Driving user value through AI


Finding low-hanging fruit and designing an adaptive user interface with Doordash







How might we modify Doordash’s interface to increase orders from users with a minimal risk of user drop-off?


This was a quick assignment to design an adaptive user interface. In this project, I learned about creating effective value propositions using AI. Professor Zimmerman encouraged me and my peers to look for “low-hanging fruit”– applications of AI that simply involve regression with an existing dataset. 

This was a group project. I worked on creating user personas, identifying value propositions, and creating the user interface.
Class
Design of AI Products and Services (05-317)

Time
October 2022



Finding a value proposition for high-paying users


By designing for the most valuable user, we can make a tremendous impact on revenue (see the 20-80 rule, where 80% of consequences come from 20% of causes). We identified Doordash’s most value users as families and businesses that order lunch for their employees. These users share the characteristics of having large and predicatable orders. We created personas for our high-value users and identified opportunities to design for. 

Even if we could marginally increase orders from our top 20%-paying users with this design, we would still drive a significant increase in revenue from that aforementioned 80% return. With an existing dataset and a relatively simple prediction task, this would be a low-risk and high-return investment.* 

Additionally, by suggesting pairings of orders, we target users who make larger orders ($50+ per order) that can help offset marginal costs of delivery and increase per-order profit.  
* There are still risks such as a user perception of a violation of privacy. Additionally, notifications pose a risk of being too overbearing on certain users if used incorrectly. 



Displaying low-risk predictions


The visual design system of this is strictly informed by the existing Doordash app design.

By using push notifications, we can predict when a user is most likely to want to re-order something. This would use a regression-type model based on existing data about meals ordered, location, and time. Because we’re actively pushing this recommendation system, we can turn high-value users into higher-value users by building habits on Doordash. 

The suggestions page displays a ranking existing orders from the user by how likely they are to order them. This reduces cognitive load for choosing food, and encourages users to order multiple items. 

By taking the form of a card, this page is also low-risk to the users. If the user taps into the notification and doesn’t want to proceed, they can easily go back to the Doordash main page. The “change or add items” button ensures that users have a readily-available option for changing their order if it’s close to what they want to have. 






How users use this AUI




1. 
David got back from work on Tuesday evening and needs to prepare dinner for two kids. He realizes he doesn’t have much time.
2.
David gets a notification from DoorDash about it being dinner time and recommending dishes that he typically orders for his family.
3.
David taps on the notification, and a popup in the DoorDash app shows a list of recommended dishes. David quickly decides and orders four meals.