Nouman's Portfolio

Twitch - Chat Redesign

Improving Clips engagement by enhancing it's discoverbility.

Back

Full Document

My Role

UI/UX Designer (solo)

Mentors

Prof.Kevin Park

Prof. Leah Smith

Andrea Pacheco

Timeline & Status

4 Months

Overview

This project was my final thesis for the Master’s program in UX/UI Design & Development at New York Institute of Technology.

Twitch is the world’s largest live streaming platform

where millions of people come together every day to

create, watch, and interact in real time.

PROBLEM

"Twitch clips are created every day, but most of them go unseen and unengaged. The platform struggles with low clip discoverability and poor engagement."

SOLUTION

"I redesigned Twitch’s live stream page chat section to integrate real-time, user-generated clip highlights."

By introducing clips in chat as recommendations using the existing clip flow, I made clips more visible, accessible, and engaging.

THE PROCESS

Following UCD Framework

I used a user-centered design approach — starting with research, followed by analysis, design, testing, and iteration to reach the final solution.

DISCOVERY

What problems I Identified?

AVG TWITCH STREAM

3.5 Hours

AVG CLIP VIEWS

CLIPS ENGAGEMENT

Why does this matter?

For long-form live streams, users often miss key moments from their favorite creators.
Catching up afterward is frustrating and usually requires:

Excessive time and effort
Scrubbing through long VODs
Depending on third-party platforms like YouTube, Reddit or X

When users miss moments, Twitch’s clips should be the easiest way to catch up, but they aren’t.
Clips are hard to find in the moment, and this directly impacts their engagement.

IDEATION

Exploring Opportunities

After analyzing research insights and identifying gaps in the current Twitch experience, I conducted a competitive audit to understand how other platforms solve similar problems around content discoverability and user catch-up.

Using those findings as inspiration, I moved into ideation — starting with a Crazy 8s session to rapidly generate ideas. From there, I developed and refined 2 two design directions that were both feasible and effective in solving the core problem.

HYPOTHESIS

"If clips are surfaced contextually within the live stream chat, where users are mostly active, then it will become easier for viewers to discover and engage with clips in real time, ultimately increasing overall clip visibility and reducing user drop-off."

SOLUTIONS

Designing Systems

Based on my ideation and hypothesis, I created 2 design versions for integrating clips into the Twitch live stream page. Both versions aimed to improve discoverability without disrupting the live viewing experience.

TESTING & ITERATION

6/8 Users preferred System 2

To validate both design directions, I conducted A/B testing with 8 participants. Each user interacted with both versions in a controlled test environment and shared their feedback based on clarity, accessibility, and ease of use.

Iteration Based on Feedback

Following the A/B testing, I implemented a key improvement based on user feedback:
many users appreciated seeing clips in chat, but found it difficult to Browse through clip messages in a fast-moving live stream.

CONCLUSION

A better experience for users. A smarter solution for the product.

Through research, testing, and iterative design, I validated that my solution improved usability, reduced cognitive load, and directly addressed the problem users face when trying to discover clips during live streams.

By surfacing clips contextually and making them easier to access in real time, my design not only solved a core user frustration, but also addressed a key business challenge for Twitch — improving clip discoverability and increasing engagement on the platform.