An Optimal Way to Ride in Groups
A Case Study | UCSD course: Cogs127 Design Human-Data Interaction
In this Design Human-Data Interactions course, I worked in a team of three on generating new ideas to improve existing Uber mobile application. We were inspired by our common experience on real-world problems on the existing Uber app and hoped to extend the function for a solution. Our team came to an agreement on creating a function that can help Uber users, especially young adults, better organize rides with large group of people.
Our extension is named UberGroup. I was the generator of all UI components, creator of the functional flow (Flow 2), and the director of the user research interviews. I analyzed the research data and reflected on the user research and testing. My focus is to preserve the users' habits while providing them with a new and comfortable experience with UberGroup.
"Uber riders have difficulty coordinating Ubers for large parties to arrive to the same destination which leads to frustration, anxiety and separation of group."
We focused on younger adults who use Uber to coordinate transportation for events as our target user, because this group have potentially similar thoughts as ours.
From our own experiences and casual observations, Uber does not satisfy user’s convenience in traveling with large parties. Our problem was indeed a prevalent issue and a majority stated they would like to see the changes we offered. Thus, with our redesign, I believe Uber will be a more user-friendly and user-centered application that can help riders meet their goals with Uber more comfortably, efficiently, and conveniently.
Communicate location quickly and effectively for a specific period of time
Plan specific places to meet should some members become separated from the larger group
Organize rides easily for more than 6 people and/or more than one car full of people
Understand the problem
Driver VS Riders?
Find users' thoughts
What we do later?
Analyze the gap between these and optimize user experience
Our team conducted 11 one-on-one interviews with participants, who are 10 Uber users and 1 Uber driver. 10 young adults are interviewed to collect enough data about our target users, while one Uber driver provided us reference to the other side of the story on group riding problem. Interview is an intimate and personal method of gathering data by asking questions and observing users' behaviors. Our pool of interviewees is composed of college students, whose ages are around or slightly over 20. Since these issues are prevalent in our lives, we wanted to focus on individuals similar to ourselves and therefore mainly interviewed UCSD students or young adults in our age range.
UCSD, EE student, Junior, 20
"It’s tough when people do not communicate about where to meet or show up on time. Getting people to cooperate when they are intoxicated is hard."
Has a hard time organizing his group during an event. Feels like they either ignore him or due to issues with apps on their phone right now nothing exists to streamline this process.
Project Engineer, San Francisco, 25
"Uber should have a feature where they ask exactly how many people there are."
Finds it hard to meet up with his group of friends when taking Ubers to a venue or large event. Uses Uber as main mode of transportation when attending these large events.
* Selected the most representative interviewees selected from our interviews to represent the majority:
UCSD, HCI student, Senior, 21
"One person ordering there and back and Venmo both people, sometimes you have to order multiple, and two more people to order it. Sometime going back, hard to find a sober people to order on multiple phone, sometimes put the wrong location."
Has trouble splitting the Uber cost in both sober and intoxicated circumstances. It’s better if multiple Ubers can be ordered at once. Communication can be very difficult after drinks. Can be dangerous when intoxicated people are divided into different cars.
"I had intoxicated passengers before but the largest group is only 3 of them. I just drove them home and nothing happened. It was a normal ride except they were laughing and joking loudly."
Had no problem with either large group or intoxicated people. However better communication is needed between passengers and drivers, such as providing more information and making the map more accurate for picking up and dropping off.
Our strongest findings gave us evidence that 6/10 passengers reported having difficulties Ubering with large parties when intoxicated. 11/11 participants showed support and endorsement for our suggested app extension. Multiple users expressed the desire in the ability to order multiple Ubers on one device.
Stronger evidence for difficulties in ordering Ubers for large parties was seen for intoxicated individuals, as only 4/10 participants reported having trouble in such circumstances when not under the influence, which was unexpected.
Due to our limited abilities in interviewing Uber drivers, it was inconclusive whether these problems persist among Uber drivers as prevailingly as they do among riders. It is also inconclusive that how our new UberGroup function will influence the drivers, who are also "users" for Uber company. Due to the inconsistency in how frequently our interviewees use Uber, we cannot foresee how the trend may differ for more frequent Uber users.
