MealGenie
Overview
I designed an AI-powered app called MealGenie that provides an easy and convenient way to plan and prepare meals and discover new recipes.
Meal planning is a big part of my week but I hadn't found a great solution for preparing my recipes.
I had recently discovered ChatGPT and wanted to experiment with creating an individualized weekly meal plan for myself. I told it my likes, dislikes, and other relevent dietary information and asked for a personalized meal plan, and I was pleasantly surprised with the results.
I became interested in designing this app as a way to share this experience I had using ChatGPT with others. In particular, I've shared it with my mom to encourage better eating habits.
User Research & Personas
My user research included surveys and interviews. I received over 100 responses in total from a survey that I posted across several subreddits.
- Respondents were 60 female : 40 male and most were primary cook
- Over half reported that they had very little time to plan/shop for meals
- Very few said they were on a tight budget
- Wanted more flexibility with recipes (ability to swap/alternate ingredients)
- Wanted wider range of recipes and better recipe search features
I was able able to narrow down the potential target markets in the form of three personas:
Betty
Betty, the Busy Parent, needs a quick and painless way to plan meals for the week. She works and is the primary cook for this dual-income household. Grocery budgeting is not a huge concern. She has kids who also need to be happy with the meals.
Jango
Jango, the Fitness Guru, has specific dietary needs to achieve his his goals. He's tired of eating the same thing all the time, so he's looking for variety. He will need a way to swap out items in a recipe to fit his Paleo diet. He is an individual who cooks only for himself. He is sometimes budget-conscious.
Maria
Maria, the Parent of a Child with a Food Allergy, needs nutrition information and ingredients listed front and center. Her son Alex has an allergy to gluten that she takes very seriously. Even she eats gluten-free stuff most of the time because she's adapted to it. Dinner is a big part of their family experience together with her son. Swapping ingredients is crucial when she goes to look at recipes.
Insight: Money/budgeting is less of a factor than expected
- Survey results showed only 10% of respondents said they were on a tight budget and 30% don't budget at all
- Nutrition information (counting Macros) was more important than budgeting
- Some focused on fitness
- Decided not to focus on people on a budget (not people who want to use items already in their pantry).
Insight: Some people use multiple apps to plan meals
- Some survey respondents entered several apps that they currently use
- Interview showed people are looking for one app that does it all
Insight: Many people are looking for a better search UX
- Survey respondents reported wanting more recipe variety and more customization options when searching
- Users seemed dissatisfied with their search results and the tools they had available to find good recipes
Hypothesis: Could AI help people get more personalized meal plans?
- Users are looking for more customization options
- Rather of keeping a huge database of possible meal plans, AI could probably produce more personalized results.
Design & Usability Testing
I focused on a simple interface that utilized Tabs with Swipable Buttons and also provided a Text Search Box for adding custom items to the users' likes, dislikes, etc.
There is a delay while the results are generated, so I added a fun animation to keep the user engaged.
- Used maze.co for usability testing, with 20 paid testers
- Of those that reached the end, 4/5 stars was the average rating
- Most testers were excited and delighted that it worked at all: "I was pleasantly surprised with the results I was given and how easy it seemed to generate a meal plan with the requirements I asked."
