Add a Feature: Setting the conversation level with your Uber driver
Jessica Yu
UX /UI Designer
Made with

Add a Feature:
Setting the conversation level with your Uber driver

Adding a feature that (almost) everybody wanted but nobody felt comfortable using until one small discovery changed the tide.

The Challenge

I was tasked with developing a new feature for an existing app within a 4-day timeframe for a project focused on quickly iterating on proposed solutions under tight timelines while supporting proposed claims and finding ways to incorporate new features into an existing platform. 

  • The app: Uber
  • The feature: Let riders "control the conversation level" with the driver
  • Deliverables: High-fidelity prototype that reflects best path forward based on research, iteration, and testing
  • Platform: iOS
  • *All visual designs were based off of Uber Design Guide before the 2018 rebrand

We’ve all been there.

Maybe you just got off your 11 hour flight from London to LA and all you want is a pillow and a bed.

Maybe you’re headed to an important pitch and you need that extra 30 minutes to prepare.

Maybe you're just not in the mood to talk.

But how can we achieve a seamless integration that confronts a socially sensitive dynamic?

Add a Feature: Setting the conversation level with your Uber driver

The Client

Uber currently sits as the “most valuable US start-up, by a long shot” with a current valuation of $68 billion while operating in 83 countries in about 674 cities.

Presently, the app features 11 different types of rideshares where the services range from carpooling with strangers to ordering your own private luxury cars and even having the ability to order a wheelchair-accessible ride, and much more. 

All in all, Uber has done a great job in addressing an enjoyable ride for all different types of individuals, no matter the differences they may hold, so integrating a feature to allow more comfort for the rider seemed like an easy task.

Seemed like an easy task.

Probably would use
How would you feel if you could control the conversation level with your driver?
No way


At the discovery phase of my project, I conducted 6 user interviews in order to get a better understanding of the problem. 

I also took to more unconventional methods: in the 8 different Facebook groups related to digital-, nomadic-, entrepreneurship-, travel-related topics, I posted the same question "How would you feel if you could control the conversation level of your rideshares?"

There, I conducted some pseudo interviews by pulling on the threads of comments as they were coming in and started multiple threads really digging into the responses they were giving. 

I also pulled in 6 volunteer interviewees from the respondents of those same groups and started to discover there were 2 different types of users. 

After wrapping up interviews, I took to affinity diagraming to help me distill the insights and discover the main themes. 

Affinity Diagram #1
Affinity Diagram #1

Insights: Affinity Diagram #1

  • showed concern for the driver and the discomfort (regarding the rating system) the feature would cause them
  • their openness to having a conversation was heavily dependent on the situation & their mood
  • reflected extroverted qualities, but the desire for the feature was present


Situational mood vs. Social accordance

Persona #1
Persona #1

Persona #1

I initially assumed that there were "types" of users: extroverted vs. introverted. I quickly realize that this type of user wasn't separated into categories but rather the context and what mood they were in. 

Here, we came up with Jin who was the ideal persona that represented the "average" user.

Affinity Diagram #2
Affinity Diagram #2

Insights: Affinity Diagram #2

The two affinity diagrams were separated because of the last couple interviews where respondents had mentioned social disabilities where I conducted an interview with an individual who placed on the autism spectrum with Asperger's Syndrome.

Asperger's Syndrome is a developmental disorder where the individual has "significant difficulties in social interaction and nonverbal communication". In normal day to day life, behaviors related to autism can typically go unseen to a layperson.

In other words, day-to-day interactions that "average" people may think are easy and natural are actual hurdles & pain points for others with these social disabilities.
Persona #2
Persona #2

Persona #2

"…the people in the high functioning end of the autistic spectrum don't have any kind of support out there, even if it's as frequent as 11 for every 1000 Europeans, and 1 every 60 Americans. We don't have the option of saying "hey, I suffer social anxiety, please don't talk to me because I rather keep to myself." Mimicking and faking social traits it's exhausting and we have to do that every day to look neurotypical."  - Interview #6

Receiving these insights took me by complete surprise and made me realize that this feature was no longer just a desired function of the app for some but it was a necessary feature for others.


Emotional & Mental Wellbeing vs. Social accordance

So we chose persona #2. 

Although this individual didn’t represent the large majority, she did represent a certain population where this certain feature would make a huge difference as how they lived their daily lives and how they enjoyed using the app.

How Persona #2
changed the tide

Once I decided to focus on Persona #2 as my user for the project, I realized that this pivot could be of positive use for the company and its notorious reputation in recognizing this minority group of users. I figured the feature could be implemented in accordance with an Uber PR campaign raising awareness for social disabilities. 

I returned to the users who were against the implementation of the feature and had asked what their thoughts were if the feature were implemented in support of users with social disabilities. And all of them were in favor of the feature if it meant helping out others who weren't mentally or emotionally adept at handling day-to-day social interactions. I had also asked how they felt about using the feature themselves and, surprisingly, all of them reported that it made them feel more comfortable in using it themselves as well. 

Highlighting the reason for the integration of the feature, addressing a minority group that would immensely benefit from the feature, proved to be pivotal in its communication to the public. 

So here we are. 

Problem Statement

Socially-challenged individuals need a way to feel comfortable & at ease in their rideshares because social interactions give them stress & anxiety and affect their overall wellbeing.

Selected screen to implement feature
Selected screen to implement feature

Starting Point

The ride selection screen was the chosen screen to implement this feature as it seemed the most intuitive and least intrusive. 

The previous screen is targeted to search for the locations while the next screen sends your request. 