Research to Problem
Users find difficulty with ordering multiple Ubers in a single group
Users would rather have a way to order multiple Ubers from a single device
Most of our data collection fell inside our target user population
Positive feedbacks on our general ideas about UberGroup
Core Pain Points
Coordination between individuals at night is difficult:
When Ubering to large events and venues, either riders in their group were intoxicated or they were headed to an event with high noise levels, making it difficult to communicate in general.
Difficulty in Coordinating multiple cars to leave and arrive at one place:
Riders expressed inconvenience in having to split up groups for Ubers and in having to gather up after the ride. The inability to know where the other group’s Uber is headed, when they departed, and when they are estimated to arrive makes this process difficult.
Inconsistency in ways of ordering Ubers for large parties:
There are multiple ways to choose ride options and combinations for selecting the cars among many Uber ride types. Users need to calculate and consider the best way to order cars, usually on multiple phones.
Offers transportation for up to six passengers from one device
Can't really call multiple cars on a single device
Not as well known as Uber
Clunkier UI than Uber
Easy transport for big groups
Helps commuters through use of shuttles
Not designed for spontaneous request service
Currently only available in AUS
Easy book for future trips
Personalize-can choose specific drivers
Only used for rides to the airports
No feature for large groups
Integrates many methods of transportation service
Not widely available
Not as well known and will impact fewer people than others
Britney just started her new job as an elementary school teacher. She doesn’t go out at night with her friends too often but when she does, she finds it really hard to uber and meet up together. She isn’t tech-savvy so she depends on her friends to lead and guide her to their destination, but sometimes feels lost and worried when she has no idea where they are headed.
Steve is a young professional who works in technology. He loves going out to venues and bars with his other friends but finds that he is always the one organizing their transportation. He constantly worries about small details and wishes that he could be the sole person to coordinate how his group got around.
Elizabeth is a CEO of a fashion & design company. She works in high-speed pace everyday and often travels around the world for business with her team. She wants to save some time from her busy life by being productive when traveling , especially for business trips, but never does because they don’t order a car large enough for the team. She constantly has to call everyone to talk even though they are all headed to the same venue.
We originally drafted two UX flows so that we could test two novel user experiences in the user testing and get feedbacks from different angels.
Flow 1（ & 2）
Designed for personas Britney and Steve:
ride with a large groups usually for special occasions.
Designed for persona Elizabeth: a group riding service by Uber for a certain period of time with designated large group of people.
Timed group creation page
Number of Passengers
Input Number of passengers
Uber provides optimal ride options
Create a group for a business trip
Set time limit
Enroll a timed group(eg. 3 days)
Uber assign cars
Uber calculates spending and order cars
Enroll the group
One pays all
Split based on cars
Everyone in the group enter the code to on their own phones
Everyone in the group enter the code to on their own phones
More than 6
Uber outputs a secret code for riding
Users enter their assigned cars
Users check others location on App
Enter a conversation and function page
Initiates a trip
Prototyping and testing
We made one paper prototype for each flow. Besides creating prototype for the user that decides to create the group, we create a simple prototype for users who are joining the group. The first prototype makes the individual ride the priority and allows users to select the UberGroup option from the scrolling menu, similar to how you would scroll up on Uber if you wanted to call and UberBlack or UberXL. The second prototype features on the group creation interface. Included within the group functionality is time restricted location sharing, an internal chat, and the ability to call Ubers from the chat menu.
Our users liked the ability to add friends to create a group and pick the specific car allocation and the ability to share the group. They also liked the ability to see instantly when and where others in their group would arrive, especially at locations that are large and hard to find others in. Buttons and procedural flow need to be made more obvious to the user, such as buttons on sharing group code or the invite friends feature.
Easier for users to navigate as it was better integrated into the existing Uber user flow.
Offered better group navigation and a more centralized information flow.
Use flow 1 to iterate according to the feedback in user testings
Prototyping and testing
Our high fidelity prototypes cooperated the users' and TA's feedback, which coordinate better with the Uber UX flow and design, and are more user oriented. We altered the features to allow for easier switching between cars and reducing the amount of excess information we displayed. We also altered buttons and layouts. In our ride customization page specifically, we altered the design so that it did not look so similar to the Uber ride selection page and would cause less confusion.
