Case study
Expense sharing
Adding the ability to split expenses with your friends over Messenger
UX
UI
Mobile app
Case study
Expense sharing
Adding the ability to split expenses with your friends over Messenger
UX
UI
Mobile app
Case study
Expense sharing
Adding the ability to split expenses with your friends over Messenger
UX
UI
Mobile app
Summary
Messenger is a popular instant messaging platform. It allows people to instantly message their friends and share media. Like other instant messaging platforms, people rely on it a lot to coordinate their social events, including asking for reimbursements after someone pays for everyone at an eatery that does not split bills. Currently there are no special features within Messenger to make this mentally taxing and potentially labour intensive task easier.
I propose a feature for Messenger to facilitate expense sharing within an Australian context (where I am based). This feature would allow users to share expenses, request reimbursements (evenly split or itemised), and track payments.
User interviews and competitor analysis revealed key social dynamics affecting expense splitting. I designed and tested two flows to a high level of fidelity, which rated highly for ease of use and user benefit. While the feature is well-designed for Australia, it requires adaptation for other contexts due to varying payment systems and social dynamics.
Methodology
Competitor analysis
User interviews
Empathy & story mapping
Personas
Finding opportunities (POVs/HMWs)
Defining the solution & features
User flows
Low fidelity wireframes & initial testing
Prototyping & usability testing
Revisions
Summary
Messenger is a popular instant messaging platform. It allows people to instantly message their friends and share media. Like other instant messaging platforms, people rely on it a lot to coordinate their social events, including asking for reimbursements after someone pays for everyone at an eatery that does not split bills. Currently there are no special features within Messenger to make this mentally taxing and potentially labour intensive task easier.
I propose a feature for Messenger to facilitate expense sharing within an Australian context (where I am based). This feature would allow users to share expenses, request reimbursements (evenly split or itemised), and track payments.
User interviews and competitor analysis revealed key social dynamics affecting expense splitting. I designed and tested two flows to a high level of fidelity, which rated highly for ease of use and user benefit. While the feature is well-designed for Australia, it requires adaptation for other contexts due to varying payment systems and social dynamics.
Methodology
Competitor analysis
User interviews
Empathy & story mapping
Personas
Finding opportunities (POVs/HMWs)
Defining the solution & features
User flows
Low fidelity wireframes & initial testing
Prototyping & usability testing
Revisions
Competitor analysis
I performed a competitor analysis, looking at 4 different products:
Three of these are standalone mobile or web apps for splitting expenses with others, while Meta Pay is a payment option that has been integrated into Meta’s platforms in certain markets. However, Meta Pay is primarily aimed at vendors selling products or services over their social networks - it isn’t suited to peer-to-peer payments or bill splitting.
Strengths
Most of apps allow you to accrue a tab and pay it off at your leisure, avoiding unnecessary transfers if you and your friends alternate paying for each other
Most of the apps integrate payment option to allow direct reimbursement
Weaknesses
All of these tools require everybody you are splitting expenses with to sign up and link accounts (Meta Pay leverages its existing social networks).
Payment and reimbursement options are restricted to what the specific apps support, which may be limited and country specific.
Ongoing costs associated with handling payments and maintaining the platforms means these platforms charge to split bills. Alternatively Settle Up only handles bills splitting calculations - not the payments themselves (but this makes it less useful).
User research
I chose to interview 5 people remotely for 20 - 45 minutes each:
Aged in their 30s and 40s
Use a mix of instant messaging tools
Mostly renters
All had jobs; one was also studying as well
One participant owned an apartment and had a teenage son
The line of questioning was focused on understanding how people split bills with their friends over instant messaging tools, to understand how we can simplify this process.
User research
I chose to interview 5 people remotely for 20 - 45 minutes each:
Aged in their 30s and 40s
Use a mix of instant messaging tools
Mostly renters
All had jobs; one was also studying as well
One participant owned an apartment and had a teenage son
The line of questioning was focused on understanding how people split bills with their friends over instant messaging tools, to understand how we can simplify this process.
