How thoughtful VUI design can reduce food waste, support sustainable cooking, and simplify meal preparation? By focusing on user needs and aligning the experience with intuitive voice commands, this skill offers users a seamless, environmentally friendly, and enjoyable cooking companion.

Amazon Alexa Skill / UX Research /  VUI Design / Prompt engineering / Node.js /  Testing

Project
VUX / VUI
Solo Student project
Duration 2 months
2024

About

Making a meal can be difficult in today’s busy world. To help solve this problem, I designed & deployed an Alexa voice skill to empower users to reduce food waste and simplify meal prep by guiding them to create recipes based on ingredients they already have.

Problem

The Meal-Prep Challenge: Reducing Food Waste in Busy Lives

Preparing meals is challenging for busy individuals, often leading to food waste as people struggle to creatively use leftovers or prepare meals with the ingredients on hand. Users need simple, quick solutions to access recipes that minimize food waste without adding stress to their routines.

My Solution

A Voice-Powered Solution: Simplifying Meals & Reducing Waste

I created an Alexa Recipe Skill that provides access to various recipes based on leftover ingredients, reducing food waste, simplifying meal prep, and enhancing convenience. The skill guides users through recipes step-by-step using voice commands, providing an intuitive cooking experience.

My Solutions

Simplifies Meal Prep: Users save time by quickly finding recipes based on the ingredients they have, reducing the need for meal planning.
Minimizes Food Waste: The skill encourages users to repurpose leftover ingredients, creating a more sustainable cooking habit.
Hands-Free Convenience: By using voice-only navigation, users can cook without having to touch devices, ideal for a busy or messy kitchen environment.

Impact & User Value

Voice User Interface

Design Process

Using Design Thinking as a foundation, I broke the project into five key stages to ensure the skill would be both functional and user-centered:

Empathize: Conducted user interviews to uncover real-life challenges around meal planning, food waste, and ingredient management.
Define: Identified common pain points, like food waste and lack of meal inspiration, which guided the feature development.
Ideate: Brainstormed solutions to make the skill adaptable and accessible, enabling voice-only recipe browsing and step-by-step guidance.
Prototype: Created an interactive, voice-first design prototype that incorporated critical features identified in user feedback.
Test: Iterated based on user testing, fine-tuning features to optimize clarity, ease of use, and alignment with user expectations.

User Research

“I want to make the most out of food that is already available at home”

To understand user behavior and challenges, I conducted remote interviews with five (5) participants, focusing on:

Cooking habits in a busy lifestyle
Desire to reduce food waste by creatively using leftovers
Need for quick, easy access to recipes without the need to handle screens or devices physically

Insights:

Insights (shown below) shaped the skill's core functionalities, with a focus on guiding users through meals that maximize the use of available ingredients, reduce food waste, and streamline the cooking process.

Key Insights / User Interviews

1.

Struggle with Meal Planning:

Participants found it challenging to create weekly meal plans, often leading to food waste and repetitive meals.

2.

Frustration with Food Waste:

Users expressed a desire for ways to use up leftovers, linking it to cost-saving and environmental benefits.

3.

Lack of Recipe Inspiration:

Many found it difficult to come up with ideas for using leftover ingredients.

Research Solution & Persona Development

Building the Perfect Team: Personas Drive the Development of the Recipe Skill

Analyzing both the user persona, Jenny, and the system persona, Chef Eliza, helps us understand why each is essential for the Alexa Recipe Skill's success. They offer unique insights into the problem-solving process, guiding decisions about feature design, user interaction, and overall functionality.

User Persona - Jenny:

The Busy, Budget-Conscious Cook: Jenny’s persona represents my primary users—individuals juggling tight schedules, health-conscious goals, and budget constraints. By focusing on Jenny’s needs, I prioritized simplicity, speed, and sustainability, ensuring the skill would save time and encourage eco-friendly habits.

Captures Core User Needs
- Value: Jenny’s profile highlights essential pain points—reducing food waste, easy meal prep, and using available ingredients. This clarity ensures the skill addresses the most impactful features for users on tight schedules.
Guides Feature Focus
- Value: Jenny’s goals help prioritize features, such as:
  - Simplicity & Speed: Quick, accessible recipes.
  - Sustainability: Creative use of leftovers.
  - Ease of Use: Hands-free, voice-only interactions.

