Feature Add-on
Implementing AI within MyFitnessPal
Product
My role
UX/UI Design, UX research, information architecture, wireframing, prototyping, usability testing.
80 hours
Figma, Figjam, Maze
DesignLab Capstone
Background
Tracking calories, macros, and other nutrition information can be highly time-consuming. This rings true, especially for busy individuals with demanding jobs, families, and other responsibilities.
Problem
Entering calories, macros, and meal planning doesn't need to take up all your time. Implementing AI into an app like MyFitnessPal (MFP) can help alleviate users' obstacles when tracking health and fitness goals. AI can take the hard work out of tasks such as planning a meal, ensuring staying within your goals, and learning more about health and fitness.
Design goals
A user-friendly and efficient chatbot design
Ensure the branding of MFP is interwoven within the new feature design
Create useful flows and possible capabilities that can potentially help users achieve their goals faster and easier.
Discover
What we want to find
Determine the level of experience and comfortability users have with AI.
Understand how the use of MyFitnessPal fits within the user's daily life.
Understand the challenges with current and past users of MyFitnessPal.
Learn what features are highly utilized and valued by the user.
Methodologies
SWOT analysis
User interviews
Usability testing
Define
A SWOT analysis is used to understand the current state of the market, identify areas of improvement, focus on strengths, seek opportunities, and navigate through risks and weaknesses.
The following companies were analyzed:
MyFitnessPal
Carb Manager
Lifesum
Lose it!
Fit Genie
Click to view analysis
Overview:
4 people between the ages of 20-45 years old who use some kind of calorie tracking app.
All interviews were conducted through Zoom
An affinity map was done to synthesize all the data
Synthesis:
After interviews, it's clear users are dissatisfied with MFP due to its costly paywall, lack of resources, and features. All four interviews highlight this trend: users initially start with MFP but switch to other apps offering free and superior features. Despite being a top calorie tracker, MFP risks becoming obsolete if it disregards user needs. Investing in AI might be costly, but it's essential for MFP's relevance.
100%
of interviewees use some kind of AI on a daily basis
4 out of 4
found it frustrating that some of the simplest features are behind a paywall on MFP.
Click to view full data
Affinity Map Themes
AI is friend, not foe
Motivations
Breaking up with MyFitnessPal
Pain points
All about the features
Understanding the user
Meet Alex the Health-Conscious Working Dad
After conducting user research, it was important to develop a user persona, because it aim to bring more empathy and understanding to the user's goals, pain points, and behaviors. Ultimately, helping me stay focused and better align my design decisions.
Click to enlarge
To better empathize with the user's needs we crafted POV statements and How Might We questions to develop a deeper understanding of the problem as well as present opportunities for innovation to help generate possible solutions.
Precious needs to be able to track her home-cooked meals quickly and accurately, because tracking home-cooked meals can be so time-consuming, that she often gets too preoccupied with her responsibilities as a professional and a parent to stick to her goals.
Justin needs feedback when tracking his calories because he struggles to understand if the foods he is eating are benefiting him and leading him in the right direction towards his health and fitness goals.
Fanny needs an effective meal plan that aligns with her goals but is also budget-friendly and achievable because she usually has difficulty researching recipes that coordinate with her nutrition and financial goals.
How might we help Precious log her recipes quickly and effectively to save her time and energy?
How might we help Justin better comprehend what he puts in his body is helping him achieve his goals?
How might we help Fanny stay within her means to cook easy meals that align with her goals?
Ideation
Complete personalization questionnaire
Create a budget-friendly meal plan and add it to my grocery list
Generate recipes with groceries : Food database and/or photo scanning
Get feedback on your daily diary
Develop
Say hello to MyPal
An important design consideration that needed to be thoroughly considered was what the chatbot was going to look like.
These icons were narrowed down and created to finalize which design was best. The far right icon represented a bit more fitness, while the one in the middle resembled more of a bot, and the left was the cleanest design and the headphones made it fun and unique.
The idea to combine two icons came to be because I wanted to emphasize not only a chatbot but also give the icon personality and fitness characteristics. Thus, the icon for MyPal was finally decided.
Keeping the research and the branding of MFP in mind, sketches were drawn up to continue to flush out any further issues. Sketches and wireframes allow us to quickly put our ideas into function to examine further pain points and to start to visualize our solution.
Deliver
Prototyping this feature within MFP seemed easy at first, however, there are some limitations for a very localized feature. Triggering different functionalities for a chatbot became a bit more tricky than expected. For example, identifying separate points of entry to ensure the user journey within the prototype made sense and that the users could go through the prototype and complete the task flows.