This screen also allows for any preferences to be made such as a scheduled pickup time (excluding Express Pool & Pool rides) and checking your payment details. 

I specifically chose the UberX screen to show how the feature would interface with a more complicated user flow with the addition of pickup time selections. 

Ideation & Testing

Sketch Testing #1
Sketch Testing #1


I usually start the design process with low fidelity sketches. This is the way I iterate through many design options quickly through brainstorming & testing as quick as possible before translating them into wireframes.

Especially given the small timeframe, I felt it was important to sketch, test and reiterate ideas that would or would not work. 

Test #1: Mimic the pickup time button on the left or have its own section on the right?


  • 5/5 preferred the small button
  • having its own section felt like it was just as important as selecting which type of ride 
  • less intrusive & more comfortable

Sketch Testing #2
Sketch Testing #2

I wanted to keep close to the branding of the existing app and an essential part of a company's brand is their language. I felt conflicted between "conversation" and "chattiness" since both invoke different feelings even though they may be synonymous with each other. 

Test #2: Language usage & users' feelings towards tone of voice


  • 4/5 felt "conversation" was more appropriate for Uber's tone of voice
  • 1/5 felt "chattiness" was more inviting & friendly 

Sketch Testing #3
Sketch Testing #3

Deciding how many buttons to express the levels of conversation proved to be more important than what I had initially assumed. The amount of buttons & the type of words used played a big hand in how comfortable the users felt using the feature.

Test #3: Is there a difference & preference for having 2 vs. 3 buttons?

Test #4: Language choice between "quiet type" and "quiet mood" 


  • the concern for the driver was the most important apparent in this testing
  • 5/5 preferred "quiet mood" as it implied a temporary status instead of a permanent characteristic
  • 3/5 preferred just 2 buttons as the third felt intrusive to the driver forcing them to have a conversation they might not want to have

Low-fidelity wireframe
Low-fidelity wireframe

Wireframes & Prototyping

I almost always create wireframes in Sketch and make them interactive with InVision invoking a more "real" feeling when users test the prototype. I've discovered that the feedback at this stage can be more rich in detail whether or not something works. 

Test: Did the location of the button make a difference in the existing user flow of the app? How so?


  • 5/5 preferred to have the button located on the lefthand side of the screen
  • felt more "natural" since most users read the screen left to right
  • thought this feature would be used more than the pickup time feature

Add a Feature: Setting the conversation level with your Uber driver

UI Design

I strictly adhered to the 2018 Uber design guidelines as the project was executed before the 2018 rebranding. I kept consistent with the color, font and wording all the way down to the icon size and spacing & padding used throughout the app.

Feature Button Design

Creating the buttons and adhering to the style was the most challenging in terms of visual design. I wanted to keep aligned with the company and make the buttons seamless so I searched for some Uber designs and constructed the buttons you see to the left. 

I also left the default option to "Open" as that was the natural state of rides are these days: you're open to conversation based on the vibe of the driver and the ride.


You can also see some copy I had written in keeping aligned with Uber's tone of voice used throughout the app. Uber also utilizes copy when describing the type of rides riders can choose from. 

Add a Feature: Setting the conversation level with your Uber driver

Request Button

Keeping consistent was key as I did not want to break the existing interface when it came to ordering a UberX and selecting a desired pickup time. 

Again, I specifically chose this user flow as it illustrated how the request button would change in more complex user flows. 

Final Prototype

*It's important to keep in mind that the user flow for the prototype was created in the point of view of persona #2 (selecting a "Quiet" conversation level and then a scheduled pickup time from her work).

Further Explorations

Button Interface

Currently, the button automatically converts to one single bar when the rider requests a pickup time, making it impossible to cancel or edit the pickup option. You have to either cancel the ride and restart the entire process or request the ride and go into your profile to cancel it and then restart a new request process.

Many of the users were confused at this option as they wondered if they can go back and change the preferred conversation level, but I decided to follow the current user flow and relay back to the product team to revisit this interface (if I were a part of the real product team). 

Moving the conversation level to the user profile section

Another option that arose through the project was the option of moving the conversation level selection to the profile page where it could be defaulted to one state or the other.

As I had considered this option, I wanted to avoid making the conversation selection a static item for majority of the users out there who would also like to quickly opt for a quiet mood depending on their context. 

I would have been interested in A/B testing out the two placements or seeing how much usage the feature was getting and properly allocate its placement based on gathered data.

PR campaign in social disability awareness & support

I thought this could be an excellent win-win for all parties involved: users with social disabilities, "average" users and Uber as a company. Through research, the majority of "average" users felt a sense of relief and newly found support for the feature when more awareness was created concerning those with social disabilities. I also thought Uber could do with a positive PR campaign in helping them with their brand image & reputation due to past controversies & blunders.


Given the short timeframe, there was a lot more I wished I could have done. 


I would have loved to do more testing with the language. The wording is the most pivotal tool when it comes to communicating with the users throughout the interaction of the app and I would have loved the opportunity to further refine the proper wording to further invite the usage of the feature since many users were concerned with how it would affect the driver. I think it would have made users feel more safe when using the feature knowing it would not have any negative effects on other parties. 

Staying user-centered

The most important thing I learned throughout this implementation of the feature was the crucial role of usability testings. User-centered design really has to do with what the user’s actually want and not what you assume that they would want. I found incredibly powerful information from the feedback that either confirmed my assumptions or redirected my doubts to ideas that were more powerful in achieving the objective of the app. I found that listening to their feedback as well as digging into their explanations proved to be the most fruitful of actions that could be taken through designing something that is of real functional use for the public.