Lyft’s Trust & Safety Initiative
Anyone who’s ever used a ride-sharing app understands the convenience of Lyft. You tap a button and a driver picks you up. Behind this basic premise, however, are the many complications involved with navigation, environment, and human error. I joined the NYC team in 2018 to focus on solving the problems behind the pickup experience.
Being based in a dense urban environment allows one to see all the cracks in the user experience. The aforementioned issues were compounded by the unpredictable factor of human social interaction, something unavoidable in a two-sided marketplace.
Many riders are apprehensive toward the idea of getting into a stranger’s car, and the fear of riders getting into the wrong vehicle was becoming a concern at the legislative level. In our understanding the problem through user research we landed on 3 key insights:
- Riders are unclear what information to focus on, and when. The vehicle make, model, and license plate are key to identifying a ride, and should be emphasized.
- The app isn’t responsive to real-world conditions. UI included an animating map overview, but otherwise information was static throughout.
- There’s a lack of established trust between rider and driver. The app displayed minimal information about the driver.
Using these insights for a design sprint, we generated design concepts ranging from the visually unintrusive, to the more prominently displayed friction. We could see that designing for trust on Lyft’s platform had much untapped potential.
Enhancing the arrival
To address the insights from UXR our first step was to help riders better find the correct car, as well as to more easily identify the driver.
The results from this experiment were successful all around, with reported no-shows, cancellations, support tickets, and even pickup times decreasing. This was a step in the right direction, and we still had many more solutions to implement.
Addressing legislation
The moment immediately prior to entering a vehicle is crucial, and a newly drafted bill requiring drivers in certain regions to display an identifying placard on their vehicle was an opportunity to design for this moment.
Building on the previous experiment we added an entry point to optionally scan the placard, which would either alert the rider of an incorrect vehicle or confirm the start of the ride (vehicles missing the placard could also be reported).
Ultimately usage of the scanning feature remained relatively low, especially by veteran users. Was the current implementation too lightweight? We gathered feedback on one of our more heavy-handed ideas, PIN code pairing.
No silver bullet
Both riders and drivers pushed back against the additional friction to their ride experience. It was clear that when accounting for human error it wouldn’t be a single solution that would prevent 100% of riders from accidentally approaching the wrong vehicle. We had to also design around the moment of arrival, and leverage other signals and technologies to do so.
The future of safety
As of today the NYC team has shifted priorities, emphasizing building trust in the waiting experience as opposed to during the moment of arrival. But our foundational work remains pivotal in having helped shift the conversation away from seeing safety features as one-size-fits-all. Various teams now work collaboratively to contribute to the suite of functionality, with Safety Tools and Location Sharing being one of the latest additions available in-ride.
As we leverage new signals and technologies to improve Lyft, it’s exciting to be able to work toward a more equitable and trustworthy future for both riders and driver. Interested in learning more? Let’s chat.
