
VOICELETS
Enriching the new generation of listeners with
A fresh way to emote uniquely and making audio content more accessible


Overview
Amazon Music is a music streaming service that offers a large library of songs and podcasts. It's available on many devices, including Amazon Echo, Fire TV, etc.
Amazon partnered with Pratt to conduct a hackathon to innovate upon the current state of the product and how AI can be integrated to enhance the experience which is where this project was born.
DURATION : 4 weeks
PROJECT TEAM : Shivani Kolte and Anne Kuo

CLIENT : AmazonMusic
RESEARCH METHODS : Primary + secondary Research, Product study, Consumer Analysis, competitive analysis
My Role
-
Ideation and conceptualisation based on market gap and prduct feasibility for the company
-
Product design sprint and development strategy
-
Prototyping and User Testing
The Ask
Given the current state of amazon music, how might AI enhance customer experience with music and/or podcasts
Our Solution
What if you could..
Share music clips like you send stickers?
AND Edit and Share bits of podcasts the same way...


(fig 1) Voicelet (Music) in Chat
(fig 2) Voicelet (Podcasts) in Chat
How it works
VOICELETS
are shareable audio stickers of customisable sections from the songs and podcasts you’re listening to.
For Music
AI Integration :
Songs
Lyrics
Segmentation + Pattern Recognition
Identification and categorisation trending section
Voicelet

For Podcasts
AI Integration :
Podcasts
Transcript
Segmentation + Pattern Recognition
Automatic Speech Recognition (speech to text)
Voicelet

Our Approach
Through preliminary research and a kickoff meeting with our Amazon mentor, we identified the requirements and outlined the product priorities, establishing guiding principles for our strategic approach as follows :

HIGH QUALITY LISTENING EXPERIENCE
The Amazon brand which caters to a-z needs of its users, it expects to find a way to deliver an inclusive and high quality listening experience

ENCOURAGING FANDOM SUPPORTING CONNECTION
To empower fandoms and leverage their connection to uplift artist through strong community connections
RESEARCH
Striking a chord with listeners
" Cumbersome to annotate for later use."
Understanding their interaction behaviours and consumption priorities with audio content
" Sometimes I'll stop listening halfway through and then the next time, I'll forget the previous content."
" I have to come out of that experience to share it and people don't even open the links

Provide a way for listeners to share and connect through common musical taste and choices uniquely
Key Opportunities

Enhance the accessibility of audio content to include a more representative spectrum of the user set at every stage in their music app journey

Improve the traceability of audio content to make it more browsable
102
Survey Responses
12
Interviews
5
Competitors Analysed
HOW MIGHT WE
Enrich listeners with a tool that makes their experience more accessible, easily and truly indicative of their feelings?
HOW WE APPLIED THE INSIGHTS
Matching their beats, sampling arrangements
Conceptualising based on the insights to inform a working solution in the form of a testable prototype.
DESIGN PRINCIPLES

Accessibility
The idea originated from the need for transcriptions, as many podcasts lacked them, causing listeners to lose engagement due to difficulty in following the content.

Discovery
A common issue was resuming audio content, especially during passive activities like commuting or chores. Transcribing and sectioning audio content could help listeners remember and navigate better.

Social and Community
To foster engagement and fandom community, we introduced shareable audio bits. This concept was inspired by the popularity of static and moving stickers among younger audiences, leading to the creation of VOICELETS.
Integrating voicelets into the customer journey to validate the need

(fig 3) Voicelet Saving and Sharing journey

Working out the feature and the user flow to strategise the visual design according to the Amazon Music information architecture

(fig 4) Voicelet (Music) in Chat

(fig 5) Sectioning and saving audio content

(fig 6) Sharing music in chats
We studied chat interactions and saving and sharing patterns across various apps to understand the existing mental models.

(fig 7) Competitive analysis
Prototyping variations of the voicelets look and experience based on the research

(fig 8) Prototyping efforts
HOW WE ITERATED TO SUIT OUR USER NEEDS
Debuting the prototype and facing the music
User testing to iterate based on feedback and deliver a sound solution.
Adding Signifiers to the preview
Before
- The ‘X’ colliding with the ‘play’ cta in the chat interface.
- Sending it without previewing made users feel less in control

After

Giving users more security by providing a clear “back” button and preview window.
Explore Voicelets
After multiple iterations and 5 user testing session :
01
Sharing Voicelets
For Music


You find lyrics that speak to you
Easily edit and share it with friends via chosen social media platform

Play the song you vibe with

Find the lyrics you resonate with you

Find curated and trending voicelets suggested by Amazon Music

Share them with one click across social platforms

Even in chat, express yourself uniquely
02
Saving Voicelets
For Podcasts


You find sections of podcasts you’d like to save
Easily edit and save it to the library.

Browse through podcast before playing it.

Play the sectioned and labelled podacast

Follow along with transcripts

Save the specific sections that resonate with you

easily find them in the saved section
03
Voicelets Extended
For Music and Podcasts

Saving podcast voicelets and browsing through them in a breeze


The s'aved' section gives an easily accessible window into all your saved voicelets

Pull up any voicelet within the chat experience

Playable song voicelets available in chat drawers which can be pressed to send like a sticker (but with sound!)
Integration with Amazon Ecosystem
Voicelets with Alexa
I wanna be able to save it while I cook or clean, will just saying it do it?
"
"

Conclusion
The Voicelets project revolutionizes how users interact with and share audio content. By allowing the creation of customizable audio snippets from music and podcasts, Voicelets address key user needs for more expressive communication, better accessibility, and seamless sharing.
Our thorough design process, from research to testing, led to a user-centric solution that integrates smoothly with Amazon Music. Voicelets enhance the listening experience, fostering community and connection by making audio sharing more personal and engaging.
Work for the future
Sectioned commenting
Ability to comment in sections within the transcript to socialise and share thoughts.
Personalised Voicelet reactions
Expanding the scope of voicelets' way of expressing uniquely to personalised audio stickers of yourself.
Challenges
-
Conducting InDirect User Research - Setting Amazon Music apart
We encountered difficulties in obtaining direct user research due to the low user base of the app in a reachable network. Consequently, our research was primarily based on listening behaviors and parallel usage patterns, providing us specific insights accordingly. This approach gave us a good idea of the competition and how to leverage Amazon Music's values and set it apart.
-
Naming of the new feature
Through limited user testing we were able to get feedback on the name of the feature which was very enthusiastically accepted for denoting the short form nature of the feature, but there were confusions about the word voice which (off the bat) made it sound like users own voice would be used
Delightful Discoveries
-
The love of and NEED for stickers
It was very validating finding out that the audience we were targetting found stickers as such a crutch in communicating and expressing better through the digital medium as it can be difficult to perceive correctly - the way someone has said something. It meant that with our idea we were a step closer to connecting people with each other even more strongly.