About Quick Bites
1
Provides easy, healthy meal ideas
Meal ideas for breakfast, lunch, dinner, and even healthy snacks. A delicious meal in less than 10 minutes!
2
For busy students and young professionals
These individuals live on their own and have to cook for themselves, while maintaining a busy schedule.
3
Project timeline, tools used, and my role
Timeline: October - December 2024
Tools used: Alexa, JavaScript, Figma, Otter.ai, Google Sheets
My role: Voice Designer
Background
The discovery phase for Quick Bites commenced in October 2024, focusing on competitive research and user interviews. I identified a common challenge among students and young professionals: the difficulty of preparing quick, healthy meals due to their demanding schedules. To address this group’s needs, I focused on delivering healthy, versatile meal options that can be prepared in ten minutes.
Goal of this project
Help busy students and young professionals prepare simple meals with using voice design technology.
Deliverables: workflow, personas, user stories, sample dialogs, intents, user flows, script, and user testing, and meal recipes
How I worked
I slightly modified the Design Thinking process to tailor it to my Voice Design project. I started with ideating on the UI design right away, as this made it easier for me to imagine the skill's content and my voice system's personality.
Then, I conducted research on competitors and my users to have a better understanding of what my skill should entail. These insights were applied in the user stories, flows, journeys, sample dialogs, and scripts. Once I built my skill in Amazon Alexa, I tested it with users and gathered their feedback for iteration and eventual publishing of the skill.
​
Last but not least, I also wrote some recipes myself that you can try out!
UI design
As Quick Bites promotes a healthy diet, I envisioned the brand color to be a fresh shade of green. Green stands for nature and freshness, and is often associated with healthy food.
The logo of the skill is a tomato; its vibrant red color serves to complement the brand color.
Competitive analysis
The research phase was kicked off with investigating how the current voice assistants provide recipes to users. I especially looked into Amazon Alexa’s competitors, Google Assistant and Apple Siri.
For comparison, I tasked both of the assistants to give me an easy breakfast recipe.
​
Insights​
​
-
Both Google Assistant and Siri were only capable of linking me to recipe websites.
-
Quick Bites will stand out from its competitors by offering users step-by-step cooking directions.
Google Assistant
Apple Siri
User research
After learning about the competitors, I reached out to potential users of Quick Bites for a short interview. I asked them questions about their cooking and dietary habits, such as:
-
issues they face while cooking;
-
how they manage time between their schedule and preparing meals;
-
and how well they do with maintaining a healthy diet.​​
​​
Insights
-
Users feel more comfortable with a voice assistant that sounds more human-like, using human conversation markers.
-
Users look for details in a recipe, e.g. the duration of the meal preparation.
-
Users need quick and easy instructions since they’re busy with other things.
These and more insights were illustrated in the persona below.
Jay - The independent student
Along with the user persona Jay, I created a system persona for Alexa. A system persona acts as the foundation for designing a voice assistant that resonates with users, aligns with the brand, and delivers a seamless and engaging experience. It’s like giving the assistant a ‘character’ that users can understand and connect with.
Helen - Your personal home chef
User stories
As an important preliminary step for creating sample dialogs and user flows, I used the insights from the previous phase to define several user stories. These user stories reflect how users would use Quick Bites and under which circumstances. After a brainstorming session, I prioritized four user stories.
The most important two are highlighted here:
Sample dialog
With the user stories as a starting point, sample dialogs were created. By anticipating various user inputs, the sample dialogs helped me with visualizing how users would eventually interact with the system in real-life, and which kind of responses would be expected from Helen.
Here are two of the sample dialogs I used as inspiration to create the Quick Bites experience. Please keep in mind that these are not the final responses that made it to the skill. They were refined in a later phase.
User flows
As I was getting closer to coding my skill, I chose one of the sample dialogs to make a user flow for. This user flow highlighted the journey a user would take to prepare a meal with Quick Bites.
Here, I included a few features that I wished to implement within the skill:
-
Choosing a recipe
-
Asking for an alternative recipe
-
Adjusting the ingredients
-
Following the instructions and asking to repeat a step
-
Saving a recipe
The graphic below shows the user flow. To improve visibility, I split the user flow into two parts, starting on the left.
User story: As someone who’s lazy to cook, I want to be able to cook simple, healthy meals, so that I won’t depend on fast-food anymore for quick meals.
Script
The script is a spreadsheet that serves as the blueprint for Quick Bites. With my script, I was able to:
-
Outline sample user utterances needed to interact with the skill, such as launching it, requesting a recipe, or asking for help.