Original Uber functions
As currently in the Uber app, user see the home screen with a map and click the input box to select a destination to go before group riding selection
We implanted the UberGroup as an additional car type in the original car selection list, in order to follow the users' habits while using this application.
Number of People
To allow quick input of numbers, we switched out design from horizontal sliding for selection to keyboard input.
By using an example of number “16”, we showed the maximum number of cars allowed in this function to users
Original design: Horizontal sliding menu
Two ways of sharing code
Share the secret code with friends face-to-face by showing them by phone or orally
As riders join the trip, their first names are shown at the bottom so that the phone owner knows when to start a trip
In cases that your friends are not here yet or are temporarily off, you can also save them spots by sharing the code to their phone by texting
Original design: only has the function of sharing via text messages
Instead of showing multiple options, we made the screen a best option as default and allow the riders to customize the car selection by clicking the button on top
Original design: let the user select an option from "fastest", "cheapest", "most comfortable", and etc.
Gives users a chance to double-check the passengers before selecting the payment
Inheriting from the US convention and original Uber cost split function, we design a function that allows the owner to split the bill prior to the trip, which would be easy to manage and plan
To increase the efficiency of selection, and ensure that all the information of names and cars are saved and visible, we use a pop-up window for selecting the payment
Add and delete riders
By clicking on the name tags, the owner can manage the passengers by adding, deleting riders and switching cars
View your friend's info
Additional info for your car
Additional car icons are added to the original trip map so that each rider can view the locations and ETAs by clicking on different cars
As usual, you can see all the information of the car you are riding, such as the plate number, departure and destination location, and etc, while you can only see the location and ETAs for other cars
Our UberGroup function did not sufficiently consider users who want to be on the group rides for that particular time, but do not have an Uber app installed, or whose phone is out of battery. They will not be able to being added, splitting the costs, and seeing the location of the others. I cannot solve the very last problem because we cannot propose something like implanting a location sharing microchip to a person just for knowing other friends' Uber locations. However, for the other two, I propose that a temporary link as a potential solution. This is inspired by Uber's original function of sharing a trip via a link.
This temporary link is an add-on for the ride sharing function via text message. When the orderer chooses to share the code via the text message, there will be a plain text message with the code only, and a sharing link which connects to the Uber app. If people have a Uber app, the link will open the app for them; if they don't have Uber, this temporary link will be sent to them via text message and clickable even after the ride ends. People need to input the code from the text to the app or link to get onboard. This link serves the Uber app's function for people who do have an Uber app but have a phone with them.
Originally, the riders need to input the code or click on the text message to go the Uber app page so that they are confirmed to be on board. Now if the link is not clicked by the time that the orderer want to start the ride, there will be a secondary option-allow this person to be a "temporary free rider". What this means is that this person will still have a seat on the ride, but they can confirm that they were on the ride later and pay for the splitting costs afterwards. The confirmation process will still be entering the sharing code. This "temporary free rider" can enter his payment method, or use the link to go into Uber app and pay with existing payment methods. This link then helps to solve the problem that people do not have an access to their phone but want to be on the ride.
In our case study, we purposefully included both flow in each step before the high-fidelity prototype. We narrowed our redesign down to the first prototype driven by testing data. I am glad that we used the limited data to reflect human-centered experience on entry point, user conventions, visual design and layouts, and interactions. I think more minor issues will be reflected if we can interview more people and analyze the user testing data in a more quantitative way. This extended UberGroup feature will be potentially better than the original app, because it can extend the user group for Uber by helping users to reduce the workloads for ordering Ubers for a large group of people.
We did note exclude our second prototype until the final prototype, because I believe a more functional version is maybe anti-conventional but worth exploring, although current data do not support it. Good designers and developers do not just follow the convention but lead users to new and better conventions. I am willing to challenge conventions and create better features for users in my future projects.
Improvements for our study can be made by considering the edge cases in advance and access more data from both various Uber riders and drivers. It is possible that we found our this function is not helpful for many users after enough number of testings. This can be due to the abundance of alternative group ride apps in the future or people do not think the existing problem is a significant issue.
For future projects, It is valuable to apply the problem solving skills and user-centered redesign I learned from this case study. This experience is very helpful for me to explore more on existing application about their conventions and potential improvement through extending or redesigning the current functions. The short term case study with a team inspired me to work in more team projects creatively and coordinately.