Uber Eats Study
I worked as an Uber Eats bike courier and heavily documented the experience, creating a personal project based off of my findings. What excited me about being a bike courier was the challenge of documenting an environment constantly in motion under time pressure, Uber’s infamously deceptive user interface, and my love for biking.
Questions I addressed:
What level of information and detail is necessary for an audience?
What are methods for effectively documenting an environment in motion?
How can different technologies be utilized to create a unified graphic?
How can UI be used to subtly influence behavior through dark patterns?
How can a UI be deconstructed to uncover it's motives?
My methodology revolved around documenting as much as I could, constantly sketching and adapting the story I would tell, and synthesizing different elements to tell a comprehensive story.
I initially looked for trends in the data I collected, although given the potential variability and potentially insufficient sample size, I was wary to make statistical conclusions. The constant logging helped me reflect and record thoughts as I went however, making an informed assessment of Uber Eats far easier.
I took extensive screenshots of Uber's user interface as a primary form of documentation because the interface constantly changed throughout the summer. Screenshots were helpful in identify dark patterns, Uber promotions, and navigation flaws.
Sketching and Adaptation
My sketching served the role of aiding exploration and comprehension, exploring possible graphics to create and comprehending the confusing Uber user interface through mapping out different screens.
Much of the work surrounding this project involved iterating through different concepts while determining which information to show versus which information hide.
In describing the Uber Eats experience from the perspective of a biker, it was necessary to balance my personal experience with the larger context of Uber Eats in Pittsburgh.
Restaurants within a boost zone (determined roughly by the red shaded areas) had a significantly higher delivery fee compared to restaurants which weren't; 32% of my total pay came from the extra boost charge. Every restaurant I ever delivered from is shown on the second map, red are boosted pick ups, black are not.
While the overall distribution of restaurants I picked up from were boosted, roughly 20% of all pick ups weren't boosted. The final graphic shows the relative distribution of pick up locations through size of restaurant dot.
Each one of my drop-off locations is shown on the first map below. From an aggregated view, the distribution of drop off locations seem quite favorable, with the vast majority of deliveries being close to my starting location in North Oakland and Shadyside.
The red navigation lines denote "paid" distance traveled, while the black navigation lines denote "unpaid" distance traveled. Uber only pays deliverers for the distance between the restaurant and the drop-off location, not including the distance it takes to travel to the restaurant.
This trip shows 5 deliveries from 4 restaurants, the first two orders being from the same restaurant. The total time for Sequence A was 1 hour and 8 minutes (not including the 30 minutes it took to travel home). Idle time: 4 minutes or 6% of the total time.
This trip shows 4 deliveries from 4 restaurants. The total time for Sequence B was 1 hour and 44 minutes (not including the 6 minutes it took to travel home). Idle time: 27 minutes, or 26% of the total time.
I mapped every pick up, delivery, and intermediary travel route in order to visually compare the quantity of paid versus unpaid distance. The higher density of certain travel routes denote the "beaten path". The red navigation lines denote "paid" distance traveled, while the black navigation lines denote "unpaid" distance traveled.
Uber's Dark Patterns
Uber's UX was full of subtle design nudges which attempted, at times successfully, to influence my behavior. Some nudges were obvious, for example, an opt-in Arbitration Agreement intended to prevent legal action against Uber, while other nudges involved third party companies and rewards with hidden consequences.