Shared Homes Finder

Intelligent, thoughtful roommate matching.

Skip case study, view wireframes with commentary >

Try out the working prototype >

Case Study

Project Summary (the "Why?")

The task at hand was to create a product for Homes.com, designed for a new market not yet targeted by the Homes.com team; specifically, people looking to create shared homes.

The project brief imagined that Homes.com, a site best known for showing listings of homes for sale, wanted to provide a service that matches people to create households of multiple roommates.

Problem Statement

I defined the problem as follows: 

There’s an increasing demand for shared living situations, but not a good service that intelligently matches like-minded individuals who share lifestyles and are seeking new homes.

In other words, there are services for finding shared homes:

...and services for finding like-minded individuals who match closely with you (i.e., dating and friendship services): 

...but no services that combine the two together.

Solution Statement

Create an “OKCupid for roommates,” adding the housing data of a service like Homes.com to the intelligent matching of a detailed online dating service like Match, OKCupid or eHarmony.

Research and Team

(the "Who?")

After understanding the problem, I teamed up with two other students in a small design group, and began learning about the target audience for SharedHomes.com, and what they were looking for in a roommate matching service.

Learning About Our Customers

We started by sending out a survey of 21 questions to a wide audience of peers, including people of all ages and demographics. They were asked a variety of questions about their experiences searching for shared living spaces in the past or present.

We also interviewed several roommate seekers as well as a real estate agent specializing in rentals. 

We learned that many of the qualities people cared about most in a prospective roommate were not related to basic demographic information, like age, gender, race or religion. Instead, our customers focused on nuanced issues of living styles, personal compatibility, level of sociality, etc. 

User Personas

Based on our findings, we created three personas with similar needs as the people we had surveyed.

We made our primary customer "Tripper," a 28 year old sales rep new to town, looking for something more than just the humdrum inadequate Craigslist-type service to find a good set of roommates. Tripper is an extrovert and wants to make sure the people he lives with match his social style and his habits.

A younger customer, "Alice," and an older retiree "Trudy" became secondary customers that we also worked to accommodate, but without the same focus.

Team Responsibilities

Our team of three divvied up roles and created a process based on our unique strengths:

Jonida: User Researcher

Conducted interviews, compiled survey data.

Lead - Homepage & Dashboard

Sephora: Visual Designer

Refined the style guide & user personas.

Lead - Filtering & Search Results

& me: Product Manager

Defined business goals, project milestones & timing.

Lead - Sign Up & Profiles

Results

(the "How?")

After exploring business needs, understanding our customer better, and mapping out team process, we set out to design our product.

Competitive Analysis

We looked at several shared housing websites, as well as our favorite dating site, OKC.

Overall, we decided our challenge was to combine the best features of each:

• the ease of use of Craigslist;
• the smart survey questions and profile display of OKCupid;
• the organic easy-to-use filtering options of Air BnB; and
• the creative yet simple approach to displaying housing info and photography on Zillow. 

User Flows

We then created an outline of the paths through the site navigation for our three user personas. How could we get each of our customers paired with a great roommate or home?

Site Map

We then translated those flows into a hierarchy of our new website. Tactical decisions were made:

• who wireframes which pages during our 2-week sprint;
• which pages we should show in our prototype (represented in orange below) as critical to the main use cases, and
• which we would just mention in presentation without actually laying them out (in white below).
• We also made some final decisions about what features we wanted to have on our most critical page, the dashboard. These features are represented as ovals below.

Ready to see the product itself? 

View wireframes with commentary >

Skip wireframes, go to final prototype >