Voice User Interface (VUI) UX Design
CASE STUDY
Healthy Tasty
OVERVIEW AND CHALLENGE
Five Minute Recipes: Allow users to select from among several options for quick meals, then follow step-by-step directions to prepare the selected meal.
PROBLEM STATEMENT
Making a meal can be difficult in today’s busy world. To help solve this problem, I designed and deployed an Alexa skill that allows users to browse several recipes, then select one and have Alexa assist with the meal preparation.
CONSIDERATIONS
● How do people generally search for recipes or recipe inspiration?
● How will users navigate the available recipes without visuals?
● Is there a particular audience for the skill?
● How do people prefer receiving directions for meal preparation?
The goal of this project was to illustrate the voice design process from start to finish within the context of building an Alexa skill. In the following sections, I looked at the design criteria and basic requirements for your recipe app.
THE PROCESS
I was tasked with creating an App which answered the brief set out above. By following the process of UX Design, I was able to create an Alexa Skill.
I used UX processes and applied the principles of Voice User Interface (VUI) Design to create an ‘Alexa Skill’ providing healthy meal options for breakfast, lunch and dinner.
TOOLS
Alexa Developer Console, Git Hub, AWS Lambda, Pen and Paper
DURATION
October – December 2020, 3 months
ROLE
VUI, UX and UI Designer
RESEARCH
GETTING STARTED
I first researched and learnt basic Java Script and HTML in order to understand how to structure the Alexa Skill and therefore work out the relationship between the skill and what is making it work. I looked into the cases for using voice over visual design and thought about the benefits of using this direction to best deliver the solution to the problem statement. I considered existing non-voice apps and how a voice app could provide a solution to an already existing problem in a better way than using a visual interactive solution – Buying a train ticket as an example.
BUILDING A PERSONA
I next worked on building a Proto-Persona, this is a person that we can refer to who would be a typical user of our skill – in this case, Diana – a professional dancer who needs healthy meal options that are quick to prepare in her busy schedule. From here I created a System-Persona (Kate) who will be the voice assistant delivering the recipe solutions; it’s important that these two personas relate with each other and build a connection.
HOW THE INTERACTIONS WORK
The next stage was to further understand how the user will use the skill for this I created User Stories which are typical scenarios our persona will use on a day to day basis. This process also highlights areas where further interactions need to be added in the form of Utterances, Intents and Slots – all of which need to be added to so that the assistant is able to answer on multiple levels and start to a conversation. To further enhance this are I researched Sample Dialogues to work out answers and potential gaps in the voice interaction.
USER FLOW FOR VOICE
From the research and feedback I had already gathered I created Personas of people from different backgrounds, all with different needs. I needed to understand each Persona. This enabled me to map out some scenarios that would be required from the App and work out goals in the form of User Journeys. From here I could work out the journeys users could take and map these out accordingly.
VOICE USER SCRIPTS
Through working out the structure of the skill more thoroughly, I was able to set up the Prompts that Alexa will say to help. I structured the skill by setting up Utterances to Launch, Help and Choose Meal Type. It has been very interesting to investigate how many possibilities are available for a relatively small request.
CONTEXT AND MEMORY
As the Alexa technology progresses within AI it is fascinating to see how this technology is evolving.
When we initially open the skill we could get a Novice prompt to give us a useful guide on how to use the skill and the features it offers. A Normal Prompt could make me feel at home by greeting me by name or a Proactive Prompt could encourage me to continue with my previous searches for gluten free food, as an example: it’s important at this point to give the user the option to filter by preferences such as vegetarian or an already saved search.
ACCESSIBILITY AND SAFETY ISSUES
It was an interesting study to discover that established platforms such as SIRI do not really support users with visual impairments but rely on giving a list of google suggestions as a solution. I also researched more about the safety of children online or in apps and skills and how there are so many ways we need to improve these areas and have them in mind as we we continue to create products and solutions. As designers it’s our job to push boundaries but also to consider how we protect minors or keep in mind the accessibility of other users. Fortunately Designing for Voice is one of the most transparent ways of creating experiences that we interact with.
It will be something I take forward and think about in my work. You can read more about this in blog posts I have written on these subjects.
