How could we intelligently match drivers and riders to encourage daily carpooling?


While new ridesharing options such as Uber and Lyft are quickly expanding, carpooling has been an underused method of transportation with huge potential market especially for daily commuters. It is odd that only 10% of American commuters are HOV (high-occupancy vehicle), and encouraging commuters to carpool has remained a huge challenge for the government.


Commutify aims to improve the carpooling experinece for commmuters and encourage people to carpool more often on a daily basis. It is an carpooling applications designed for commuters, which aims to better accommodate both driver’s and rider’s needs, and create optimal matches between drivers and riders in an effective and efficient way.


Individual project


4 weeks, 2019

My Role

User research, UX design, visual design


Sketch, Illustrator, and After Effects

Final Deliverables

A. Create Regular Schedules

In the “schedule” page, the user could create regular schedules with specified route and time, which will be automatically applied to the corresponding days in the upcoming weeks. The requests will be automatically posted on the online platform. The regular schedules could be easily turned on or off in order to accommodate more flexible schedules.

B. Send Requests to Potential Carpoolers

In the “calendar” page, the user could easily view and manage all the rides being planned out for the upcoming weeks. The user could enter the ride request page to browse the list of all potential carpoolers. The user could then filter carpoolers, and send requests to the ideal ones. 

C. Confirm Rides and Set Up Groups

Multiple days could be grouped together, so the user could send one request to book multiple rides with one particular carpooler. When the other commuters approve the request, all rides are confirmed, so the user could have multiple rides with the same carpoolers for the upcoming week.

Design Process

01 Early Research

A. Literature Review

Even with the HOV lane and carpool parking spot policies, encouraging commuters to carpool remains a challenge. While 85% of all American commuters are dependent on the automobiles for weekday commuting, almost 90% of them are SOV (single-occupancy vehicle) and only 10% of them HOV (high-occupancy vehicle), according to the American Community Survey (ACS).

In this project, I decided to unpack the commuters’ current carpooling experience and analyze the factors that discourage commuters from carpooling and the frictions that cause inconvenience.

B. Preliminary Research

I started my research process with gathering information from various sources. Two papers I found online provided me an overview of the key work-related (non-household) carpooling factors and their effect sizes:

Key Insights:

Quite unexpectedly, the research reveals that the effective matching between potential carpoolers itself would act as the biggest incentive for commuters to carpool, rather than monetary or instrumental factors. It appears that if people are introduced to an ideal carpooler with matched schedules, they usually won’t mind taking another person for a ride.

However in reality, the recurrent and pre-scheduled nature of carpooling makes the matching process extremely challenging compared to other ridehailing services such as Uber or Lyft:

How could we make the matching process easier and more accessible, so that more commuters would be encouraged to carpool and enjoy the benefits of carpooling?

C. Qualitative Interviews

Given the information collected, I decided to put my emphasis on the pre-ride matching process of carpooling. I conducted interviews with 5 people (3 riders and 2 drivers) who regularly carpool to commute in order to understand the user pain points existing in the current pre-ride process.

Key User Pain Points:

1.    Schedule management: lack of flexibility and stability
For the young millennial professionals, their working schedules are not “9 to 5 every single weekday”, but rather a combination of both regular schedule and flexible schedule. When people are having fixed schedules, recurrent appointments on a regular basis are desired; when people are having flexible schedules for a particular period of time, more irregular appointments with different time or routes are desired.
How could we coordinate people’s different types of schedules and better balance between flexibility and stability?

2.    Request creation: lack of effectiveness and efficiency
Interviewees complained that with the current carpooling apps, they need to send a new request for each ride. Even though they may have the same schedule for the whole week, they need to send requests for each day separately and be matched with different carpoolers for each ride, which involves huge communication cost.
How could we synchronize drivers’ and riders’ needs and increase the effectiveness and efficiency of the matching process?

3.    Unpredictable carpoolers: lack of safety and trust
Carpooling is not an inelastic demand, and therefore people’s psychological and social factors rank equally high as economical or instrumental factors. Both drivers and riders want some levels of autonomous control over the matching process.
How could we ensure the privacy, safety and comfort for all the parties involved in the carpooling experience?

D. Competitive Analysis

I then took a close look at the existing carpooling apps and how they are approaching to the above user pain points. A better solution is needed to respond to the above three problems.

02 Ideate

A. Wireframes

The initial wireframes focus on the rider’s persona:

B. User Testing and Feedbacks

I then asked my interviewees to test my first prototype, and specifically asked them whether this prototype was effective in addressing the three user pain points.

In my first prototype, the users could create single ride request for a chosen day in case they have more flexible schedules, or create the same requests for multiple days together in case they have more stable schedules.

Problem 1:    Setting up a new schedule for each week involves much repetitive manual labour, and people consider their schedules more stable than that. People are also afraid that they might forget to create schedule for the upcoming week.

Problem 2:    This prototype doesn’t really solve the second user pain point. Even though users could now create multiple ride requests together, each ride request is still independent from each other. Users have to separately enter each ride one by one in order to contact potential carpoolers for that particular ride. This means that even though one user might have the same route and schedule from Monday to Friday, he/she is going to be matched with five different carpoolers for each ride.

C. Design Choices

Try to solve Problem ①
How could we coordinate people’s different types of schedules and better balance between flexibility and stability?

Try to solve Problem ②
How could we synchronize drivers’ and riders’ needs and increase the effectiveness and efficiency of the matching process?

Try to solve Problem ③
How could we ensure the privacy, safety and comfort for all the parties involved in the carpooling experience?

Other Screens:Trip in-progress