System Persona - Chef Eliza:

Friendly, Knowledgeable, Supportive: Chef Eliza was designed as an engaging companion rather than a robotic assistant. Her calm, clear instructions and encouragement create a positive cooking experience, essential for building trust and guiding users smoothly through recipes.

Sets the Interaction Tone and Speech Style
- Value: Chef Eliza’s friendly demeanor makes the skill feel conversational and supportive, enhancing user engagement and enjoyment.
Builds Trust and Confidence
- Value: As an experienced, eco-conscious guide, Eliza fosters trust in her suggestions, helping users, especially novices, feel reassured in their cooking endeavors.
Personalizes User Experience
- Value: Eliza adapts instructions based on user input, offering tailored recipe options that meet individual needs, which is essential for users like Jenny.
Reinforces Sustainability Goals
- Value: By emphasizing food waste reduction and eco-friendly practices, Eliza aligns with users' values, promoting sustainable cooking habits.

Research

Why Both Personas Are Essential for Project Success

1.

Aligning User and System Goals

User Goals: Jenny seeks efficiency, affordability, health, and sustainability.
System Goals: Chef Eliza offers tailored, quick, budget-friendly, waste-reducing recipes.
Value: This alignment ensures that every feature meets genuine user needs, increasing usability and relevance.

2.

Human-Centered Voice Experience

Chef Eliza’s voice-only guidance addresses Jenny’s needs for accessible, intuitive cooking support, compensating for the lack of visual cues.
Value: Makes the Alexa skill user-friendly and practical, designed specifically around user behavior.

3.

Promoting Positive Behavior Change

Chef Eliza supports sustainable cooking habits, helping users reduce food waste by creatively using ingredients.
Value: Fosters eco-friendly habits, increases engagement, and meets user needs for mindful, waste-conscious cooking solutions.

In summary, Jenny and Chef Eliza are indispensable to the skill’s success. Jenny’s persona drives design decisions that make the skill relevant and practical, while Chef Eliza brings warmth, expertise, and encouragement, transforming the user experience into one that is both functional and enjoyable. Together, they create a user-centered, sustainable cooking assistant that addresses key pain points, empowers users, and promotes positive, eco-friendly habits.

“As a user, I want to be able to give voice commands so that I don’t need to use dirty hands to touch any screens.”

User Stories

From Wishes to Dishes: User Stories that Shape the Cooking Journey

These user stories are vital because they reflect user-centered design choices informed by real needs and preferences uncovered in interviews. By aligning the skill’s features with these validated stories, the design process ensures a practical, accessible, and enjoyable voice experience that simplifies cooking, reduces waste, and meets diverse dietary needs.

Here’s an analysis of some of the stories, why it’s important, and how the insights gathered validate these needs and inform the design process.

“As a user, I want to be able to give voice commands so that I don’t need to wash my hands and unlock my screen all the time.”

Prioritizing Hands-Free, Voice-Only Interaction.

Importance & Value: User interviews revealed that people often multitask while cooking, and are frustrated by screen-based interactions that require clean hands. Implementing robust, responsive voice commands validates this need and supports a primary motivation: enabling hands-free, uninterrupted cooking.

1.

2.

“As a conscious individual who wants to avoid food waste, I want to be able to choose the main ingredients, so that I can make the most out of my leftover food.”

Reducing Food Waste Through Ingredient-Based Searches.

Importance & Value: Insights about food waste motivated the inclusion of ingredient-based recipe searches, allowing users to find recipes tailored to their leftovers. This feature directly addresses user pain points, supports sustainability, and increases the skill’s practical utility.

3.

“As a vegan, I want new recipe ideas that are suitable for my diet, so that I can quickly choose meal plans.”

Incorporating Flexibility for Dietary and Budget Needs.

Importance & Value: Users’ dietary preferences (e.g., veganism) and financial considerations influenced the inclusion of filters and budget-friendly recipe options. These features ensure the skill’s relevance across different dietary needs and financial situations, validating that the skill addresses a broader user base.

4.

“As a user, I want to be able to request the recipe instructions while I am making the food and ask for them to be repeated so that I can be sure I’m following the instructions correctly.”

Clear, Repeatable Instructions for Enhanced Usability.

Importance & Value: Feedback on recipe clarity and repetition led to the design of step-by-step, repeatable instructions, ensuring users feel supported and confident throughout cooking.

5.

