Overview

As part of understanding the problems that university students face, my team and I decided to focus on the topic of managing shared bills between roommates. Our team set out to understand how people currently manage shared bills and pinpoint major concerns in the process.

Preliminary Research

To uncover existing bill-splitting practices between student roommates, we used topic-related keywords like ‘roommates’, ‘bills’, ‘splitting expenses’, etc. to find articles and research papers. We found that shared bills proved to significantly impact the lives of students staying in shared spaces. Despite many product offerings currently in the bill sharing space (i.e. Splitwise, Splittr, etc), users still feel feelings of dissatisfaction and distrust with their roommates when splitting shared bills. We then used this research data to create an affinity diagram.

‍

The high-level themes discovered through affinity mapping, such as ‘fairness’, ‘conflict’, and ‘money problems’ showed a trend of mismatched expectations between lenders and lendees when splitting bills.

The Problem

Our research helped us identify that there's a mismatch in expectations between people who share bills. This helped create project-level research questions.

   1. What factors affect someone’s expectations when sharing bills?

   2. Why is there a mismatch of expectations between people who share bills?

Target User Group

We then chose to narrow the scope of the user group to help gather denser data points. We identified our target users by using the following criteria -

• University students between the ages of 18-30 years old
• Reside in off-campus housing
• Share at least one common expense (i.e., Wi-Fi, dish soap, toilet paper, etc) with a roommate

The ‘off-campus’ parameter was chosen as students who live on-campus (i.e. residence dorm) receive most shared expenses from the university (i.e. Wi-Fi, toilet paper, etc), which could be classified as an individual expense.

Interviews

The goal of the semi-structured interviews was to understand participants’ attitudinal beliefs around splitting bills. We created a set of 5 core questions and asked follow-up questions wherever necessary.

“A lot of it is based on trust because the person is taking the responsibility for it...”

“...our method of communication was inefficient since it was so indirect. We just wanted to avoid conflict.”

“... In person, there are no missing lines of communication. You can see facial expressions and it’s quicker too.”

A cultural model was used to analyze this data.

‍

We then identified the following key points.

   1. Interactions and payments between roommates are reliant on trust and shared values.

   2. Participants preferred face-to-face communication over other methods.

   3. Roommates prefer to track payments using traditional datasheet tools like Excel or newer methods like Splitwise.

Diary Study

We then decided to conduct a diary study to collect behavioural data from participants. The questions for the diary study were provided to five participants via a Google form that they could complete daily. There was a set of 3 core questions, along with follow-up questions set up using branching methods on the form. Participants were requested to complete the form every evening once they were confident no more purchases would be made for the day. The anonymity of the participants was maintained by assigning participant IDs to each participant.

‍

At the end of the diary study, the collected data was exported to Google sheets to conduct quantitative analysis. The frequency counts and mean results were compared across participants to identify common patterns of behaviour between them that linked to the research questions for this study. Using this information, we created a flow model to better understand more about how users communicate and divide tasks between them.

‍

From this model, we were able to identify the areas of conflict and within the communication and information flow between roommates.

   1. Messaging between flatmates on virtual group chats was not followed upon.

   2. Face-to-face communication can create conflict when conversations need to be tracked.

Design Recommendations

From our research, we understood the journey the participants experienced during the bill splitting process.

• Forming a pre-agreed contract that outlines shared expectations on the methods/frequency of bill related communication, methods of splitting bills, methods of product replacement, and payments
• Adhering to pre-agreed upon methods of product replacement and bill payments
• Process of payment transfers for shared expenses

We then proposed eight system design recommendations.

1. Form Pre-Agreed upon Contracts that can be modified with everyone's approval.
2. Create a list of recurring shared items unique to the household members and categorize items into pricing tiers.
3. Simultaneously notify other roommates that a shared resource is close to finishing and allow for others to “claim” it.
4. If there is no “claim” after multiple notifications, the task should be automatically assigned to the roommate who has volunteered the least.
5. Allow users to either direct message their roommates with a request or collectively send a calendar invite to all roommates to request an in-person conversation.
6. Allow users to upload an expense and select the members who will be using and subsequently paying for the resource.
7. Send automatic periodic reminders to members who have outstanding payments.
8. Inform users about whether roommates have fulfilled a repayment request from another roommate.

Future Steps

If this project were to continue, we would address all the limitations of our current work and conduct more research with a larger spread of target users. We would also potentially try expanding the team to have 2-3 people focus on determining social factors and bill tracking. In that way, we could create more potential design recommendations to enable cohesive social relations between people who live together.

Learnings

   1. Identifying research limitations helps understand the research scope.

     During this project, we made note of the limitations our research had -

• We were unable to conduct more thorough research due to a lack of time and resources.
• Convenience sampling made it easier for us to find participants, but this sample wasn't generalizable.
• The diary study happened to coincide with the busy schedules of a few of our participants during reading week.

     Since we identified these during the research phase, we realized that the data generated could not be generalized to every single roommate, but it could apply to college students in shared living situations with some more research.

   2. It's important to prioritize ideas that address user issues.

    Besides the eight design recommendations, there were two other ideas we considered but didn't include. The first one was the option to track expenses, however results from the diary study overwhelmingly showed that most users were not interested in a way to track expenses. The other idea included considering social factors, but since the social relationship between roommates was beyond the project scope, we decided not to conduct further research. Hence, by choosing ideas that were backed up by research and addressed user issues, we made sure that our design recommendations were actionable and helpful.