The challenge

To prototype an interactive voice experience for children, that enables the Disney Princess Franchise to test and learn about an emerging technology (voice) and present their findings to the wider group.

Discovering what the voice experience could be

Meaningful testing with a very young user base

Expanding product with a road map of future releases

Analyse the competition

Researching what was already in the market, how they work and what makes them engaging highlighted areas to focus on who building our own skill. With a 5 year old user quick testing of children’s game and entertainment skills was done, and it quickly became obvious which approaches worked and which didn’t when it came to creating a voice experience. The use of voice actors rather than the assistants voice, the addition of music and sound effects as well as keeping choices simple where all highlighted by the testing as key to creating the most usable experience. The testing also highlighted key types of skill we could explore further with user interviews.

Talking to the users (well actually their parents).

Because the user base was so young the interviews were set up with parents, who would be key to driving adoption. They would likely be the ones to install the skill on their smart speaker, and influence their children to use it.
‍
Questions were asked about their and their children’s use of smart speakers in their home. By enquiring about the type of games and activities they did, the length of time they did them for and the parents views of the princess franchise, a map of similar themes was built. From this a survey was created and sent out to 500 people, to expand on the user information that had already been captured.
‍
At the same time various staff on the franchise team were interviewed to get a broader understanding of the reasons behind the proposed brief and the wider princess franchise.

Discovery

Using the data gathered from the interviews and surveys, a 2 day “mini sprint” was run with the franchise team. By doing various tasks we quickly narrowed down choices for the logistics of the experience, content, audience and the right technology to use. The content of the skill was speedily created by doing several rounds of crazy 8’s and prioritisation a shortlist of ideas.

Speech bubble saying "The repeat things again and again and again"

Speech bubble saying "Her knowledge of what it can do is limited by my understanding."

Two ideas, a classic party dance game using well know princess music and interactive stories based of various princesses, were chosen. The two experiences were then journey mapped, taking into consideration both physical and voice actions from the child and parent, sounds effects and music, and channels for technology, and distribution.

Creating “on the fly” lo-fi storyboards allowed us to to refine the idea and start concentrating on actual basic content and scripting. The final idea was picked based on the budget and the upfront work that would be needed to create interactive stories, on the part of the franchise team.

The Final Idea

Princes dance game including 3 games of musical bumps, musical statues and a free dance setting, using popular princess songs.

Users

Girls ages 4-6. Can be used individually or in a group.

Build

Prototype using voice actors, showcasing several princesses.

Objectives

To further push brand, with the knock on effect of pushing sales in other areas of the franchise.

Platform

Amazon Alexa
designed to port to Google

CORE VALUE

An activity the whole family can enjoy

A game to keep children entertained

HOOK

Popular music that the users already know and love.

Large range of princess music available

STICKINESS

Repetitive play with children’s favourite songs & Princesses

New games/Princesses added periodically

How It Was Solved

By prototyping in Voiceflow, a working model of the skill was quickly realised. This first model, using the assistants voice was used to refine and expand on the scripts defined in the storyboards. Once a solid script was written then assistants voice was replaced with my voice recordings to make the prototype more relatable to the young user group.
‍
By doing several rounds of testing with 15 children, aged 3 to 7, further refinements to the journey and script were made. To make the experience as relevant and unthreatening as possible for the young users all testing was done in a home setting.

Something that became immediately obvious was that, given the age of the users and the potential for spoken mistakes, the negative journeys were just as important as the positive ones, and guiding the user back onto the positive path in a frictionless way, was an important process to concentrate on.
‍
It was very important to get the scripts and journeys as accurate as possible before budget was spent on professional voice recordings.

The Results

Using the results of the iterative testing and improvement a comprehensive voice map was finalised. Using this the brand team applied specific tone of voice and recorded the first round of professional recordings. Due to budget for the purpose of the final prototype only one princess song was used.

Insight from the project

User testing with 5 year olds must be handled in a very different way to normal adult user testing. They must feel at their ease in a familiar environment. It's better to test in groups.

Asking a child for subjective feedback is hard and interview questions may need to lead them slightly to get a response. eg. “Did you enjoy that?” rather than “What did you think of that?”. Be prepared for one word answers.

Parents would often subconsciously lead or speak for the child.

When developing a voice experience, multiple quick prototype, test and iterate cycles are essential to quickly refine the correct journey and scripts.

Negative journeys in voice are just as important as the main journey.

Prototyping a voice experience for 5 year old princesses.

Full Case Study