“As a user of the recipe app, I want to be able to pick the meal category and choose the meal type so that I can quickly find recipes that match my cravings/ingredients on hand.”

Quick Recipe Search by Category and Type.

Importance & Value: Insights about food waste motivated the inclusion of ingredient-based recipe searches, allowing users to find recipes tailored to their leftovers. This feature directly addresses user pain points, supports sustainability, and increases the skill’s practical utility.

6.

“As a visually impaired user, I want to receive recipe instructions through voice guidance so that I can easily follow them.”

Voice Guidance for Visually Impaired Users.

Importance & Value: Interviews and empathy mapping indicated the value of inclusive design. To accommodate this, I prioritized clear, detailed voice instructions without screen reliance, ensuring visually impaired users feel fully supported and empowered in the kitchen.

User Flows for Voice

Mapping User Flows for Seamless Voice Navigation

User flows map out how all the intents in a skill are related to one another—they show what a system can do as well as how a system will respond to various inputs. Just like intents, flows show the what, not the how.

User flows are a way to look at the logic of the system without going into detail regarding what the system will actually say.

From Script to Skill

The Script Behind the Cooking Experience

Scripts and flows go hand-in-hand. Writing scripts is very similar to going from low- and mid-fidelity wireframes to higher-fidelity mock-ups.

A script is organized by intent. The script has a tab for each state, and aligns with the flow. Each section include the utterances that get the user into that state as well as the system’s responses to those utterances. I wrote multiple utterances that all lead to the same response and for variation, I wrote multiple responses, as well.
The script also contains any error prompts, confirmations, and disambiguation.

Below are shown some example snippets from the script sheet.

Usability Test

Usability Insights Revealing the Need for Multilingual Support

Test Objectives:

Evaluate the voice-driven recipe search process.
Assess ease of navigation and overall user experience.
Identify gaps in project definition needing correction.
Pinpoint pain points and areas for improvement.
Measure clarity and usefulness of step-by-step instructions.

Participant Demographics:

5 participants with varied backgrounds, technical skills, and cooking experience.
Age scale: 16-52

Testing Methodology:

Wizard of Oz testing simulating real-time Alexa interactions in a controlled environment.

Script

Key Insights

/Usability Test

Most participants found it easy to start and navigate the skill.
One participant, with low-tech savvines, needed a bit more time to get accustomed to the voice commands.
Grammar issues caused some confusion, especially for non-native English speakers.
Some recipe instruction parts could be shorter.

Issue #1: Lack Of understanding for Finnish dish names (Multilingual support) (High)

Description: Few participants (2/5) noted that the skill doesn’t understand Finnish dish names, making it difficult for non-English speakers to use.

Suggested Change: Focus the skill language for specific country / region / languages.

Evidence: Participant had difficulty giving prompts for the system.

Issue #2: Lack of clear instructions for measurement system (High)

Description: One participant (1/5) noted that the skill doesn’t inform, is the recipe instructions for litras or gallonas (us metric system) and celsius or fahrenhait.

Suggested Change: Add clear instruction to the beginning of recipe (and how to convert the metrics?). Focus the skill language for specific country / region / languages.

Evidence: Participant tried to ask help from the system.

Issue #3: Lack Of Multilingual Support (Moderate)

Description: Few participant (2/5) noted that the skill doesn’t support multiple languages, making it difficult for non-English speakers to use.

Suggested Change: Implement multilingual support (FIN - ENG) to cater to non-English speakers.

Evidence: Participant had difficulty giving prompts for the system.

Issue #4: Lack Of Side Dish ideas / beverages recommendations (Moderate)

Description: One participant (1/5) noted that the skill doesn’t support ad-ons for drinks, side dishes and dessert recommendations

Suggested Change: Implement recommendations for ad-ons.

Evidence: Participant wished drink and dessert recommendations.

Conclusion

The usability test for the Social Savory Club Alexa recipee skill revealed the importance of region and language understanding. While the skill was appreciated for its clear instructions, good repetition feature, and user friendly interaction, improvements for language understanding and specific recipes for different region are needed.

Key areas for improvement include supporting multiple languages, implementing an ad-on feature for drinks and desserts, improving navigation and information about metrics system, improving language learning for recipe names based on specific ingredients and regions. Addressing these issues will enhance user satisfaction and accessibility. Some issues should be addressed promptly to improve usability and prevent user frustration when launching the skill.