TWITCH:UX CASE STUDY
Project update: Since developing this case study and solutions (as of 06-2019),
Twitch has since adopted similar improvements to their systems.
THE TWITCH CHAT PROBLEM
Twitch chatters shy away from using chat feature because they feel drowned out from spam when they try to speak so it is not quite doing its job effectively.
Streamers struggle keeping up with chat and viewers and don’t feel like there is any interaction due to massive spamming.
PROJECT DESCRIPTION
The proposed solutions will hope to alleviate the major flaws and annoyances of using Twitch’s chat feature so that it will provide the clearest form of communication for streamer and viewers.
USER INTERVIEWS
10 Male and Female mixed ranging from 16-32 years of age who consist of streamers and regular viewers of Twitch.
Conducted a survey to gain insights which were synthesized in Trello.
PERSONAS
The proposed solutions to test were based on two personas. Winston the streamer and Brandon the viewer /chatter.
INFORMATION ARCHITECHTURE
SOLUTION 1: EMOTE OVERLAY
Based on first principles, I assumed that cleaning up the abundance of emotes in chat could help separate legitimate interaction and communication versus the hive mind aggregate of reactions being portrayed.
SOLUTION 2: CHAT THROTTLING
Another proposed improvement that was hypothesized was the throttling of chat speed and the amount of messages chatters would be able to enter repeatedly. By limiting the spamming to a timeout each time messages are being copy and pasted in the message box, this would control the visual traffic and help reduce visual chaos.
PROTOTYPING AND WIRES
I began to build out working inVision prototypes of the proposed chat improvements to begin user testing.
Testing Results and Learning Outcomes
What happened when users chat throttled? What does that interaction look like? Did it improve the experience of chatting?
Overall, most users tested with positive results and they felt that this improved the app.
LOOKING AHEAD FOR FURTHER ITERATION
I concluded that the value of understanding users pain points / habits and needs from using Twitch is paramount to making the right call and improvements to the app.
Testing was invaluable. It made sure that I was on the right track with the assumptions and the data and results from the test proved it as a successful change to be implemented.
Also, one for the memory bank is that people are more willing to accept a problem and avoid it rather than be vocal how bad something is. I would have also liked to play with the UI more especially figuring out details how emotes would work in conjunction with text and see if there are any alternatives there.