-
Document the recipes available in Quick Bites and Helen’s responses.
To make conversations with Helen feel more natural, I’ve created multiple response variations for each intent. The code randomizes these responses, ensuring interactions feel dynamic and less robotic.
Below is an impression of my workflow for this process, concluding with a summary of the insights I discovered.
Key takeaways
​
Creating the script was an invaluable experience, as it helped me clearly define what features were within the scope of my skill and what fell outside of it.
Since I used a code template provided by a developer beforehand, there were certain limitations that limited me from implementing some desired features. Below is a wishlist for future updates to Quick Bites:
-
Personalization: I’d like Helen to ask users for their names, enabling her to greet them personally during future interactions.
-
Ingredient Customization: Users could request recipes based on ingredients they already have at home or ask for alternative ingredients to accommodate allergies.
These are the features included in the current version:
-
Repeating a step on request
-
Providing help when asked
-
Randomized responses to create a more human-like experience
-
Offering specific recipes within these categories: breakfast, lunch, dinner, and snacks
-
Listing ingredients before beginning the instructions
Recipes
For this skill, I wrote 3 recipes for each category, which were breakfast, lunch, dinner, and healthy snacks. The recipes had to be healthy and easy to make!
Here are some examples of the recipes used in Quick Bites:
Usability testing
Prior to the usability testing, I set up a test plan with the following research goal: Test the key features and functionality of this prototype, specifically targeting our primary persona, Jay.
​
Test objectives:
-
Evaluate recipe clarity and usability
-
Identify any usability issues, like misunderstanding commands and navigation difficulties
-
Measure the success rate for users to complete key tasks, such as finding a recipe, starting the instructions, and navigating through the steps
-
Evaluate user satisfaction and engagement
Process
-
I made use of the Wizard of Oz testing technique.
-
Tests were conducted either in the participant’s home or remotely via Zoom, to create a realistic cooking environment.
-
During the session, I acted as Helen, responding to the user’s commands and simulating the skill’s functionality.
Participants
-
Five participants aged 22-28, consisting of students and young professionals.
-
Selected based on their busy lifestyles, driven by university commitments or demanding work schedules; often leave them struggling to find time to prepare healthy meals.
Discovered issues
The usability tests aimed to assess the user experience of the Quick Bites Alexa skill, ensuring it aligns with the needs of the target audience.
The test report summarized the top 5 issues with the voice prototype, ranked from most important to least important:
-
Vague instructions
-
Limited recipe customization
-
Lack of saving features
-
Missing allergen and nutritional information
-
Context retention
Some quotes stated by the participants during the testing:
Key takeaways
​
The usability testing highlighted key areas for improvement in the recipe skill, including:
-
The need for clearer instructions
-
Greater personalization
-
Better context retention
-
Additional features: recipe saving, ingredient-based searches, allergen information, and potential supermarket integration.
Participants appreciated the skill’s potential for convenience and efficiency but found its current functionality limited.
Design iteration
Earlier, we explored discovered issues found during usability testing. This section dives deeper into how I worked on resolving the most urgent issue: vague instructions.
Improving recipe instructions quality
Evidence: 4/5 participants found hollistic terms like “a generous layer” unclear.
Suggested change: Provide clear instructions to improve user satisfaction.
Changed recipe structure
-
Recipe instructions were rewritten with more specified measurements.
-
Short steps were combined into 1, so users don’t have to say ‘Next’ every few seconds.
-
Added more variety in the recipe database, so users can have more fun with the skill. Some recipes take a bit more time to prepare than others, but always less than 10 minutes!
Before
After
Design simulation
Here is an impression of Helen’s responses. The responses are accessible to the user as either a combination of text and voice, or voice-only, depending on personal preferences.
Moving forward
Moving forward, I should prioritize the following aspects:
-
Enhancing clarity
-
Integrating personalization options
-
Addressing technical constraints to improve context-awareness and interactivity
​
Incorporating these changes will not only address user pain points but also elevate the overall user experience, making the skill more practical and appealing to use.
Based on user feedback, the following features will be prioritized in a future update:
-
Changing recipes based on preferred ingredients
-
The possibility to save recipes for future use
-
Skill recovery after receiving a misunderstood response
I will keep on iterating and eventually publish Quick Bites for a greater group of users. Thank you for following me on my journey. Feel free to contact me if you have any questions or want to know more about me!