User research
I chose to interview 5 people remotely for 20 - 45 minutes each:
Aged in their 30s and 40s
Use a mix of instant messaging tools
Mostly renters
All had jobs; one was also studying as well
One participant owned an apartment and had a teenage son
The line of questioning was focused on understanding how people split bills with their friends over instant messaging tools, to understand how we can simplify this process.
Findings
Existing bill splitting tools
A few interview participants currently used existing tools to split bills with others. One used Splitwise to split household expenses with their partner. Another used an app called Beem (which I was unaware of when doing competitive analysis) to split household expenses as well as shared expenses with friends. An advantage of these apps is to create a tab that they then settle occasionally.
Reciprocity
Participants generally preferred tit-for-tat expense splitting - “you pay for this one, I’ll pay for the next” - especially with friends they regularly shared expenses with. People don’t like to talk about money (nor be seen as a penny-pincher) and don’t always expect to be fully reimbursed. In addition, tit-for-tat allows alternative methods of compensation (e.g. by providing their house as a venue).
Effort
Most don’t like expending the effort required to accurately account for expenses, calculate splits and chase up people who owe them money. Often the figures are fudged because the exact dollar amount isn’t important and will be forgiven.
Unequal shares
Often bills are split evenly if everyone consumes approximately the same, however sometimes some individuals consume significantly more than others which requires a more complex itemised split. In addition, some people are prepared to pay more than others and some may have a mental budget of what they are prepared to spend, both of which may cause problems.
Social dynamics
Multiple participants indicated paying for others can be a power play: some like to pay for others as a way of showing power, while others are eager to reimburse to not be “subservient”. Underpinning these are feelings of shame or frustration for not being able to pay for others, and a feeling of overburdening others. However, friends may pressure other friends to pay someone back.
Findings
Existing bill splitting tools
A few interview participants currently used existing tools to split bills with others. One used Splitwise to split household expenses with their partner. Another used an app called Beem (which I was unaware of when doing competitive analysis) to split household expenses as well as shared expenses with friends. An advantage of these apps is to create a tab that they then settle occasionally.
Reciprocity
Participants generally preferred tit-for-tat expense splitting - “you pay for this one, I’ll pay for the next” - especially with friends they regularly shared expenses with. People don’t like to talk about money (nor be seen as a penny-pincher) and don’t always expect to be fully reimbursed. In addition, tit-for-tat allows alternative methods of compensation (e.g. by providing their house as a venue).
Effort
Most don’t like expending the effort required to accurately account for expenses, calculate splits and chase up people who owe them money. Often the figures are fudged because the exact dollar amount isn’t important and will be forgiven.
Unequal shares
Often bills are split evenly if everyone consumes approximately the same, however sometimes some individuals consume significantly more than others which requires a more complex itemised split. In addition, some people are prepared to pay more than others and some may have a mental budget of what they are prepared to spend, both of which may cause problems.
Social dynamics
Multiple participants indicated paying for others can be a power play: some like to pay for others as a way of showing power, while others are eager to reimburse to not be “subservient”. Underpinning these are feelings of shame or frustration for not being able to pay for others, and a feeling of overburdening others. However, friends may pressure other friends to pay someone back.
Finding opportunities
From POVs I identified some key problems that could be addressed:
Helping people to more easily split irregular expenses with housemates
Helping people to more easily split expenses and request reimbursement for outings with their friend groups
Helping people to collectively agree on a shared budget for outings with friends
Helping people who like to go tit-for-tat with friends to take turns providing for others, so there is more transparency and trust
Finding opportunities
From POVs I identified some key problems that could be addressed:
Helping people to more easily split irregular expenses with housemates
Helping people to more easily split expenses and request reimbursement for outings with their friend groups
Helping people to collectively agree on a shared budget for outings with friends
Helping people who like to go tit-for-tat with friends to take turns providing for others, so there is more transparency and trust
Defining the target audience
Based on my research, I defined an archetype with challenges, frustrations and goals, representing the target audience. These are their goals:
Cat, the social butterfly
Wants it to be easier and more transparent to communicate and share expenses with others
Wants to spend less time and effort tracking finances
Wants to ensure she gets reimbursed for household expenses
Wants to spend less time and effort chasing after friends and flatmates for payments
Wants to not worry she is being taken advantage of when it comes to shared expenses
Wants to avoid awkward conversations about money
Defining the target audience
Based on my research, I defined an archetype with challenges, frustrations and goals, representing the target audience. These are their goals:
Cat, the social butterfly
Wants it to be easier and more transparent to communicate and share expenses with others
Wants to spend less time and effort tracking finances
Wants to ensure she gets reimbursed for household expenses
Wants to spend less time and effort chasing after friends and flatmates for payments
Wants to not worry she is being taken advantage of when it comes to shared expenses
Wants to avoid awkward conversations about money
Defining feature set
I wrote some HMWs to narrow my focus:
How might we help friends who go out to dinner or events together to collectively split the bills over instant messaging so there is no confusion about who owes what?
