Tracto
One place for the things you track.
Water, food, vitals and exercise in one app, with reports worth handing to a doctor. A self-initiated project, from a short survey to 54 screens.
The problem
People jump between apps to count steps, log meals, check vitals and find a doctor. The tracking becomes the chore, and then it stops.
What I did
A survey and five conversations, four written constraints, a design system and 54 screens, with 22 of them spent on logging and reports.
The call
No streaks. A missed week is when you most want the readings, so Tracto counts the days you did log, and that number only goes up.
01The problem
The tracking becomes the chore, and then it stops
People jump between apps to count steps, log meals, check vitals and find a doctor. Each one asks for the same information again.
"There are too many apps doing the same thing."
Riya, 27, product manager
02Research
Six findings, two of which changed what I built
A short online survey, then five conversations. The survey was small, so I treat it as directional and haven't put percentages on it.
The phone is the tracker
Watches came second. Whatever I designed had to work with thumbs and no hardware.
Water, food, vitals and exercise, together
No single metric dominated. People wanted the set, which is the argument for one app.
Doctors are occasional
Most said only when needed. Booking is not the daily loop, so it should not sit at the centre of the app.
Ease of use beat every feature
Ranked above wearable sync and personalisation. That is a mandate to cut features.
Symptom-first made sense to people
Choosing a specialist by name assumes you already know what is wrong.
Appetite for one app was high
Encouraging, and also the easiest thing in a survey to say yes to.
Insurance advisor, medium confidence with apps
Wants to log blood pressure and sugar regularly and be reminded. Forgets to log readings by hand; most apps feel complex.
Product manager, high confidence with apps
Tracks steps, water and calories daily and wants a routine she can keep. Has no time to log manually and wants charts that mean something.
03Where the product lives
The screen that demos well is rarely the screen that matters
Five modules, and they are not equal. Onboarding runs once. Booking a doctor happens a few times a year. Logging happens several times a day, every day, and it's the only one that has to survive being boring.
- LoggingSeveral times a day. Water, food, vitals and exercise.
- ReportsMonthly, or before an appointment. Trends, ranges, export.
- DoctorsA few times a year. Symptom quiz, search, booking.
- ProfileRarely, after setup. Targets, reminders, privacy.
- OnboardingOnce. Welcome, goals, wearable sync.
Bar lengths are illustrative. They show the order of how often each module is used.
22 of the 54 screens are logging and reports, and that's the right proportion.
04What it won't do
Four constraints, written before drawing anything
I wrote these first, so later decisions had something to fail against. They ship in the app, on the About screen.
Won't tell you what your numbers mean
The app shows the reading and the trend and stops there. No risk scores, no colour-coded verdicts. Out-of-range readings say so plainly and point to a doctor.
Won't move your data off this phone
Storage is local. Export is a deliberate act with a send button, and nothing is kept afterwards.
Won't punish you for a gap
No streaks, no broken-chain graphics, no guilt copy. A missed day shows as a gap in the data.
Won't sell you anything
No supplements, no plans, no sponsored doctors, no ads, no sale of data. Reminders exist only to help you log.
The decision behind "no streaks"
Two lines sit next to each other in Riya's empathy map: she's motivated by streaks and progress, and she feels guilty when she misses logging for a few days. Both describe the same mechanism: a streak motivates precisely because breaking it costs something.
For a step counter, that trade is probably fine. For blood pressure it isn't. The person most likely to miss three days is the person having a bad week, and a bad week is exactly when you want the readings.
So Tracto has no streaks. It counts the days you did log, and that number only goes up.
05The product
The reading is the product. Everything else is packaging.
Logging is the daily loop, so every interaction is built to finish in one tap wherever it can. Water is four buttons, with no keyboard unless you want one. Blood pressure asks for three numbers and the time, and nothing else. Quick-add exists because the survey put ease of use above every feature, and because Rajesh said he forgets when logging takes too long.
- Health
Tap anything to log it
Every metric sits on one grid with today's value, and recent entries underneath.
- Blood pressure
Three numbers and the time
A reading above the usual range says so, says it can happen for ordinary reasons and that Tracto won't tell you what it means. Your doctor can.
- Day one
Nothing logged yet
Start with whatever is easiest: one glass of water counts as a log. The home screen says readings never leave the phone unless you export them.
06Reports and doctors
What the logging is for
Reports hand a doctor something useful. Booking stays at the edge.
07Structure and system
Five tabs, one blue
08How I used AI
A second pass over my synthesis
Where it helped
- After the five conversations, a second pass over my anonymised notes to check I hadn't missed a theme.
- Drafting alternative wordings for out-of-range readings, so I could pick the one least likely to read like a diagnosis.
What stayed mine
- The empathy-map insight about streaks, which came from reading Riya's lines side by side.
- The four constraints, and the call to put 22 of 54 screens into logging and reports.
- Naming what this case study can't prove.
09What I would test next
What the next round would test
There is no usability testing in this case study, because I haven't run it. Every decision is argued from a small survey, five conversations and a look at the apps people already use.
Streaks against days logged
The obvious counter-argument is that streaks work, which is why every habit app uses them. Run both for eight weeks and see which group still logs in week seven, and how each describes the app after a missed week.
Where "not diagnosing" stops helping
Someone who sees 148 over 94 and is told only that it's above their usual range may find that useless. I'd test where people put the line between refusing to diagnose and refusing to help.
See the full case study
The complete Tracto case study, with the empathy maps and every screen, is on Behance.