Sous App Design

Empowering Users to Reduce Food Waste

AI

Product Design

End-to-end

Product Research

Timeline: 3 months

Design Role: Design Lead, Product Research, Product Design

Design Team: Jared Munoz (Product Research), Marcus Coley (Product Research)

Project Description: Using artificial intelligence to empower users to reduce food waste by showcasing recipes that use their near spoilage items.

PROJECT IDEATION & DISCOVERY

Originally a personal project to cut back on my own waste.

I swear I lack object permanence when it comes to my fridge, and my food goes bad far more likely than I care to admit. I wanted to develop a system that would help me keep track of my food and give me ways to go through it before spoilage.

So I started looking into it, turns out food waste is a much bigger problem than I thought.

PROBLEM STATEMENT

In the United States, a significant amount of food waste occurs at the consumer level, with estimates suggesting that 70% of food waste happens in homes [1]. This equates to wasted money, resources, and environmental impact.

Why it matters?

Economic impact: US households waste an estimated $1800 per year on uneaten food [1].

Environmental impact: Food waste contributes to greenhouse gas emissions, equivalent to millions of cars on the road [1].

Social impact: Food waste occurs while millions of people face hunger in the US [2].

SOLUTION

To create a platform that seamlessly scans, categorizes, and tracks your groceries while showcasing personalized recipe suggestions to make the most of your ingredients before they expire.

RESEARCH

Method: Desk Research

Desk research showcased large target areas many people struggled with.

Confusion over Expiration Labels: Terms like "sell by" and "use by" cause consumers to discard food prematurely, wasting perfectly good produce.

Poor Planning and Overbuying: Impulse purchases, lack of meal plans, and misjudging needs lead to excess food that spoils before it can be consumed.

Over-preparing Meals: Increased portion sizes and forgetting about leftovers contribute significantly to food waste.

Busy Lifestyles: Younger generations (18-34 year olds) struggle to plan meals due to busy schedules, leading to more waste.

Improper Storage and Lack of Visibility: Food gets pushed to the back of the fridge and forgotten, contributing to spoilage.

Method: Survey

Surveying our users greatly reaffirmed our problem statement.

90%

of users did not plan meals in advance.

80%

of users cooked 2-5
times a week

74%

of users considered forgetting what is in their fridge and pantry their biggest struggle.

Key findings:

This project is tricky, we have statistics that show what causes the most food waste, but we also know that we have to meet people where they are. I did not believe that creating a service that required users to change their behaviors a lot would take hold, after all, we as humans love homeostasis.

Most people don’t meal plan, whether that is due to busy schedules, forgetfulness, or not liking it, we couldn’t expect users to start meal planning.

User’s considered themselves forgetful when it came to their food, in regards to leftovers, and fresh ingredients.

Method: Interview

Going deeper into user struggles: time, and pantry, and recipe management.

"I usually search for recipes online. Sometimes it's hard to find something that uses ingredients I already have on hand, or the recipes seem too complicated."

- L.G. student at Texas State University

I tried a meal planning app a while back, but it required a lot of manual entry for ingredients, the time requirement ruined it for me.

- E.C. student at Texas State University

Key findings:

When looking for recipes online users struggled to find recipes that utilized their ingredients.

Users are busy and won’t participate in systems that take a lot of time like manual systems.

All of our key findings

Translating insights into design solutions.

We had all this data, and we knew how large the problem was but now we needed to think of solutions.

Idea #1:

Knowing users overbuy, create a
food-sharing network that users can
exchange food they can’t or won’t use.

Knowing users overbuy, create a food sharing network that users can exchange food they can’t or won’t use

Knowing users overbuy, create a food sharing network that users can exchange food they can’t or won’t use.

Idea #2:

Create a meal planning and recipe service that allows users to make meal plans and buy ingredients from their plan.

Idea #3:

Catalog ingredients and suggest recipes that utilize them, prioritizing ingredients near spoilage.

(My selected ideation)

Why?

In our research, we observed that most people weren’t accustomed to meal planning, and we couldn't expect them to make significant behavioral changes.

Regarding food sharing, people tend to be forgetful, and developing a service solely for sharing unused food becomes ineffective if they forget what items they have in the first place.

Method: User Journey

Understanding the current user experience.

What were the pain points, and opportunities, and how can we meet the people where they already are?

Meet the full-fridge-fumbler

A young adult trying to save money, eat healthy,
and save their food from going bad.

Defining our priorities.

With a better understanding of the user’s pain points and our opportunities, it was time we started thinking about what this project should include. I hosted a feature prioritization session for my team to go over our ideas and concepts that we believed would most help the user.

Crafting the user flow.

With all these ideas, we had to plot out how users would navigate them and ensure they could be utilized in the most intuitive way possible.

My early designs.

Now we get to the design, and to be honest, these early wireframes were a massive fail. Users struggled and were overwhelmed, and early in the process, the idea was driven by an AI feature users would realistically struggle with.

Upcoming deadlines rushed the project.

With a deadline approaching, I needed to quickly roll out this project and draft a UI that would later be user-tested.

User testing

Like I said these early designs were a fail and a lot of work needed to be done. Here are the main findings.

Positives

At the least users liked the idea, affirming that users would utilize a service like this. After all our strategy is to reduce food waste by helping users with their meal choices and cooking, and by giving them recipes that match their ingredients we could achieve that.

To be improved (a ton)

Let’s start with the big one, the AI chatbot

Perhaps my biggest blunder, admittedly, I was riding the AI wave and attempting to implement a chatbot just for the sake of it. My original thought process was that if users didn't know what they wanted to eat, they could use the feature. However, the issue with chatbots is that they require good prompts for meaningful outputs, and those who don't know what they want would inherently struggle with prompts. Coupled with a lack of AI acceptance at the time, most people were already relatively weak prompters.

An overwhelming pantry

Users felt somewhat overwhelmed when they could see the upcoming expiration dates of all their food, as it seemed like they needed to find a way to use them immediately.

The recipe page needed a rework

Users rightfully pointed out that there needed to be a way to generate ingredient substitutions since the service most likely was not going to be spot on with its expiration predictions 100% of the time.

Users wanted a better recipe overview encompassing cooking times, difficulty, and other information that could affect their recipe choice (ex. one pan dishes)

Iterative design changes.

Fixing my mistakes, and making sure that users could confidently use the service to target their needs.

Still AI, less chatty

With users unsure about the chat bot and the issues I discovered with the implementation of one. I opted to change the way users would interact with the AI.

Instead of a chat bot, users enable ‘smart search’ which enhances the search by allowing users to search by flavors, ingredients, recipes and other factors. Triggered by a toggle this ensures user’s are in charge of the AI at all times.

A cleaner pantry

To reduce the stress inflicted on users when bombarded with the expiration dates of all their food, we opted to only show the rough expiration dates of the food spoiling soon.

A completely improved recipe page

A focused homepage

First, goodbye to a page entirely for scanning, to encourage users to use the service properly I moved the scan function to the homepage.

Second, a focused expiring soon, to cut back on the same issue users faced with the pantry with being shown a lot of expiration dates.

Final Designs

Thank You!

You’ve reached the end... why not view my other projects?

Let’s Connect

Email

Resume

LinkedIn

Let’s Connect

Email

Resume

LinkedIn

Let’s Connect

Email

Resume

LinkedIn