UX Case Study:
Bill splitting is a common UX problem addressed with apps
As a frequent group diner, I knew no-one used a specialised mobile apps to split a bill. I wanted to find out out how others got around this problem. It seemed they faced an issue that no existing app seems to be trying to solve.
Interviews with people who have split a bill before
8 bill splitters were interviewed most between mids 20s to mid 30s. All were smart phone users. An interviewee mentioned using their phone calculators to do a split, otherwise, none of the other interviewees used an app to do it.
On the process of bill splitting, people say they:
have asked the restaurant to do the split if it was simple enough (3 persons)
were not too particular about splitting accurately (6 persons)
do not use an app to help them with splitting a bill (8 persons)
find the math of splitting tedious (2 persons)
On collecting money, they:
find reminding people to pay is embarrassing (3 persons)
would forget to collect the money owed (3 persons)
would let it slide if the amount was small enough (4 users)
find honestly forgetting to pay is a real problem (2 users)
Summarising the findings
with a persona
The product would be focused on Sue, who is really concerned with maintaining her social relationships and has to deal with forgetful people. She is a highly empathetic person who feels that reminding peers who forget would somehow affect the relationship with them, perhaps demonstrating a lack of trust. She would rather avoid sending reminders or let it slide.
On bill splitting mobile apps
A quick google many bill-splitting apps showed up, though none of my interviewees have used any of them. I did a competitive analysis and summarized functions that I thought were essential for a bill splitting product, mostly in terms of how easy or hard to use.
Problem to address and possible solutions
None of the competitor apps addressed the issues of:
people forgetting to pay
reminding people to pay is uncomfortable
It would seem, a solution has to go further than bill splitting. To avoid the discomfort, it would seem that there should be somehow a reminder system from a third party that would help take away the discomfort of asking personally.
The solution I proposed is a feature that can be incorporated into restaurants with digital tablet menus: a bill splitting function with the option to send a reminder for those who can’t pay on the spot.
The onus is then on the debtor to pay as reminders are sent from the restaurants. No one has to download any app. It made the most sense to me as it matches the customer’s journey at the end of the meal when bill splitting is discussed. Also, all the data that is required from a split is in the restaurant’s system, requiring less input such as dishes that were had, prices, taxes and service charge.
Restaurants' tablet menus
I proceeded to do a quick survey of restaurants' tablet menu apps. Some characteristics include:
It is part of the restaurants' point-of-sale system.
Restaurant staff set discounts and happy hour prices.
Some apps have a bill-splitting function
Of those I have seen, none had a reminder function.
Apt sys and Tab Square are 2 local e-menu tablet providers
Apt sys has a personal mobile ordering app for restaurants, which also lets customers earn points.
How MenuMinder would work
When it is time to settle the bill, a waiter may guide a group if they would like to split the bill and recommend them to use the function in the menu. The user flow would thus be as below:
Iterations of the uneven split and reminder functions
The challenge of bill splitting is the uneven split of a bill. Based on the interviews, most would probably still go for the even split with this app. The uneven splits would be for situations where the disparity in what was ordered is too great, e.g. people who drink alcoholic beverages versus those who do not. I have tried to simplify the process:
using initials rather than inputting full names
direct matching of foods to persons from mid-fi version
based on the input of names, users would just need to add a phone number if they wanted reminders
The tablet app is designed for iOS. The following are iterations at different fidelities for the uneven split process and sending reminders.
Medium fidelity wireframes
Uneven split - assigning of items
The design for this function evolved much from its first sketch prototypes. A simple drop down to assign a food to a fellow diner proved confusing for users. I then experimented allowing users to directly link a food from the left to the right (see mid-fi wireframes). This was intuitive enough for users but in terms of an order summary, it seemed messy to trace the items that one has picked.
In the high fidelity designs, I designed a drag and drop system whereby users drag food items from the bottom of the screen into lists created for each diner.
Pay for a friend/ combine payments
In earlier iterations, I wanted to allow users to pay for a friend. Users were confused here. Reframing it as combining payments did not work either. Reflecting on the test results, I hypothesize that if a friend was short of money and needed help, they do not a system to help them with calculating this. This function was thus removed.
Send a reminder function
This function got some interesting feedback. 1 users says he would find it awkward to use with close friends or bosses, but may use it with those less close. He then pondered and added he may use it with friends with a history of forgetting. Other comments included:
Would use this function for large amounts owed
Afraid of giving up number or email address for fear of spam - added a liner "No spam will be sent"
High fidelity prototype
In this high fidelity prototype, combining payments was removed. I also added infographics in for contextual help and onboarding with the payment process, assignment of items and what to expect after a reminder is sent.
Further development on this feature would depend if it is to be integrated into an existing point-of-service tablet menu solution. This would require further understanding of the point-of-service system for restaurants.
The high fidelity version would require usability testing. Visual design may occur at the same time though it may require a versability of design as most tablet apps can be customised by restaurants to suit their corporate theme.
This was my first UX project and synthesising interview transcripts proved to be quite challenging. The affinity mapping method seems seems to make sense if one was really selective or has fewer data points to begin with. It probably is a good activity to get a team or stakeholders involved and get familiarized with the data. I had 8 users and after affinity mapping for a while, I was dissatisfied and felt I was leaving data out without knowing if I was doing so.
I decided to pick out the main points of all my interviews, placed them as a list in excel and mark all the interviewees who have said similar points. From there I could tell who said what and the patterns that were emerging in a more organised way without making assumptions. This to me was a visually more organised, thorough and objective way to look at the points that were made.