How might we help friends who like to do favours for each other to be confident their generosity will be reciprocated so they can have more positive relationships with each other?
How might we help people who share expenses to easily calculate splits and share with people over instant messaging, to reduce the effort and work required?
Based on the problem space, I decided what features should be included in any solution and prioritised them.
Required for MVP
Posting expenses to a message thread and automatically calculate splits
Reimbursement tracking
Upload receipts & invoices - OCR would automatically read the items to be split
Payment details sharing - users could persistently share their PayID or other details to receive reimbursements
Next in priority
Event threads - temporary message groups for organising events and tracking associated shared expenses
Pay it forward - users can volunteer to do a favour or buy something for someone else; the user who accepts is reminded to pay-it-forward
Requesting advance shared payments - users can post proposed shared expenses and others can indicate how much they are willing to pay
Defining feature set
I wrote some HMWs to narrow my focus:
How might we help friends who go out to dinner or events together to collectively split the bills over instant messaging so there is no confusion about who owes what?
How might we help friends who like to do favours for each other to be confident their generosity will be reciprocated so they can have more positive relationships with each other?
How might we help people who share expenses to easily calculate splits and share with people over instant messaging, to reduce the effort and work required?
Based on the problem space, I decided what features should be included in any solution and prioritised them.
Required for MVP
Posting expenses to a message thread and automatically calculate splits
Reimbursement tracking
Upload receipts & invoices - OCR would automatically read the items to be split
Payment details sharing - users could persistently share their PayID or other details to receive reimbursements
Next in priority
Event threads - temporary message groups for organising events and tracking associated shared expenses
Pay it forward - users can volunteer to do a favour or buy something for someone else; the user who accepts is reminded to pay-it-forward
Requesting advance shared payments - users can post proposed shared expenses and others can indicate how much they are willing to pay
Key flows
I diagrammed key flows for:
Posting a shared expense to a group message thread
Viewing and updating a shared expense
Diagramming the flows enabled me to resolve specific implementation details, in a way that made sense to the user.
Key flows
I diagrammed key flows for:
Posting a shared expense to a group message thread
Viewing and updating a shared expense
Diagramming the flows enabled me to resolve specific implementation details, in a way that made sense to the user.
Low fidelity wireframing & testing
I initially created some simple low fidelity wireframes to illustrate the two flows, and tested these with two of my original interview subjects. Some key feedback I received was:
Overall, the participants thought it would be useful and increase their chances of being reimbursed for shared expenses
The initial steps were easy to understand, while the latter steps were a little confusing, especially with regard to terminology
There was too many steps and confirmations overall
There should be more flexibility in the ability to split expenses (either evenly or by item) and it should be earlier in the flows
There should be more information and functionality within the group message thread view
Low fidelity wireframing & testing
I initially created some simple low fidelity wireframes to illustrate the two flows, and tested these with two of my original interview subjects. Some key feedback I received was:
Overall, the participants thought it would be useful and increase their chances of being reimbursed for shared expenses
The initial steps were easy to understand, while the latter steps were a little confusing, especially with regard to terminology
There was too many steps and confirmations overall
There should be more flexibility in the ability to split expenses (either evenly or by item) and it should be earlier in the flows
There should be more information and functionality within the group message thread view
High fidelity prototyping & testing
With the feedback from my low fidelity wireframes I created high fidelity prototypes for the following flows:
Splitting an expense evenly over a group message thread
Splitting an expense by item over a group message thread
These flows were then tested in moderated tests over video chat with 5 different users, all of whom used Messenger at least some of the time and had previous experience splitting bills.
