Alexa Allegro Skill
Allegro is a conversational music interface (VUI) that suggests classical music based on the user's mood, desired state of mind, or current activity. Designed as a prototype Amazon Alexa Skill, users will be asked to voice their mood or current activity to their Alexa device.
Users craft their personalized music discovery journey, using responses like “I’m feeling sad” or “Play something to help me focus.” After each exchange, Allegro recommends a relevant classical music piece. Allegro is meant to inspire better mental health and curiosity for classical music.
ROLE
Conversational Designer
UX Researcher
TEAM
Angela Wang
DURATION
4-Week Remote Design Sprint
TOOLS
Voiceflow
Alexa Developer Console
Lucid Chart
D E M O
Allegro, I'm feeling
Happy
Sad
Excited
Music to my ears, if I had any. Wanna hear something lively?
01.
T H E G O A L
Allegro was created to break negative stigmas surrounding classical music and inspire people who would not otherwise be interested in the genre to become curious about it. Allegro makes exploring classical music approachable through conversations that are based on users’ current mood. Therefore, users relate to music more since the music is picked based on their present emotions or activity.
02.
R E S E A R C H
To begin building Allegro, we focused on understanding current motivations for new classical music listeners and their challenges. Additionally, we researching competitors to help Allegro differentiate itself from existing market competitors like Pandora. For these reasons, we decided it would be best to conduct user surveys and competitive analysis.
35 participants took part in an anonymous questionnaire hosted on Google Forms. The questionnaire contained a total of 16 open-ended & multiple-choice questions. Our questions were designed to learn more about people's experiences listening to classical music and the correlation between classical music and mood. Ultimately, our goal was to find the habits and motivations for current and new classical music listeners.
Below are our findings from both research methods:
Findings
Participants typically listen to classical music while performing specific activities: 31.4% of participants listened to classical music when studying. Users also prefer to use classical pieces as “background music,” especially when they need to be hands-free.
Participants’ mood is sometimes affected by the music they listen to. 65.7% of participants acknowledged classical music having some impact on their mood.
Participants listen to classical music on their desktop or mobile: 85.7% of participants listen to music on mobile phones. 62.9% listen to music on their desktop.
The tempo is not the only factor defining a piece’s mood or impact on listeners: 54.3% of participants said they listen to slow-tempo music when they are down. 37.1% also stated they listened to slow-tempo music when stressed. Based on results & user interviews, we conducted more academic research to build a more accurate song selection based on additional music factors. (e.g., Intensity is measured by decibel & pitch by hertz)
03.
P E R S O N A L I T Y
Our VUI is leaning towards casual, warm, and calm. We chose these tones because we learned from our survey participants that most people listen to classical music when they are engaged in an activity, when they are in a particular mood or want to feel a certain way.
04.
D I A L O G F L O W & S A M P L E S C R I P T S
Prior to creating a sample script, we developed a dialog flow to walkthrough how users would interact with Allegro. During this development, we paid attention to not only focus on happy paths or new users but also to accommodate for returning users or instances where Allegro failed to understand the user. Likewise, we finalized our data sets.
The dialog flow as well as a small preview of our data set is presented below:
Sample Scripts
HAPPY pATH
USER
Alexa, open Allegro
VUI
Welcome to Allegro! Let’s discover a new classical piece together! I can make recommendations based on your current mood or activity. How should we begin?
USER
Mood
VUI
Alright, l can make music recommendations based on your current mood. How are you feeling right now?
USER
I’m feeling sad
VUI
Sorry to hear that. Classical music can be therapeutic and help lift your spirits. Wanna hear something calming?
USER
Sure
VUI
Okay, I found {%relaxingSong}. Here’s a preview. Want to continue listening?
USER
Yes
VUI
I hope it helps brighten your day. Be sure to check back in for a new piece anytime!
We conducted 2 rounds of usability testing. For the first round of low-fidelity testing, we recruited three participants to review our scripts for new and returning users. To give us feedback on aspects such as their impression of the personality of the VUI or confusing interactions. For high-fidelity testing, an additional three participants user test helped us discover new utterances, check for usability issues, and dialog confusion. Overall, participants found Allegro to be friendly and fun to use. However, we did need to make changes regarding users' control over the audio preview, adding angry emotions to our data, and including a re-prompt if they missed the chance to respond
Hover over the images below to preview two of our scripts for new users:
unhappy pATH
USER
Alexa, open Allegro
VUI
Welcome to Allegro! Let’s discover a new classical piece together! I can make recommendations based on your current mood or activity. How should we begin?
USER
Mood
VUI
Alright, l can make music recommendations based on your current mood. How are you feeling right now?
USER
I’m feeling optimistic!
VUI
Music to my ears, if I had any.
Wanna hear something lively?
USER
Not really
VUI
Alright, let’s find you a piece in another way. You can say things like “play something to help me relax” or “hear something soothing.”
USER
Play something to help me focus
VUI
Okay, I found {%relaxingSong}. Here’s a preview. Want to continue listening?
USER
Yeah, I like this
VUI
Great, enjoy!
C O N C L U S I O N
Given the time constraints and technical limitations such as the audio length limits, we were unable to fully release the skill. However, it is in beta-testing in my Alexa Developer Console and we hope to develop it further using an API. Likewise, our participants also gave us valuable insights into how we can further improve Allegro’s performance in the future. We will consider adding these suggestions for our next iteration of Allegro.
-
Consider adding different welcome back messages. Unique welcome messages will make the experience more interesting for users who use the VUI multiple times a day.
-
Account for different user types. There may be two paths: one for classical music experts and one for new listeners. Classical music experts may not expect suggestions or previews, but they may want to listen to music similar to the ones in their current playlist. They may also want to create a playlist to "save" the songs they liked. For new listeners, it might be a good idea to add some background information on the pieces to make Allegro a bit more educational.
