A real-world usability study of how riders understand, monitor, and act during the waiting phase of an Uber ride.
Skills
Usability Testing Product Design · User Research
Team
Ruohan Li · Annie Huang Alex Foster · Zhian Hu
Duration
January 2026 March 2026
Stakeholder
Uber HCDE 517
Project overview
Understanding uncertainty in the space between booking and pickup.
This project examines how riders experience and navigate uncertainty during the waiting phase of Uber rides, using real-world usability testing to identify breakdowns and opportunities for clearer, more trustworthy interactions. The project is sponsored by Uber, collaborating with UW HCDE's 517: Usability Testing Course.
01 · The problem
Waiting looked simple. The system status was not.
Consistency and standards
Language implied completion while the rider was still waiting.
01
False completion
Using the word “Congratulations!” implies a successful outcome. This creates a false sense of completion when the user is actually still in a waiting state.
02
Mismatched language
The phrase “Confirming your ride” is used while the system is still searching. In a user's mind, “confirming” happens after a match is made, not during the search.
03
Unclear system status
The UI simultaneously tells the rider it is “looking for a ride” and “confirming one,” leaving the user unsure of exactly what stage the process is in.
User control on the pickup page
When something felt wrong, intervention paths were hard to find.
01
Limited controls
When a driver arrives, users have limited control if the situation feels incorrect.
02
Common errors
The system does not provide easy paths for intervention when the driver is on the wrong side of the street, the pickup point is unclear, or the app shows “arrived” while the vehicle is facing the wrong direction.
03
Communication barriers
There is a need for better opportunities for driver-rider communication to resolve these “something feels off” moments before a ride is canceled or abandoned.
02 · User tasks
We tested the wait as it actually happens.
Participants moved through three connected moments: committing to the ride, communicating an ETA, and monitoring the driver's arrival.
01
Confirm the ride and wait to be assigned a ride
Goal: Understand how participants interpret commitment and expectations while the system finds a ride, and verify whether “finding a ride” status is clear.
User success metric: Participants can clearly explain what they expect to happen next and what they have committed to.
02
Send a message to a friend
Goal: Understand participants' perception of ETA and how they communicate it with others over text.
User success metric: The participant is able to send a message to a researcher with the time of their arrival.
03
Monitor the app while the driver is on the way
Goal: Observe how participants track progress and manage uncertainty.
User success metrics: The participant understands where the driver is and what they should do while waiting.
Participant profiles: seven college students, all iPhone users, with varied rider frequency and Wait & Save familiarity.
Where & when did we test?
A real route, in a real urban waiting environment.
Starting location: Westlake Link Station
Destination: Chinatown Link Station
Participants requested a ride from the starting location to the destination.
Real-world urban environment with high ride demand.
Conducted from late morning to afternoon, 11 AM–3 PM.
Finding 01
Cancellation discoverability changed after driver assignment.
Before driver assignment
Participants easily located the cancel button.
6 out of 6 participants quickly pulled the draggable sheet open [A] to locate the cancel button [B].
“So my first impression is that it’s not very visible at first glance. But I know that if I pull [the tab] up, it's gonna have like the little button right there for you to cancel.”
Participant 1
After driver assignment
Participants could not locate the same action as easily.
2 out of 6 participants viewed other screens or buttons before finding the cancel button. 4 out of 7 participants scrolled to the bottom of the screen, as they had before driver assignment, and then went to the top to locate the cancel button in Ride details.
Opportunity
Keep cancellation visible and spatially consistent.
01
Relocate the cancel button to be instantly visible without additional navigation.
02
Make the cancel button location more closely match before and after driver assignment stages.
Finding 02
Staying informed during the wait was inconsistent.
When does it get inconsistent?
Expectation dropped when the system changed without surfacing the change.
Inconsistent notification activity
Critical arrival messages did not reach everyone.
01
Missing arrival alert
4 out of 7 participants did not receive the “Your driver is arriving” notification. Two of the four did not receive it inside the app, while one of the four did not receive it outside the app.
02
Uneven message sequence
One participant, P3, received the “Check your ride” notification, but not “driver is arriving.”
03
Settings mismatch
A participant's iPhone notification settings did not always reflect how they actually wanted to be alerted during a ride. (P6)
Opportunity · Granting user affordances
Let riders choose how a time-sensitive ride reaches them.
Finding 03
Waiting information was clear, but driver matching remained opaque.
User insight
Riders understood the time range, but not what the system was doing within it.
Opportunity
Make the matching process legible without adding noise.
Finding 04
Riders lacked reassurance that the remaining wait was still acceptable.
Evidence 01 · Price is not the primary driver
Waiting tolerance mattered more than savings.
Riders evaluate the value of Wait & Save primarily based on whether the waiting time feels acceptable in their situation. Price savings play a secondary role. 6 out of 7 riders referenced waiting tolerance, while 2 out of 7 referenced price savings.
Evidence 02 · Decision logic
The decision depended on whether the wait still fit the rider's situation.
Before confirming, estimated waiting ranges help riders judge whether the wait feels acceptable. Once riders enter the driver-matching waiting room, the interface provides limited cues to confirm whether the wait is still within their tolerance and whether the cost-time trade-off remains worthwhile. This may lead to impatience and cancellations.
Opportunity
Restore reassurance during the waiting process.
What we would do next time
Protect the signal, the participant, and the consistency of the study.
Save the post-task questionnaire for after the ride
Allow participants to give more honest feedback.
Give participants a break from high-pressure tasks
Reduce pressure from tasks and potentially from conversations during the ride.
Prepare more probe questions
Account for variable wait times and gather more consistent, richer insights.
Improve data consistency
Keep the same roles across all sessions and have more than one notetaker.