Every participant rated the flows either a 6 or 7 out of 7 for ease of use, and the flows were well understood and made sense once the participants had been through them once.
Most of them also said that the feature provided significant utility and they would use it. However, there were some minor usability issues and potential gaps in functionality that were highlighted.
Certain social dynamics were also not accounted for. For instance, it is considered gauche in certain cultures to ask for reimbursement, so others often ask for reimbursement on behalf of you. Likewise, people may wish to pay on behalf of others.
High fidelity prototyping & testing
With the feedback from my low fidelity wireframes I created high fidelity prototypes for the following flows:
Splitting an expense evenly over a group message thread
Splitting an expense by item over a group message thread
These flows were then tested in moderated tests over video chat with 5 different users, all of whom used Messenger at least some of the time and had previous experience splitting bills.
Every participant rated the flows either a 6 or 7 out of 7 for ease of use, and the flows were well understood and made sense once the participants had been through them once.
Most of them also said that the feature provided significant utility and they would use it. However, there were some minor usability issues and potential gaps in functionality that were highlighted.
Certain social dynamics were also not accounted for. For instance, it is considered gauche in certain cultures to ask for reimbursement, so others often ask for reimbursement on behalf of you. Likewise, people may wish to pay on behalf of others.
Priority revisions
→ Some users suggested a live OCR camera view, to make the process of capturing a receipt faster and less error-prone
Some users weren’t sure where to tap to see the itemised receipt from other steps in the flow
→ Added clearer affordances
Multiple users indicated the ability to split the cost of a single item (e.g. a shared dish) was important
→ Added long-press gesture to split a single item between any number of people
Multiple users indicated that there is an important use case for being able to adjust expenses and the splits after initial entry. Also, they also mentioned being able to see the remaining amount left to be paid on the bill is important.
→ Added ability to adjust bill/split, and remaining total to message thread view
Priority revisions
→ Some users suggested a live OCR camera view, to make the process of capturing a receipt faster and less error-prone
Some users weren’t sure where to tap to see the itemised receipt from other steps in the flow
→ Added clearer affordances
Multiple users indicated the ability to split the cost of a single item (e.g. a shared dish) was important
→ Added long-press gesture to split a single item between any number of people
Multiple users indicated that there is an important use case for being able to adjust expenses and the splits after initial entry. Also, they also mentioned being able to see the remaining amount left to be paid on the bill is important.
→ Added ability to adjust bill/split, and remaining total to message thread view
The future
Currently the prototype is in a good state. The usability testing illustrated it provides a lot of value to users and is reasonably intuitive. The priority revisions address the minor usability issues and gaps.
However, more thought should be given to thinking about how this solution would work in different countries, cultures and contexts.
For instance, currently this solution is designed for an Australian context - prices generally include tax, there is no tipping, and people transfer money to each other using PayIDs. Obviously this wouldn’t work as-is in a country where any of these things isn’t true.
Furthermore it doesn’t fully account for the social dynamics of expense sharing. This goes especially for different cultures, such as places where it is considered gauche to ask for reimbursement and others do it on your behalf.
The future
Currently the prototype is in a good state. The usability testing illustrated it provides a lot of value to users and is reasonably intuitive. The priority revisions address the minor usability issues and gaps.
However, more thought should be given to thinking about how this solution would work in different countries, cultures and contexts.
For instance, currently this solution is designed for an Australian context - prices generally include tax, there is no tipping, and people transfer money to each other using PayIDs. Obviously this wouldn’t work as-is in a country where any of these things isn’t true.
Furthermore it doesn’t fully account for the social dynamics of expense sharing. This goes especially for different cultures, such as places where it is considered gauche to ask for reimbursement and others do it on your behalf.
