DoorDash is an on-demand delivery service that connects customers with local businesses.
What were we solving?
DoorDash is growing at an incredible rate. This means adding both new cities and new Dashers as quickly as possible. This means the costs of supporting and regulating dashers is increasing at the same speed. Reducing or eliminating these costs allows for even faster growth.
What was our solution?
To reduce support workload a self-regulating system was our chosen approach. This in short means collecting information from consumers, post delivery, and parsing this to establish what feedback is attributed to the Dasher and what is attributed to the resturant.
This project was interesting from a consumer flow point of view. The consumer has already recieved their food and is no longer engaged. We then need to re-engage them and pull out vital information.
Asking Consumers for Feedback.
Timing Feedback Requests
One of the trickier problems we had to solve was requesting feedback, this needed to be post meal but before the memory waned. Thus started some of the tastiest research I've participated in, we found ~30-40 minutes provided enough time for most consumers to finish their meal, while still providing an accurate account of their experience.
Perfect vs What Went Wrong
Unfortunately not every delivery goes perfectly. We needed to find out why deliveries weren't perfect, and where to fix them. Any delivery with a rating ≤4-Stars expands the feedback modal to include multiple options. These are also vital in attributing the ratings correctly to the drivers. A 2-Star rating due to poor food shouldn't impact on a Dashers rating for example.
Putting Dasher Ratings to Work
Translating the user ratings into something substantial to DoorDash and their Dashers was the real fun in this project. Luckily our aims for both sides align incredibely well, for DoorDash we want to create a self-regulating system, reducing workload on support staff and keeping Dashers performing at their best.
For Dashers our aim is to provide easily digestible feedback on how they're performing in a variety of areas, along with feedback on how to improve, based on their scores in each area.
Dashers are able to access their stats and deep dive data at any point, in addition to this we sent weekly push notifications prompting them to look at their ratings. If they don't see them, they won't be motivated to improve them!
Expanding on Feedback Groups
When diving into one of the three feedback groups, we wanted to provide two things; quick understanding of the numbers and immediately actionable steps to improve. With such varying data we decided it was worth investing the time into custom data visualization for each segment of feedback.