Design Saves Lives: A Dasher App Case Study
Putting safety in the driver’s seat.
Image by rawpixel.com on Freepik
Overview
Context
Dasher is the driver-side app of the American food ordering and delivery company, DoorDash. The algorithm connects customers with nearby restaurants and delivery drivers. However, the current order decline sequence in the app poses safety issues for drivers.
The Problem
The current UI of the Dasher app presents an order decline sequence that is confusing and distracting for drivers. When a driver receives a delivery request, they are given the option to accept or decline the order. If they decline, they are then presented with a mandatory feedback survey (meaning that drivers must complete it before they can continue using the app). These prompts take the driver's attention away from the road and can be difficult to navigate while driving, increasing the risk of accidents.
This also results in a large number of responses that are not reflective of the actual reasons why drivers decline orders. As a result, many drivers simply select random answers to quickly move past the survey and get back to work. This not only undermines the usefulness of the feedback data for the company but also creates frustration for drivers who feel that their time is being wasted.
The current order decline sequence not only allows distracted driving, but incentivizes it. Dashers are ranked by their Top Dasher Status, which is determined by several factors, including their Acceptance Rate.
Upon receiving an order, the app gives 40 seconds for the driver to either accept or decline:
~50%
of food delivery workers have been involved in a delivery related accident.
Source: Workers Institution of Cornell University30,000
vehicle accidents each year as a result of distracted driving.
Source: NHTSA Annual Report (2017)
DECLINE
Select 'Decline' buttonDECLINE CONFIRMATION
Press 'Decline OrderFEEDBACK
Select order decline reason and 'Submit'.This totals 4 clicks…whether parked or going 65 mph on a speedway (not including reading and processing time), compromising the safety of drivers and other road-users.
I conducted a survey with 18 Dashers to further understand pain points in their interactions with the app. Here were some of the key responses.
User Research
“Declining orders needs to be fixed. I just pick random answers to get it off my screen.”
“My Dasher status gets docked on things that are out of my control. Like when I’m driving and can’t accept an order, it still affects my rating.”
“Lately I just let the timer go so I don't have to answer the questions.”
Solution Space
This project should seek to reconcile…
Safety: Prioritize driver safety by minimizing cognitive load and the amount of clicks needed to decline an order.
Response Accuracy: Allow the app to obtain accurate and helpful data on why dashers decline orders.
Ideation
I began by creating a mindmap of problem areas, solution spaces, and possible remedies for the safety concerns users expressed.
I asked two brainstorming buddies, Cassy and Lillian, to help me.From this, we came to two conclusions:
1) Declining an order needs to be a one-step process.
Safety stands as priority over everything else. Drivers should be able to receive an order, decide whether they want to accept it, and execute within a matter of seconds.
2) Dasher needs a way to incentivize voluntary and accurate feedback.
Currently, Dasher’s order decline sequence is not only lengthy, but mandatory. Collecting less data that is accurate is more valuable than collecting more data that is noisy.
Next, I needed to discover when a voluntary feedback survey would solicit the most (accurate) participation from Dashers. I conducted a survey with 16 Dashers, asking them if they would complete the feedback survey at 3 points in their Dash.
With this information, I set to the drawing board. I used Figma to explore different possible user flows.
I also began exploring the potential of a “Dasher Feedback Rating” system, which would let Dashers see their feedback rate. Maintaining a high feedback rate would help boost their Top Dasher Status.
Though the surveys would be optional, Dashers would be more likely to participate since it would be to their benefit.
Delivery Request
Delivery Request Mockup
This redesign focuses on minimizing the cognitive and visual load of delivery requests. By using clear and concise language and visual cues, drivers can quickly identify the most important information.
“Decline” button outlined for visibility, relocated to driver sideThese high-fidelity prototypes showcase how reorganizing the delivery of information and its hierarchy can diminish safety hazards for drivers while simultaneously promoting the collection of more precise data, serving the best interests of the app.
High Fidelity Redesigns
Current
Ergonomic 'Accept' buttonRedesign
Reorganized information heirarchyFeedback Survey Mockup
Upon completing each order, Dashers will receive a prompt to provide feedback regarding any prior declined orders. This practice helps to enhance safety by ensuring that the survey is not answered while driving. Furthermore, the data collected is expected to be more precise since respondents will not be pressured by time constraints or multitasking.
Dasher Feedback Rating Mockup
By implementing a Feedback Rate Rating system, Dashers will be motivated to actively provide feedback. This approach will establish a more structured and less intrusive system for gathering data on why drivers decline specific orders, as they will be responding voluntarily instead of feeling compelled to do so.
Summary
The current UI of the Dasher app presents a decline order sequence that is lengthy and distracting for drivers, resulting in many drivers selecting random answers to move past the mandatory feedback survey. This undermines the usefulness of the feedback data for the company and creates frustration for drivers. To improve this, a redesign aims to minimize cognitive and visual load by using clear and concise language and visual cues, allowing drivers to quickly identify important information. Dashers will be prompted to provide feedback after each order, enhancing safety and improving data accuracy. Implementing a Feedback Rate Rating system will motivate drivers to actively provide feedback, establishing a less intrusive system for gathering data.
This case study demonstrates that safety and collecting feedback do not have to be mutually exclusive. The two can work together seamlessly, as evidenced by the proposed changes. In addition to improving safety and data collection, these changes can also enhance the overall work experience and app experience for Dasher users. By reducing distractions and promoting voluntary feedback, drivers can focus on driving safely while providing accurate and valuable feedback. Ultimately, this leads to a more efficient and satisfying experience for both the drivers and the company.