Click to view prototype
Personalize your experience
Create meal plans and add them your grocery list
Generate recipes with ingredients within MFP's vast food database
Generate recipes even faster by snapping a photo of your ingredients
Get feedback on your daily diary and learn how to improve the next day
Usability testing
Once the prototype is set up, it is crucial to test the product to discover any pain points with the design. The usability testing was completed through Maze.co as an unmoderated test.
Make sure that users can navigate through the prototype with ease.
Ensure task completion for each chat flow
Test the user’s understanding of how MyPal can help achieve their goals faster.
Identify any further pain points that I may have missed.
Determine if the added feature increases the users’ satisfaction with the MyFitnessPal app.
Understand if the users find that using an AI chatbot for meal plans, diary feedback, or meal generation helps achieve their goals.
Understand what elements of the feature contribute to the positive or negative user experience.
Successful completion rates
Time to complete
Learnability
Positive influence on the users
Personalization Questionnaire
Create a meal plan and add the ingredients to a grocery list
Meal generation by using the food database
Meal generation by snapping a photo
Get feedback on daily diary
Results
9 participants completed the usability test
Successful completion rate
Time to complete
Learnability:
“Were there any difficulties or confusion encountered while interacting with the chatbot?”
“There were difficulties with the three tasks in the middle. There was not a place to press”
“There were a couple of tasks I was not able to complete and ended up getting stuck within other task flows”
“no there weren't”
“A bit confusing on what exactly you can ask and get answered. Would be good if there were options to ask or suggestions.”
“I did have some difficulties with the grocery list task and finding the right place to start the task from. Other than that, everything looked great and I liked the various ways one can access the chatbot.”
"No"
“I completed the test on my pc, so it was a little difficult to open the diary feedback since I didn't know I had to click and drag the macro tracking window up to see more. Also, for the meal plan question in the beginning, I had some difficulty because I thought I would have to look under "Plans" to set up meal plans. Aside from that, the chatbot was pretty helpful.”
Influence on the users
What aspects of the chatbot’s functionality or user experience do you think needs improvement?
“I think overall, it is great! No notes”
“The chatbot icon wasn't immediately clear and it was hard to figure out what it was until halfway through the usability test.”
“I can’t think of anything”
“A better explanation of how it can be utilized when opening the app. Wasn’t sure how to go about using it at first and got confused.”
“None that I can think of at the moment”
Iterations
Following the usability test, the overall results were mixed with constructive and positive feedback. The feedback revolved around needing to understand how the prototype flowed. Due to the limitations of the prototype and how to trigger specific actions, the users required clarification. In addition, working with Maze can be a challenge as testers who have never completed a test like this before, there can be a huge learning curve. It was clear that if this were an actual product, users would better understand how a regular chatbot would work. One lesson to be learned is with usability testing we have to be very specific in the actions we want the users to take, practically helping them along the way.
After discussing the results with my mentor, we decided to make small changes to improve the experience of:
Adding suggestions to the opening of MyPal each time the users want to interact with the feature. Providing users with some functionality suggestions rather than just an open message concept, similar to the personalization questionnaire.
Adding a modal for the new feature, as a tester indicated it was hard to recognize the icon for the chatbot as it blends in with the UI.
Conclusion
Adding a feature to an already existing product is no easy feat. There are numerous elements to consider such as branding, the need for a new feature, and the integration of already existing features. Throughout this project, I learned more about AI design as well as the struggles that users go through when tracking calories and trying to stay aligned with their health and wellness goals.
To review, the problem of this project was to help users navigate obstacles while tracking their health and fitness goals. This can include being tasked with time-consuming meal planning, guesstimating calories and macros with homecooked meals, or knowing if the foods you're consuming align with your goals. With AI becoming increasingly prevalent in today's workflow, utilizing a chatbot system, similar to Chat GPT, within MFP can assist users tremendously in working towards their goals. Empathizing with the users' wants and needs, it became apparent that the help of AI will not only alleviate some of the stress of planning and implementing users' goals but also allow users to take back some of their time, educate users on how to improve, and help users stay consistent with their goals with MyFitnessPal. There is still a lot I need to learn when it comes to usability testing, and if I could do it over I would like to complete moderated testing and ensure that the prototype works smoothly to gain better results.
Moving forward, I want to continue exploring what MyPal is capable of. Solving users' problems in a way that may eliminate some of the other features that don't get enough attention within the app. Furthermore, optimizing the feature so that users can integrate the assistance of MyPal in their everyday routine. People are starting to understand that collaborating with AI will not only improve how humans approach their habits but in the long run enable people to change their lives no matter their interests. Health is no exemption. If MFP introduces this feature, the result may bring new and existing users to buy in on premium subscriptions and improve the revenue for the business.
View more of my work
