Modern Living Room
Contextual Inquiry, Modeling and Low-fi prototyping
Re-Imagining roommates.com for off-campus housing

Relocating for college and finding a shared apartment off-campus is a taxing task. A number of platforms such as apartments.com enlist apartments and houses available in specified areas for rental but their services are limited to leasing apartments whereas students are generally looking at shared apartments or finding roommates to share an apartment. Roommates.com allows users to search for rooms in shared apartments or houses based on their preferences, and list rooms for rental. However, this service is not tailored to needs of university students. On exploring the website, I witnessed some design flaws. For further investigation, I conducted a contextual inquiry (CI) which revealed that some simple features and clarification of content can amplify positive user experience (UX).  I created a low fidelity (low-fi) prototype to evaluate effectiveness of proposed features.

4dc0f9aa-282f-495c-9de4-8312a2b282ba.jpg
App User
User Persona

Background

Name:                Maryam Ali

Age:                    20 years

Occupation:      Sophomore at community college

Specialization:  Business (major) and computer science (minor)

Future plans:     Transfer to Business program at Towson University

Core needs and expectations

  • Finding suitable shared apartment near Towson university without the need to visit every potential apartment

  •  Find credible roommates according to personal preferences

Web Consultation
Contextual Inquiry

I conducted a contextual inquiry to understand how potential users would interact with the website, what problems they might run into, and what their response would be to these breakdowns. Since, the problem at hand is how to modify the website to needs of a special population i.e., university students looking for off-campus housing, I contextualized the tasks accordingly.   

Tasks 

Since the idea is not limited to any specific task on the website, I covered all the basic functionality of the website in the contextual inquiry. However, for modeling and low-fi prototyping the basic unnecessary tasks were omitted and specific tasks were regrouped for further analysis. Following are the tasks the user was asked to perform:

  • Sign up for roommates.com

  • Create profile for finding a room near Towson university.

  • Find matches and examine one of the matches

  • Use simple "Search" feature 

  • Look for only "Female" matches by updating your preferences

  • Check for any messages

  • Hide your profile

  • Switch account type and create a listing to look at possible renters

  • Reach out to a potential roommate

  • Contact Customer Service and ask how to deactivate account

  • Delete account

Apartment Building
Grouped Tasks

Important tasks are grouped into 3 for further analysis.

Apartment
Task 1

Create user profile to find a room near Towson University: 

This task includes:

  1. Signing up for roommates.com

  2. Fill out information about:

    • apartment type, move-in date, rent, leasing options, and living situation

    • personal details such as occupation, gender, pets and kids, and profile picture

    • roommate preferences such as gender and age etc. 

Task 2

Look at matches based on your preferences and look for roommates using "Search" feature

This task includes:

  1. Exploring matches by analyzing:

    • personal details e.g., employment and age.​

    • place for rent e.g., pictures and location map

    • option to message for further details

  2. Search for potential roommates using "Search" feature:

    • look at sorting options 

    • try to display only female options. Requires user to update preferences

Task 3

Create a listing for a place and reach out to potential roommates

This task includes:

  1. Switch account type to having a room from needing a room

  2. Fill out information about:

    • apartment type, move in date, rent, leasing options, living situation, and pictures of place

    • personal details such as occupation, gender, pets and kids, and profile picture

    • roommate preferences such as gender and age etc. 

  3. Click on one of the potential renters (roommates)​ from matches displayed and reach out through message after analyzing the profile

Sequential Models

Sequential models detail the intent and user actions to help understand breakdowns in the pipeline. Creating these models helped me identify what changes I need to make to enhance user experience.   

Task1
Task 2
Task 3
Low Fidelity Prototype

Low-fi prototype is created using POP by Marvel app on iPad to test following enhancements. 

Check out the prototype here (Note: There is a pre-defined path since the purpose is to evaluate specific features. Click on the bottom right corner to scroll down.)

Clarifications and added features
Clarification of question about living situation
Clarification of question about living situation

Clarification of question about rent
Clarification of question about rent

To add more details
To add more details

Clarification of question about living situation
Clarification of question about living situation

1/5
Credibility
Reviews and specific profession of the potential roommate
Reviews and specific profession of the potential roommate

Better map with markups of places around
Better map with markups of places around

Street view option to explore the nearby area
Street view option to explore the nearby area

Reviews and specific profession of the potential roommate
Reviews and specific profession of the potential roommate

1/3
Gender Filter
Added gender filter to "Search" for ease
Added gender filter to "Search" for ease

Added gender filter to "Search" for ease
Added gender filter to "Search" for ease

1/1
Sort by Distance
Added "Sort by Distance" feature
Added "Sort by Distance" feature

Searches sorted by distance from Towson University
Searches sorted by distance from Towson University

Added "Sort by Distance" feature
Added "Sort by Distance" feature

1/2
User Feedback

1

CLARIFICATIONS

User was not confused about most of the statements. User smoothly set up her profile and did not indicate any specific issues.

2

CREDIBILITY

User raised no concerns about credibility of potential roommates. She specifically mentioned the interactive map to be a good feature. 

3

SORTING

"Sort by distance" feature is well-received by the user. It is an important feature specifically for university students looking for housing.

4

FILTERS

Gender filter for basic "Search" reduced the cognitive load to go and update the profile preferences every time users want to update their preference. This made the interaction effective for user.

REMARKS

Upon asking, the user mentioned that she is satisfied with the prototype. She did not suggest further changes at the moment. I understand that inquiring only one user is insufficient and limiting, therefore, CI  and subsequently evaluation of prototype should be conducted with multiple users to meet the needs of most of the targeted population. 

The Wall of Ideas
Contextual Inquiry, Modeling and Low-fi prototyping
Conclusion

Contextual inquiry followed by low-fi prototyping is an effective tool to understand user expectations, design flaws, and shortcomings in the pipeline at a very low cost to developers while keeping the whole process user-centric. I conducted CI to understand interaction of university students with the popular platform roommates.com for finding off-campus housing. User faced four major issues during the interaction i.e., 1. ambiguities and missing options while setting up user profile; 2. credibility of potential roommate and apartment is questionable because of lack of reviews and information; 3. sorting the search results is limited (no sort by distance); 4. lack of needed filters for basic "Search" thus, limiting users to seek matches by updating their profile every time they wanted to tweak the search. 

To address these concerns, I created a low-fi prototype with added clarifications, options and features. Evaluation of the prototype revealed that the adjustments enhances user experience by addressing major concerns. The user expressed that the added features are sufficient at this stage. Conducting CI and prototype evaluation with only one user, and only once is a limitation of this experiment. Recruiting multiple users and running multiple iterations of this process would yield better understanding of the problems faced by university students.