UX RESEARCH
UX DESIGN
VISUAL DESIGN
Discord is a popular communication platform that provides voice, video, and text chat features for communities to join a server and interact with each other.
However, servers are vulnerable to attacks from bad actors creating a hostile environment to hurt Discord communities.
Moderation Tickets is a system designed for community members to report bad actors or problems in the Discord server. Once the tickets are sent out, the moderation team can review complaints and reports in order to resolve them.
Results
83% increase in user satisfaction.
100% reported a much easier learning curve compared to other solutions.
CLIENT
Discord
TEAM SIZE
Individual
TIMELINE
DELIVERABLES
Prototype
DELIVERABLES
PROBLEM DISCOVERY
One day, my friend woke up and saw a message from Discord that would make anyone’s heart stop. Bad actors had posted illegal content then buried the post with spam. After that, they reported the server for hosting illegal content and the server was deleted.
The community, moderators, and bots failed to detect the post. This got me thinking: Is the current server moderation tools REALLY good enough?
USER RESEARCH
Discovering the Users' Opinions
I conducted a user research with five Discord moderators who previously had issues with server moderation. I wanted to identify the current problems with moderation. Eighty percent of the interviewees stated that current tools were lacking.
80% Users NOT SATISFIED
The vast majority of interviewees were dissatisfied with the current moderation tools.
User Stories
The participants were asked share a story where the current moderation tools provided a negative experience. User stories uncovered the shortcomings of user tools.
Identifying the Frustrations
Analyzing the user stories categorized server moderation tools into two categories: bots and ticketing systems. I sorted the users' pain points into the two categories in order to build the solution later. The most common pain points involved paywalled features and complicated setup.
User Journey
Interviewees cited a big frustration was lack of a singular tool; Moderators sometimes use bots and ticketing system at the same time.
I used a user journey map to determine "locations" to solve the problem. The chart also acted as guide to developing the UX of the user interface.
Insight Formation
I noticed while analyzing the data, that users did not complain about other moderation features such as: word filters, verification levels, and activity alerts. I discovered that most of the complaints pertained to post moderation problems; NOT server moderation.
Post Moderation Problems
• Too many user tools to learn.
• No built-in tool to moderate posts on Discord.
• Paywall aspects.
PROBLEM STATEMENT
How might we integrate a post moderation tool to align with the users' needs?
User interviews show that there is a strong demand for an integrated post moderation tool. Moderators want a free and universal tool used to find, moderate, and document posts.
IDEATION
Deciding on the Solution
Now that I narrowed down my focus, I chose to ideate on ways on how to develop a solution that can help Discord users moderate posts much easier.
Competitive Analysis
I performed a competitive analysis in order to determine the most liked aspects of a particular tool.
Bots
• Enforces rules automatically.
• Reduces the time on task.
Ticketing System
• Information logs with context.
• Notifies entire moderation team.
• Tickets can be discussed among team.
What Doesn't Work
I drafted on a few ideas that might work. However, they were ultimately scrapped as they were not deemed an effective solution.
Built-in Bot
Ticketing System
• Mass report results in frequent audio and visual pings.
• Useless if all moderators are offline.
What can Work
I settled on the idea of using a Ticketing System. Eventually, I took inspiration from IT ticketing systems in order reframe it into a built-in solution. An in-depth look at the categories resulted in ideating ideas to form the solution.
SOLUTION
DEVELOP A BUILT IN TICKETING SYSTEM
Looking back at the cons of the ticketing system; I realized that majority of the problems are capable of being fixed. By integrating a feature specializing on report tickets, we can fix the frustrations our users face.
Before and After
Third Party Ticketing System
• Easy to miss in chat channel UI.
• Vulnerable to spam.
• Complicated to setup.
• Paywalled performance.
Discord Ticket System
• Use existing UI elements.
• Integrate number of reports as a number.
• Integrate as a built-in feature.
• Feature no paywall system.
Before and After
I looked at my previous interview results and used a MoSCoW Matrix in order to create minimum viable product.
Must Have
• Multiple tickets view.
• Context Information.
• Filter for date, priority, ect.
Should Have
• Mass report results in multiple pings.
• Resolved tickets history page.
• Deleted ticket section.
Could Have
• Embedded media automatically censored.
• Name of moderator resolved past tickets.
• Potential for rogue moderator abuse.
Won't Have
• Audio pings per report.
• "Fun" related features similar to other bots.
DESIGN
Designing the Solution
The design phase focused on ideating on ways to deliver the solution and building off the mistakes of current solutions.
UI Ideation
I looked through several of Discord's menus in order to find a suitable user interface. I settled on using the Events panel as a frame of reference, based on the following attributes.
Design Language. The interface closely follows Discord's design language.
Swift Development. Using an existing UI helps developers release the solution more quickly.
Working Space. Plenty of empty space to fit elements and future features.
Low-Fidelity Wireframes
I used Figma to design the Wireframes.
User Flow
While designing the user flow, I referenced the user journey in order to avoid missing critical UX steps.
Revised User Journey
DELIVERY
Delivering the Solution
With our ideation and wireframes now complete, it was now time to design the solution.
Prototype
UI Kit Credits: Muhsin Ataul, bawwub, figr.design.
Usability Studies
I tested the feature with a new set of six Discord moderators. The participants were asked to perform the task of administering moderation through the prototype.
Limitations of Studies
• No onboarding process resulted in a slight learning curve.
• Study only simulates a hypothetical situation.
• Relied primarily on a user's past experience.
Results
Qualitative Results
Quantitative Results
Participant Feedback
I recorded a few notable quotes from my participants for feedback of the application. I documented positive and negative feedback to improve the product in the future.
Participant #1
Participant #3
Participant #5
Key Performance Indicators
I ideated on a few metrics to analyze in order to determine if the feature is a success for both the user and the business.
User Focused KPIs
Time on Task
Analyzing the time on task through the user journey. Measured through Usability Testing.
User Error Rate
The amount of errors a user makes when. Measured through Usability Testing.
Net Promoter Score
Validates the satisfaction of users on a scale of 1 to 10. Measured from User Surveys and User Interviews.
Business Focused KPIs
Retention Rate
Validating the feature increases user retention rate. Retaining users is a potential increase in revenue.
Nitro Sub Rates
Analyzing if a feature boosts user satisfaction and drives Discord Nitro subscriptions for brand support.
User Acquisition Rate
Analyzing if competitors leave their service to use Discord based on the tools we provide for them.
LEARNINGS AND TAKEAWAYS
Reflecting on the Project
Thank you for reading! I learned a lot throughout creating this case study. In addition, there were also a few key takeaways that helped me develop my UX practitioner skills.
Insights
My main learnings through this case study involved developing a better mindset for detecting problems.
Don’t focus on the ending. In my first iterations I did not have a “Past Tickets” or breadcrumbs element. It’s not part of the MVP, but it is good practice to have.
Narrowing the problem. Categorizing problems is important. Initially, I thought there was a moderation issue in general. However, I discovered the real problem was post moderation; it was not regarding filters, verification, ect.
Takeaways
Use the tools interviewees complain. Experience the issues yourself; You’ll remember the problems quickly and avoid it during design.
Inspiration can come from unexpected places. Sometimes, it’s important to look outside the box for solutions.
Account for outside issues. API breaks leads to bots being unusable. I now think about outside factors in a given solution.