After researching and synthesizing my data, I started thinking about possible features and screens for Spotify DJ. Before creating prototypes, designer and I wanted to have a good understanding of how common using the DJ processes work. This is a simple task flow for a user trying to play the DJ.
We made sure to include features that users wanted and easily discoverable.
After finishing our first iteration of DJ prototype on Figma, I have recruited 5 of our previous interviewees from the research phase to complete usability tests on our completed app to find any potential issues. I have used contextual inquiry approach to do my usability test. I created a task list and gave them to the participants as they went through the prototype. They followed TAP (Think Aloud Protocol) which made it easy to follow their line confusion or difficulty whenever it occured
People seem to not care much about controlling for DJ's voice and content
Wasn't too sure by "disliking" a song means it will be registered forever for future use
Wasn't too sure if they will be using certain "stations" such as following station
Based on our participants' feedback, there were few changes to the Spotify DJ:
Only kept the DJ's voice frequency (amount of talking)
Letting user's know that each session is unique
Added clearer tool tip with simple description
Based on the user interviews, surveys, usability tests and more, we were able to improve the DJ: a DJ AI that helps to play the "right" song. With the improved Spotify DJ, users can like/dislike songs, create mix-match with friend(s), easily change to different stations, personalize the DJ's narration, and quickly access to the it.
