Zoom Redesign Project

UC San Diego | 2023 | UI/UX Design, UX Research, Problem-Solving

About

For a course project at UC San Diego, our team analyzed and redesigned the Zoom platform to enhance user experience through data-driven insights. We conducted user interviews and gathered quantitative data to identify design flaws, usability issues, and common user errors. By analyzing trade-offs between aesthetics vs. functionality and convenience vs. multi-featured design, we explored solutions using design space charts. Our redesign aimed to improve feature visibility, streamline navigation, and reduce cognitive overload while maintaining Zoom’s core functionality.

Methodology

A) Brainstorming

During our brainstorming period, we used a mind map to decide which video communication platform to choose. We chose Zoom because many people have used Zoom before, due to the pandemic, or are currently in use. Another reason why we chose Zoom is that despite the several convenient functions for users, there are still functions that lead to user errors even if you have used the platform often.

B) Interview Methodology

We focused our demographic on young adult students who frequently use the Zoom platform for education and recreational purposes. Some participants include classmates, friends, and coworkers. Our interview consisted of three parts: pre-task questions, task questions, and post-task questions. All interview questions are followed by the master-apprentice model collected in a single document. The interview took place in person and online. We asked the participants to create and join a Zoom meeting individually and follow the given task question accordingly. Each interview question was to collect qualitative and/or quantitative data from the interaction between the participants and the Zoom platform.

1/ Pre-Task Questions

Demographics: Name, age, education, hometown, occupation, and consent.

Do you use Zoom?
- Goal: To understand if the interviewee has any prior experience with Zoom.

What do you use Zoom for?
- Goal: Bring up the interviewee's prior experiences and knowledge while using the app for specific purposes. This helps us understand the user's current mental model and knowledge in the head.

How familiar or unfamiliar are you with Zoom? (1. Not at all familiar 2. Somewhat unfamiliar 3. Neither familiar nor unfamiliar 4. Somewhat familiar 5. Extremely Familiar)
- Goal: To understand the subject's familiarity with Zoom. Assuming the participant makes an error while completing the tasks, there is a higher chance that the error would be a knowledge-based mistake. If the user shows a higher familiarity, we can infer that the participant made an action-based slip rather than a mistake because they would have a better mental model and thus it would be easier for them to map functions.

How often do you use Zoom? (1. Used once 2. Less than once a week 3. 1-3 times a week 4. More than 3 times a week 4. Daily)

(If used Zoom pretty often) What prior good or bad experiences have you had with Zoom before?
- Goal: To understand the user’s prior experience with any possible knowledge-based slips or mistakes because they might have less knowledge in the head.

What other video meeting platforms do you use?

2/ Task Questions

Unmute/Mute mic
- Goal: To see if there are clear signifiers for the user to perceive to mute and unmute when they are using zoom. It also demonstrates whether or not the feedback is noticeable enough for the user to know if they have successfully unmuted themselves.
  - Say you want to unmute and talk. How do you attempt to do this?
  - Say you want to mute after talking. How do you attempt to do this?

Sending a message
- Goal: To find out whether or not there are enough perceivable signifiers that tell the users where and how to send a message in zoom. It can reveal potential errors like sending messages in the wrong chat.
  - Say you wanted to send a message to the general meeting. How would you attempt to do this?

Sharing your screen
- Goal: To examine the discoverability of the signifiers that inform users where to share the screen and the effectiveness of the feedback that tells the users whether or not they have shared the screen with sound. Together, we can picture a system image that shows us if the gulf of evaluation is bridged.
  - Say you want to share your screen. How would you attempt to do this?
  - Say you want to share your screen with sound. How would you attempt to do this?

Using different buttons
- Goal: To see if the product has good affordances and clear signifiers for the users to know where and how to make breakout rooms, reactions, and changes in the screen layout.
  - How would you attempt to make breakout rooms?
  - How would you attempt to make a reaction?
  - How would you attempt to change the view layout of your screen?

3/ Post-Task Questions

What was your experience with doing the tasks above?
- Goal: To collect quantitative data based on the participant’s overall performance with the tasks.

On a scale of 1-10, with 1 being extremely difficult and 10 being extremely easy, how difficult or easy were those tasks? Can you expand on your response?
- Goal: To collect quantitative data based on the participant’s overall experience. The rating also demonstrates how well-designed Zoom’s user interface is. The second question helps identify specific aspects of the platforms and tasks such as mistakes/slips, knowledge in the head, and knowledge in the world.

On a scale of 1-10, with 1 being highly unlikely to 10 being highly likely, how unlikely or likely are you to use Zoom again?
- Goal: To collect quantitative data on the interviewees' satisfaction with their experience of using Zoom.

If there is anything you would like to add or change, what would it be?
- Goal: To identify any potential solutions, feedback, and improvements needed based on personal user experience.

Collected Interview Responses

Error Analysis

1) Difficulties Sharing Screen with Sound

2) Difficulties Creating Breakout Rooms

3) Difficulties Changing View Layout of the Screen

A) Knowledge-Based Mistake

Knowledge-based mistake: this type of error happens when a person is doing an unfamiliar task, using new equipment, or facing a novel situation. According to our data analysis, knowledge-based mistakes were made in the following three categories:

1. Task 3.2: Sharing the screen with sound

A total of 8 interviewees exhibited unfamiliarity or a complete lack of knowledge of the function of sharing the screen with sound. Two interviewees found the solution after several attempts but the rest never figured it out.

2. Task 4.1: Making breakout rooms

A total of 11 interviewees stated that they did not know how to make a breakout room. Four interviewees had given up on trying and the rest still failed after several blind attempts.

3. Task 4.3: Changing the screen layout

A total of 3 interviewees were found to lack experience with the function of changing the screen layout.

