The main challenge of this project was that Google Home and voice UI are new technologies and few best practices have been established. During the workshop, we had to create a whole new process for working with this technology. In addition to our lack of experience working with VUI, we had to assume that our user is even less likely to know how to interact with this technology. Because of this, our idea needed to be clear and make sense to all users, despite their lack of familiarity with Google Home and voice technology.
Because knew our user would not be used to using voice UI, we shaped our concept around scenarios where users would be more inclined to use something that was hands-free. This mental model shaped our entire approach to product development.
All the while, we were discovering our workflow for projects like this. After identifying our user, we created a persona for our bot since we needed to shape its vocabulary and conversations. We sought to create a bot persona that felt efficient, easy to use and trustworthy.
After establishing the bot’s persona, we sketched out the flow of a conversation between the user and Mood Mixer. The developers and I wrote out all of the variables and possibilities of the conversation. We wrote “repairs” for conversations (fail states) to help the user with Mood Mixer since there is not persistent help or UI. While the developers were integrating with Spotify’s API, I wrote the full dialogue for the bot and started to identify multiple keywords that could trigger the same response.
Assistant Technical Director - Arnaud Tanielian
UX Designer - Liz Wells
Developer - Tim Roussilhe