Company
TRAKABEE INC
Project Overview
Mysterious allergies attack over 25 million Americans. I started this company because people need help identifying allergies that evade normal medical tests.
Phase 1 is live on the Apple and Android stores and is successfully helping people track and log health symptoms so they can sleuth out their causes.
The next phase is being designed and developed to automate corrolations and provide live findings.
Roles
UI/UX Design
Product Research
Product Management
Usability Studies
Fundraising
Founder
Tech
Duration
Figma
Native Apps
iOS
Android
Firebase
12 months
Responsibilities
This was a lean startup team where I took on the roles of design, research, product managment and fund raising. Everything visual you see here is my work.
.png)
.png)
.png)
.png)
WHY BUILD TRAKABEE?
Symptom trackers on the market don't do the math for you, they track but they don't put cause and effect together. Tracking is slow and tedious in current apps. There are millions who suffer from mystery symptoms caused by food.
Cause: Unknown
“While we found that 1 in 10 adults have a food allergy, nearly twice as many adults think that they are allergic to foods...”
-Dr. Ruchi Gupta
THE MYSTERY ALLERGY MARKET
-
An estimated 26 million or over 10% of US citizens have medically certified food allergies
-
An estimated 19% of adults believe they have food allergies
-
That means 23 million (9%) of adults have symptoms that can't be medically certified but still bother them
-
This app can help these people
DISCOVERY PHASE
THREE DIFFERENT WAYS TO FIND A FOOD ALLERGY
Food allergies 101
-
The correlation method. This involves tracking each food you eat, and each symptom that occurs, then mapping one to the other.
-
The dreaded elimination diet. Eat a 'safe' diet for two weeks, then add in one food in every week, monitoring symptoms as you add each food. What is a safe diet you may ask? Rice or buckwheat...and cold-water fish, lamb, turkey or wild game. That's it.
-
A medically administered scratch test or blood test, however tricky symptoms often don't show up in these two tests.
​
HOW CORRELATION WORKS
Food Eaten and Symptoms Experienced Over Four Weeks
.jpg)
One allergy
.jpg)
Two allergies
.jpg)
Dairy Reaction
Wheat Reaction
Egg Reaction
Corn Reaction
Caffeine Reaction
Yeast Reaction
INTERVIEW RESULTS
I started by recruiting and talking to people with nagging allergies. I came up with a basic script and then let the conversation naturally unfold. It turns out they have a lot of opinions and so many ideas. Here're a few ideas the participants thought of.
Add Barcode Scanner
Tracking takes forever! Speed up tracking by adding a barcode scanner.
Connect to Weather
Could symptoms be triggered by something besides food? Tap into environmental databases that might influence the body, like local pollen counts and weather.
Geolocation
What if my Grandma's peanut butter (she buys Skippy) triggers hives when my peanut butter (organic, ground in my kitchen) doesn't? The location might make all the difference.
Confidence
There are so many moving parts; the results are only as good as the data I feed the app. Tell me how confident you are in each connection the app finds.
Accelerometer
What if the rash comes from exercise? Track my runs vs yoga and things like long walks. Sense different types of motion.
​
​
Don't Make Us Work
Can you give me results without me having to do anything? For instance, connect the app to fitness, calorie tracking, and health apps I already use.
EARLY INSIGHTS. WHAT'S IMPORTANT.
01
Speed
In general tracking is slow. The interface needs to take less than 30 second to track.
02
Design
Clearly identify what is causing the symptom(s).
03
Computation
Do the corrolations work for the user. No one has the time or ability to figure it out themselves.
SURPRISING INSIGHT
When a parent starts to introduce solid foods to their baby they're also forced to go through a food allergy discovery process. Similar to the elimination diet they'll slowly introduce foods, but if allergies don't appear right away they can be equally perplexed at nagging problems and will need to switch to the correlation method.
+ And +
This demographic is not only easier to identify and later target, but they notoriously share hints, help, and great apps with each other.
DEFINE PHASE
FEATURES TO INCLUDE
-
Food allergies have a cause and effect relationship. When you eat a food that triggers an allergic response, symptoms might not appear for days, weeks or even months, eluding doctors for years.
-
Using the powerful computers in our pockets, our phones, we can unravel what causes what.