We found that the causes for the knowledge-based mistakes are:

Lack of signifiers or signifiers is too hard to perceive
Difference between the designer’s conceptual model and the user’s mental model

Unlike other tasks where the interviewees can easily find a “share screen” button or “make a reaction” button, there is no clear signifier that tells the interviewees where to share the sound or make breakout rooms. The product cannot communicate with the user without a proper signifier, thus, the gulf of execution is not bridged and leads to a situation where the user does not know what to do. As for the “changing screen layout” task, the “view” button is too small to be noticed in the top right corner unless the user has previous knowledge of it. There is also a difference between the conceptual model and the user's mental model because the designer might think a small check box would be enough for the user to notice under the sharing screen window, but in fact, it is not. Possible solutions can be making the signifiers more perceivable or adding a procedural manual to the menu.

B) Rule-Based Mistake

Rule-based mistake: this type of error is hard to be detected and avoided because the situation is often complex, with multiple possible causes and a variety of reasons leading to this error. The most common cause would be invoking the wrong rule due to an incorrect evaluation of the current situation. According to our data analysis, rule-based mistakes were made in the following situations:

1. Task 3.2: Sharing the screen with sound

One interviewee continued to try sharing the screen (and with sound) while the screen-sharing function is disabled.
One interviewee attempted to share the computer sound by just sharing a video with sounds on.
One interviewee stated that they had no idea whether or not the computer sound was shared or not.

2. Task 4.1: Making breakout rooms

Interviewees 5 - 11 had tried a variety of wrong actions in attempting to create breakout rooms.

We have found that the causes for rule-based mistakes are:

Bad mapping
Unclear signifier
Not enough information presented

Most of our interviewees made rule-based mistakes due to their wrong mental model of how all the Zoom buttons work. To put it in other words, they have to guess some functions (of each button) due to their knowledge deficits, and those assumptions are oftentimes wrong evaluations of the system. Possible solutions could present the information of the current state more clearly, like showing whether or not the user is sharing the screen with sound. Additionally, make better signifiers and mapping.

C) Description-Similarity Slip

Description-similarity slip: this is a type of action-based slip, which occurs when the right action is acted on the wrong object. The users would have the right mental model, yet the similarities between the wrong and the right object would often confuse the users which causes them to act on the wrong object. During our interview process, we discovered some description-similarity slips in the following situation:

1. Task 3.2: Sharing the screen with sound

Two interviewees clicked on the “mute/unmute” button, attempting to share the computer sound because it looks like a microphone.

2. Task 4.3: Changing the screen layout

One interviewee clicked on the “view” on the menu bar (on the Mac system) instead of the “view” button in the top right corner of the zoom interface.

We have found that the causes for description-similarity slip are:

Similar icons with different purposes, lack of a precise description (unclear signifier)
Bad mapping

There are many functions in Zoom, such as, mute/unmute, screen sharing, turning on/off video, creating breakout rooms, etc. Therefore, it requires a lot of signifiers or constraints for the users to take proper actions to achieve the desired goal. The design of the signifiers needs to be distinguishable and easy to read, otherwise, it would be like using a remote control with similar buttons, making them act on the wrong object. As for the screen-sharing task, users may click on the mute button for sharing computer sound because there is no distinguishable signifier for computer sound sharing. A possible solution for this error would be creating more unique signifiers or descriptions and mapping them to bridge the gulf of execution.

This is a quantitative bar graph that displays the common errors made by participants with their given tasks, according to the interview data. Rule-based mistakes and description-similarity slips were recorded the most due to misunderstood task questions and several similar Zoom features.

This is a quantitative bar graph that shows the common trends of participants based on the interview data. All 15 participants rated their familiarity with the Zoom platform as 3.5 out of 5 or higher.

Design Space & Redesign

Design Space

We demonstrated the tradeoff between the aesthetics and functionality of Zoom and other comparable apps, including our redesigned app, through charts and graphs.

Aesthetics vs. Functionality

Zoom is a highly functional platform, offering screen sharing, a whiteboard, breakout rooms, polls, and more. However, its cluttered interface can overwhelm users. While icons help identify actions, seven out of nineteen interviewees struggled with knowledge-based mistakes, failing to execute functions due to an overload of options. A more visually streamlined design could improve usability by prioritizing essential functions.

Convenience vs. Multi-Featured

Zoom includes a wide range of features, similar to competitors like Discord, FaceTime, Google Meet, and Skype. While these additions aim to attract more users, they can also cause confusion, making it harder for users to remember each function. This complexity contributes to knowledge-based mistakes, as users expect video platforms to focus primarily on communication.

Redesign Prototype

Redesign of the user interface of the Zoom meeting. Two tabs of the Zoom application are merged into one, along with other changes to the overall layout for better organization.

Justification

1) Basic features of the Zoom meeting

Features that Zoom users use the most, such as muting/unmuting their voice in the meeting, primary video settings, screen-sharing, recording meetings, and sending chat messages, are located at the bottom of the screen, much like the original design of the meeting layout.
The “End Meeting” button is enlarged and placed at the center of the button tab to locate the button more easily at the end of the meeting. Originally, the button was located at the bottom right corner of the tab and was considerably small in size.

2) Advanced and other miscellaneous features

In the original design of the application, features outside of the meeting are located on a separate tab (in Desktop mode). For advanced features of the Zoom platform, you will have to log in to your account on the Zoom website. To prevent missing signifiers, the redesign merges all miscellaneous and advanced features into a side tab along with the ongoing meeting. The redesign would also have a feature where users can “Hide” unwanted features on their tabs, keeping them well-organized.

3) Additional features

Additional features at the top of the meeting screen include the number of participants, the number of screen-sharers, and the number of viewers of the screen share, which differs based on the individual sharing their screen.