COOKING WITH NONNA
Cooking with Nonna ia an Amazon Alexa skill that provides users with easy to follow, step by step recipes.
Overview
CLIENT:
Career Foundry
ROLE:
VUI Designer | User Research | Usability Testing
DURATION:
1.5 Months (Sept. - Oct. 2021)
TOOLS USED:
Pen & Paper | Alexa Developer Console | Echo | Google Sheets
Draw.io
Introduction
This is a case study for a Voice User Interface (VUI) project that took four to five weeks to develop for the voice interface speciality course at Career Foundry. The purpose of this project was to design an Alexa skill that allows users to find and cook a quick recipe using only their voice.
The project brief was as follows: Allow users to select different types of recipes and follow step-by-step directions provided by Alexa to prepare the meal. The skill should include the following features:
-
Breakfast, lunch, dinner, and snack recipe buckets
-
The ability to ask for a different recipe if the user doesn’t like a suggestion
-
A strategy for dealing with users who miss a preparation step or need something repeated
-
A way to check whether your user is ready to move on to the next step
Personas
One of my first steps was developing a user and voice skill persona. For voice design it is equally as important to develop a persona for the assistant or skill since a main priority is to achieve as much of a natural “human-like” interaction as possible.
User Persona
System Persona
User Stories
The user stories were developed next based on the user persona. These stories are short statements that provide a design direction while defining the requirements of the skill. They introduce a user with a specific need and present a product feature that may help the user achieve their goal.
Provide recipes for breakfast, lunch, dinner and snacks
“As someone who is always on the go, I want to be able to quickly plan my meals for the week, so I don’t have to worry about it later on or eat fast food.”
Sample Dialogues
Sample dialogues are an important part of the design process because it is the imagined scripted conversation between the user and the system. As the designer it is my job to craft the conversation to guide the user and solicit the desired response. This step in the process is where the personality of the skill and the interactions between the skill and the user start to develop.
System Flows
The next step in my design process was creating the system flows which would determine how the information would be input and output in the skill. Creating these flows helped me make sure the user could proceed through the skill without any problems by creating an interaction model. These flows are essential business documents that serve as the roadmap for application development, user testing and the review process.
Voice Scripts
Once the system flows were developed, I was able to refine the sample dialogue to serve as a script for recording the prompts as the system responses. Below is a sample of script showing the “Welcome” response that is part of “choose_type_message” intent. A spreadsheet was utilized to organize the intents and utterances of the users and the responses of the skill based on that criteria. Without the development of these scripts usability testing would not be possible.
Usability Testing
It was now time for the most important usability testing. I recruited 5 participants with 1 of them doing the test in person with an Amazon Echo device. The remaining 4 usability tests were conducted via Zoom with the user interacting with the echo device over a conference call. This testing method worked better than expected and in my opinion was better than “wizard-of-oz” testing since the testing was done with the actual device and the users could get a real sense of the interaction with Cooking with Nonna.
Click image to see Usability Test Results
Conclusion
It was interesting to see some of the differences in working with a user experience project versus a voice user interface project such as this one. One of the biggest challenges for me was creating an experience where the dialog is as natural and engaging as possible. It was fun to design, plan and guide the user through the various types of interactions that are possible with the skill. This was a great learning experience for me and I discovered that it is more difficult to design for voice interfaces than visual interfaces because there are more variables and uncontrollable circumstances when designing for voice.
A major problem that I encountered during this project is that the template that was provided for the Alexa skill was very limited. I don't know JavaScript therefore a template was used and it did not allow for some of the advanced functionality that I had planned from the beginning of the project. Since I did not know this from the beginning and I wanted to do the usability testing using an Alexa device versus "Wizard of Oz" testing I needed to modify the project to be as simple as possible.
Next Steps
Based on the conclusion I am planning on using “VoiceFlow” to expand on the project and see if I can incorporate some of the planned original functionality. I would like to incorporate the ability to: search for a recipe based on ingredients, save a recipe to favorites, set a timer and search for additional cooking advice on the internet while using the skill. From my initial research it seems like VoiceFlow does not require any coding and can provide a robust skill design.
See my other projects