-
The more data the app has to compute the closer to statistically significance the findings will be.
-
The problem with apps out there now is that they don’t do the computation for you.
DEFINE THE PERSONAS
First, we need to identify who this app is for. When I started this process it was for people who have nagging symptoms and suspect food allergies to be the cause. After conducting user research I thought a pivot would be smart. The app should focus on these groups in this order of priority.
Moms
Specifically parents of babies and especially first time parents. Based on 99% of the parents I talked to it's the mother who is keen to use apps and track and log each food their baby eats.
Adults Suffering from Food Allergies
Most adults who suffer from food allergies have a hunch. They suspect 1-2 foods are causing their symptom but they're never able to be certain.
Care takers like Nannies and Grandparents
If a parent wants to track their baby's food, then more than likely that means everyone who's looking after the child needs to get on board....or at least the app should be easy enough that they can.
TWO POSSIBLE APPROACHES
.png)
Track Everything
Track in Batches
EVERYTHING
BATCHES
-
Find more allergies
-
Higher confidence in results
-
Con: time consuming
-
Con: overwhelming amount of work
Pro
-
Time consuming
-
Overwhelming amount of work
Con
-
Smaller mandatory time commitment
-
In actuality users only have one or two mystery symptoms
-
Less variables mean faster results
-
Babies eat a small variety of foods
Pro
Con
-
Longer length of time needed for lots of allergies
BATCHES
-
Speed of input is the most important problem to solve
-
Start small - make the MVP work with minimal input from the user​​
-
Users almost always have a hunch; they suspect a few foods might be causing their symptoms.
-
A wide net can be cast later when more features are built
The Winning Approach
THE PROBLEM
ALL USERS STRUGGLE WITH RECORDING
The primary problem is that tracking is too slow. It's the thread that ran through the brainstorming ideas, users' suggestions, pain points and failure to stick with other apps.
BUILD TO THE PERSONAS
What does Each Persona Absolutely Need? We will Only Build Solutions that Help all Three.
One Hand Input
-
Only have one hand free with a baby in the other hand
-
Women have smaller hands and can access less of the screen with a shorter thumb
No Time to Spare
-
Add gamification
-
Inputting data is tedious and research showed that users want it to be fun
Need Simplicity
-
Easy and simple to not lose this demographic
-
Grandparents struggle to keep up with the technology their kids want them to use
-
Nannies have less excitement or motivation to track fastidiously
DESIGN PHASE
SOLVE FOR FAST INPUT
The path to fast input? Make it easy. I need to find a solution that works for all three of our demographics and can be completed in under 30 seconds.
POST-IT EXPLORATION
Rapid | Prototype | Sketches
Photo by Kat
BASIC ARCHITECTURE: MVP PHASE I
EXPLORATIONS OF TRACKING INTERFACE
.png){kind=tracking}
VALIDATE PHASE
RAPID PROTOTYPING
Narrow down the options and pick the best tracking interface.
.png)
.png)
.png)
.png)
.png)
FEEDBACK
Six users, including Moms, Dad's and allergy sufferers liked these two designs for the following reasons:
-
Fast​
-
Intuitive
-
Clean
.png)
.png)
USABILITY INTERACTION TESTS
INTERACTION FEEDBACK
Pros
-
Swiping interaction was extremely liked
-
Seeing the next card is helpful
-
Very fast
-
So simple = joyful
Cons
-
Confusion - users could never get a handle on which way to swipe the cards (even with a popup to explained)
-
Slow - users would scroll down to find the yes cards and leave the no cards which made for a slow, more confusing interaction
QUICK REVISIONS
.png)
SIMPLIFY SIMPLIFY SIMIPLIFY
.png)
.png)
.png)
.png)
SMALLEST SCREENS
Grandma's Old Phone Looks like This
.png)
.png)
.png)
LARGEST SCREENS
Mom's New Phone Looks like This
.png)
.png)
.png)
FINAL LIVE PRODUCT
.png)
NEXT STEPS
Phase I - Done
Design and launch the tracking and log functionality of the app on both iOS and Android app stores​
Phase II - In Progress
Add correlation functionality between food and symptoms and begin to analyze and improve sticking points from Phase I and incorporate into Phase II build
START TRACKING TODAY
.png)
