Lyft Driver Badges
Allowing drivers to be rewarded for desirable driving behaviors and riders to express preferences for these behaviors
Context
In 2024 I participated in the Lyft Hackathon as a Product Designer. Although the Hackathon was officially 72 hours long, the research and ideation for the project started weeks before and continued for weeks afterwards.
This hackathon idea eventually became the basis for Preferred Driver ↗, which launched later that year in 69 regions that represent around 27% of Lyft's total rides.
Role
Product Designer
Timeline
72 hours
Team
2 Product Designers, 1 PM, 1 Engineer, 1 Data Scientist, 1 Legal Counsel
Tools
Figma
Understanding Lyft's current marketplace of homogeneous goods
In Lyft’s current marketplace drivers offer a homogeneous service and riders are not able to choose between drivers. Consider the following examples:
A multilingual driver, who speaks both English and a second language, cannot communicate this to riders, or be matched more often with riders who speak this second language as their primary language.
A nervous airport traveler, who finds themselves anxious when reserving a 6 AM scheduled ride to the airport, has no way to express a preference for a driver who won’t cancel on them.
Problem Statement
Driver
Drivers are unable to market and differentiate themselves based on the quality of service provided to riders. An extension of this problem is that they are unable to exploit earnings opportunities based on rider preferences for quality of service.
Rider
On the other hand, riders have no opportunity to express preferences for drivers who exhibit the driving behaviors they value.
Vision: Driver Badges are rewarded to drivers who exhibit behaviors valued by riders
Smooth Cruiser
A driver who provides a high-quality in-ride experience with gentle braking, smooth turns, and gradual acceleration
Reliable Driver
A driver who provides a highly-reliable pickup experience with a short match times and low cancellation rates
Dashcam
A driver who has a Dashcam, which ensures safety and driver accountability
Spanish
A driver who can converse in Spanish
Driver badges can be viewed in the driver profile, and the benefits of achieving each badge is clearly communicated
Riders have opportunity to exploit preferences for drivers with badges that represent the behaviors they value
The total impact is estimated to be $38M PMM* and a 2.7% increase in the number of rides.
This impact is the result of the following:
Driver Behavioral Improvement
Insurance Costs - We believe that by rewarding safe driving behavior through the Smooth Cruiser badge we can reduce insurance costs and generate $26MM annual PMM.
Acceptance Rate - We believe that by rewarding high acceptance rates through the Reliable Driver badge we can increase the acceptance rate by 10pp, generating an increase of 2.7% in number of rides.
Marketplace Prioritization
Acceptance Rate – We believe that through dispatch prioritization based on higher acceptance rates, we can generate an additional $0.019/ride PMM or $12MM annual PMM.
*PMM is a measure of profit
The design process started with rider and driver user stories and a list of product requirements.
Rider user story: As a rider, I want to indicate my preference for drivers who exhibit certain driving behaviors so that I may have a more pleasant experience.
Rider Product Requirement
Outcome
Riders should be able to indicate their preferences through account settings
High
Implemented
Riders should be able to indicate their preferences through the ride request screen
Medium
Not implemented, as this would distract riders from the golden path
When a rider’s preferences are matched, this moment should be celebrated by displaying the driver’s badge(s) on the driver match screen
High
Implemented
Driver user story: As a driver, I want to understand how I can achieve and maintain desirable driving behaviors so that I may provide more valuable services to riders and earn more income.
Driver Product Requirement
Outcome
Drivers should be able to see what badges they have
High
Implemented
Drivers should be able to understand why each badge is valuable.
High
Implemented
High
Implemented
High
Implemented
Medium
Not implemented, as we did not have enough time and resources
We then had time for a very quick and simple user flow.
Because of the fast-paced nature of the hackathon, we jumped straight from this user flow to creating hi-fi mocks. The mornings were spent ideating and creating hi-fi mocks and by the afternoon we would be defending and discussing our design ideas with the team. The next day this process would repeat.
The bulk of the days were spent exploring a variety of designs to fulfill the product requirements. Here are some notable explorations:
How might riders set their preferences through their settings?
Between options A, B, and C we chose option C
How might drivers see the progress towards each badge?
Between option A and B, we chose option A
How might drivers learn about a badge and understand why it's valuable?
The main design consideration for the Badge Detail screen is the ordering of information presented to the driver. From top to bottom, most important to least, we include the score, benefits, and the detailed breakdown score
Reflections on the Driver Badges project
I’m incredibly grateful for this opportunity to participate as a product designer on this project. I worked alongside experienced, knowledgeable team members on a reputable product that reaches millions of users, performing a role that is basically the same as the role of any product designer at Lyft, just in a more condensed time frame.
This project challenged me to convince others that my skills are valuable. Prior to joining this team I, as a data analyst, pitched my design experience to the Senior Designer and PM to convince them that I’d be a positive addition to the team. I’m glad they took a chance on me.
If I had more time:
I would definitely user test some of the design decisions. We came to these decisions through discussions with the team, which is a valuable exercise, but actual user feedback would legitimize our decisions.
The user research had already been done before I joined the team. Out of curiosity and also as part of a more holistic design process, I would’ve liked to delve deeper into the user problems. I’d like to know more metrics, like the percentage of drivers and riders who would want a product similar to Driver Badges, and the demographics of these users.
Epilogue
This Hackathon was a finalist for the Lyft’s 2024 Hackathon, chosen as one of the top 30 submissions out of 100+.
Afterwards, the Product and Legal teams took this Driver Badges project and began working a product, Preferred Driver. Preferred Driver utilizes Smooth Cruiser, which is the most impactful of the driver badges we explored.
A provisional patent was filed for Preferred Driver, and then it was launched.
Preferred Driver launched in October 2024 in 69 regions that represent around 27% of Lyft’s total rides.
