Project Summary
"We want to allow users to select from among several options for quick meals, then follow step-by-step directions to prepare the selected meal."
Client: Career Foundry
My Role: Voice Designer
Deliverables: User Stories, Dialogs, Flow Charts, Voice Script, User Testing
Tools:
Adobe XD
Alexa Skills (AWS)
Illustrator
The Challenge
To create a voice interaction for the Alexa device with Breakfast, lunch, dinner, and snack recipe buckets. Interaction needs to include suggestions to users on meal type, food, or random suggestion. Users should be able to ask for more recipes and have options for missed steps or starting and stopping the Alexa skill.
The Problem: Users want to make a meal with step-by-step instructions guided by Alexa.
User Stories
Sample Dialogs
Flow Chart
In this next stage, I created a flow chart highlighting key points between the user and the Alexa Skill. There was a high complexity as users can have several utterances and offer spontaneity that needs to be accounted for. By doing this we discovered the ideal interaction with some prompting from the skill and allowed for a circular process in case there were errors within the interaction.
Voice Script
The voice script was the chance to review and set up the system's response to users' inquiries and directions. At this stage development of a user's interaction was mapped. From the previous testing of other Alexa skills, I had identified some pain points such as having the skill ask too many times for repeat utterances or ending abruptly.
Testing Goal
The Goal of the Usability Test for 5-minute recipes is to see how a user responds to having a digital voice assistant to their cooking inquiries. We’ll be looking for familiarity and comfort while using the skill as well as convenience or reaction to a simplified version. Afterward, a small survey was conducted to see if users are interested in the use of the skill.
Objectives:
• To see if users can complete the full recipe.
• To see if users identify incentives in using the skill.
• Identify pain points that would hinder the user from using the skill or wanting to use it again.
Insights:
• Two control groups presented during testing, those familiar with using skills and those who are brand new at it.
• Opportunity to have a multimodal setup many participants wanted a visual to follow along with the skill.
• Users who were comfortable in the kitchen did not like the skill at all. Identified the ideal user being a beginner to an intermediate chef.
“I can see the benefit of not having to wash my hands or clean my hands in order to view the recipes and wouldn’t have to keep going back and forth.”
“I really wanted a video or something to follow along.”
Conclusion
With minimal coding, I was able to create an Alexa skill that allowed users to interact with Alexa by following a recipe. The test results were enlightening and brought up key insights to be reiterated in the next build phase. It also helped narrow down who the users are and can design for them in the next iteration. There were a lot of pain points identified due to educating users on how to use the skill which would require a higher level of coding and development. Most users wanted to use the skill again if the bugs and kinks were worked out.
