciding factor
Students were using generic tools because nothing better existed.
What Surprised Me
The survey also revealed pain points I hadn't fully appreciated:
- 81% prefer to stay in one apartment for multiple semestersâstability matters more than I assumed
- 42% found it frustrating to find someone to sublease their apartment between semesters
- 34% had maintenance and landlord issues with no way to vet reputation beforehand
- 15% struggled to calculate total cost when rent, utilities, and living expenses were listed separately
The sublease problem was bigger than I expected. Students weren't just looking for housingâthey needed to offload housing when plans changed.
"Finding a student to sublease their apartment was frustrating between semesters."
The Competitive Gap
I audited both direct competitors (student-focused platforms like College Pads) and indirect competitors (Zillow, Apartments.com). The pattern was clear:
Mainstream platforms had the best user experience and largest inventoryâbut no student-specific features.
Student platforms had relevant featuresâbut clunky interfaces and limited listings.
No one had married Zillow's usability with features tailored to the college experience.
Designing the Mobile Experience
With the research foundation set, I moved into designing the mobile app. The website was live; now students needed to search on the go.
The Sublease Decision
When defining the app's scope, I faced a tradeoff: include a sublease section or a roommate matcher? Both had value. I couldn't do both well in the initial release.
The research made the call. 42% of students found subleasing frustratingâa concrete pain point with high frequency. Roommate matching got positive feedback in interviews, but sublease functionality addressed a more urgent, more frequent need.
I prioritized sublease. Roommate matching went on the roadmap for future releases.
"Based on the students I researched, a sublease section prioritized over finding a roommate. This finding surprised me."
The UI System
Purple Panda's brand was already established from the websiteâbright, approachable, student-friendly. The mobile app needed to extend that identity while meeting iOS conventions.
Testing with Real Users
I built three prototypes in Figma and ran usability tests through Maze with 17 participants. Three tasks, representing the core use cases:
- Create an account and complete onboarding
- Search with filters and favorite a listing
- Find a sublease and message the student
What Worked
Task 1: Account creation â 100% success rate. Onboarding questions were appropriate and completion time was adequate.
What Didn't
Task 2: Search and favorite â 85% success rate. Users struggled with the filter screen (clicking text instead of toggles) and couldn't easily see the heart icon for favoriting.
Task 3: Sublease and messaging â 63% success rate. Users couldn't find their way back to the inbox after sending a message. The back arrow was too subtle.
The problems were specific and fixable. I iterated.
After Iteration
I ran the tests again with the updated designs.
| Task | Before | After |
|---|---|---|
| Search and favorite | 85% | 100% |
| Sublease and messaging | 63% | 100% |
The changes worked. But testing also surfaced a deeper issue I hadn't fully solved: trust.
Users questioned whether they could trust a stranger subletting an apartment. Security and reputation weren't addressed in the current design. For future iterations, I noted the need for verified student profiles or review systemsâfeatures that would take more than UI changes to implement properly.
Reflection
Purple Panda taught me the difference between designing for a prompt and designing for a real problem.
When you're the founder, there's no brief to fall back on. You have to define the problem yourself, validate it with real people, and make tradeoffs that affect a real business. The sublease vs. roommate decision wasn't academicâit determined what the product would actually do.
The usability testing reinforced something I keep relearning: you can't predict where users will struggle. I thought the filter screen was straightforward. Users disagreed. The only way to find out is to put the design in front of real people and watch what happens.
What's next: the publisher side of the platform (for landlords listing properties) and the trust features that testing revealed were missing. Building a product is never doneâyou just ship what you can and keep learning.