Welbie: A one of a kind voice-activated emotional intelligence companion for children.


Research, UX, VUI, Copywriting, Motion, Branding


Carly Stilson
Maya Suzumura


9 Weeks


Voiceflow, Figma, Illustrator, After Effects, Premiere Pro, Notion

The Problem

Parents, caregivers and teachers want to provide children with the knowledge of emotional regulation. Caregivers can not always be available during the times of crisis when it’s most important to intervene and provide a learning moment. While these skills are not always accessible in homes and schools they are vital for children to develop independence and resilience in their adult lives.

The Solution

A voice activated companion for children that provides users with entertainment and emotional regulation skill building.


Research Outline

Going into research it was important to set some goals for discovery. Having a clear understanding of best practices for voice user interfaces and learning models for children helped to create guardrails for future design decisions.



Research and build understanding on current learning models for emotional regulation.


Uncover trends in how parents handle teaching emotional regulation to their children.


Conduct a competitive analysis to assess what similar toys are currently on the market.


Understand if parents find value in using voice user interface products with their children.

Learning Models

Research began with understanding current learning models geared towards emotional regulation. After learning about a wide variety of approaches two main models were honed in on that would act as the foundation for Welbie.

Ruler learning model graphic.
DBT model graphic.

Why these models?

Both of these learning models emphasize the importance of not judging emotions, learning mindfulness, and building skills that can lead to expressing big emotions in a more manageable way. Both the DBT and RULER models are filled with helpful tools and activities that can apply directly to a VUI flow. Empowering parents and children to make learning emotional regulation a more self-guided process.

Parental Survey

With these models in mind research continued with a survey of several parents. The goal being to uncover what skills they use to build emotional independence in their child, and what role a VUI might play in that.

The results showed a lot of emphasis on independence being important. However it’s difficult for parents to balance growing independence while still supporting their child in times of emotional distress.

Key Takeaways


Nurturing independence in children is important to parents.


It is difficult to balance nurturing independence with offering support in times of distress.


There are some concerns regarding letting children use a voice user interface.


Parents are already employing emotional regulation skills in their homes.

Parental survey graphic.
“I think voice interaction with a children's toy is ideal. Kids are able to gain confidence and practice speech while learning vocabulary and language structures.”

Persona Pair

The research conducted so far led to a persona pair of mother and daughter that highlights two key perspectives when it comes to learning and teaching emotional regulation.

A parent persona for Welbie.
A child persona for Welbie.


Jobs to be Done

Based on the insights discovered during research, the JTBD framework helped to develop those findings into possible goals and reasons that a user might "hire" a specific product. The "jobs to be done" statements that we developed are based on our persona pair that highlights the two perspectives of parent and child.

A JTBD for the parent.
A JTBD for the child.

MSCW Analysis

With these reasons for hire in mind, we were able to start ideating some possible features for the VUI toy. We did this by creating must have, should have, could have, and won't have columns where we then conducted a post up and sorted potential features into their appropriate columns.

A must have, should have, could have, wont have analysis.

Business Model Canvas

With a clearer vision of our direction for the voice user interface we decided to move on to a business model canvas. This framework pushed us to explore key revenue streams, partners and other important metrics that would focus our work and provide us a structure moving forward.

Key Partners

Credible support from an academic source (Yale Center for Emotional Intelligence).

Seattle Children's Hospital & Seattle Public School systems.

National Association for the Education of Young Children.


Key Activities

Building trust with our customers through a consumer-facing ad campaign.

Collecting credible endorsements from institutions focused on emotional wellbeing for children.

Providing schools and hospitals with our VUI to begin getting exposure and feedback from real users.

Key Resources

Extensive research on childhood emotional development and learning.

Marketing and ad campaigns that will build trust with our users and develop brand recognition.

Credible endorsements from institutions that focus on social and emotional development in children.

Key Propositions

Helping your child build a tool kit for emotional regulation.

Making play time a learning opportunity for young children.

Offering children an opportunity to learn skills independently and gain confidence in their ability to manage their emotions.


Visit hospitals and schools and provide the product in order to get it in the hands of our users without them having to pay for it.

Recommendations from therapists, teachers, and doctors.

Advertising through Facebook, TikTok and television so that we can showcase the product and educate our audience.

Revenue Streams

Outfits and skins that offer users a different personality to their VUI.

Subscription to a platform that offers parents a community of users and resources and extra subscription-only skins, and add-ons.

Additional physical toys and downloadable content.


System Persona

The system persona outlines the goal of our VUI and the overall tone and energy that it will offer our user depending on the context of the interaction.

In order to come to this persona we explored a range of adjectives that felt either in line or out of line with the needs of our user. Then we mapped our system's tone of voice in context specific scenarios with a personality map.

System persona for Welbie.

The Happy Path

We created our first happy path based on a scenario for our child persona Ella. The scenario for our happy path is that Ella is starting to feel overwhelmed and is looking for a way to calm down. This flow was a rudimentary exploration of what the interaction could look like in a best case scenario.

The Welbie happy path.


With the happy path laid out in Miro we conducted some research and found a tool that would offer us prototyping capabilities for a VUI toy called Voiceflow. We replicated our happy path in Voiceflow and fleshed it out further. This brought life to our system and gave us the chance to begin user testing our happy path.


We created two distinct mood-boards for the brand. After receiving feedback on these mood-boards this is the chosen direction that would later be applied to our logo, prototype, and consumer facing explainer video. We felt it offered a warmth and playfulness that was in line with our users and our system persona.

The Welbie Moodboard

Usability Testing

Wizard of Oz Testing

Testing our Voiceflow prototype was an important part of the design process. While we were only able to test Welbie's happy path on an adult we were still able to gain some valuable insights and validations.

Key Takeaways


The no-match response is working and felt light-hearted and playful to the participant.


The overal tone and energy of Welbie was well recieved by the user.


The breathing exercise felt a touch long, and possibly not engaging enough for a child.


Consumer Facing Video

The final deliverable for this project was a consumer facing explainer video that outlines the core features of Welbie and builds trust between the product and the user.


The Result

This project was a great opportunity to learn about voice user interface design, designing a system persona, and developing a product around user needs. We had a challenging time getting the product tested with children which really was a limitation for creating a truly user oriented experience and something we would prioritize more if we were to do it again.

Next Steps

As far as next steps go we would greatly benefit from getting our prototype in the hands of some children within the age range of our target user, we know we would gain a lot of valuable insights that could be applied to a version two.

Once we got that user feedback we would like to build out an additional flow that demonstrates other features and explores more possible scenarios outside of the happy path.