Nularo

Timeline

2026, live on the App Store

Role

Designer and builder, end to end

Scope

iOS app, character system, subscription

Nularo is a live iOS product that turns Apple Health into a character with an opinion. It reads your sleep, steps, resting heart rate and HRV, and gives you one honest verdict every morning. I designed it, wrote its voice, built it, and shipped it to the App Store, end to end.

Live on the App Store, shipped solo, at $39.99 a year
A wellness tracker built to make habits stick, where a character you care about carries the feedback instead of a chart
Zero logging. It reads Apple Health on its own, so the habit is opening it, not feeding it

Milestones scrolling into Overview, the creature, and the age it keeps. All three are the shipped iOS app, recorded on device.

The product

A creature that ages on your habits, and never asks you to log anything

It reads Apple Health on its own. You meet it once, and after that it just has opinions.

Where the years came from. Sleep duration, bedtime consistency, daily movement, active minutes and heart, each carrying the years it added.

You meet it before it has read anything.

What it reads: sleep, bedtime, quality, debt, resting heart rate, HRV.

User research

What people say when a tracker stops working

I ran 32 interviews, then tested what they told me against 1,692 App Store reviews pulled from Apple's own customer-review feed across the eight trackers people actually use, taking the 561 rated three stars or below.

Who wants to spend 10 mins or longer per day logging their sleep…? They have a bunch of stuff you have to click through daily it’s just obnoxious and too active.

You have to plug in all your own habits. The available options are limited. It’s basically a blank slate.

Customers get asked for more input to feed profit margins. Tiring

It automatically marks days as completed streaks unless I open the app and manually change it. Which forces me to remember everything. Completely defeats the purpose of this app.

My tasks keep randomly disappearing and I have to constantly re-add them, not to mention I lose my streak. This is the whole purpose of the app…

Every time I set this up and start using it something goes wrong… I cannot trust it to work properly and I can’t trust the statistics.

…it can feel like a chore after I while. Im on my 298 day and it’s a lot of work to keep up

This app actually helps me get things done, and without making me feel bad!

Theme 01

Input is the drop-off

Theme 02

A character outlasts a chart

Theme 03

Streaks punish more than they motivate

User pain points

Every tracker asks for the one thing its user has already failed at.

The category treats logging as the easy part and behaviour change as the hard part. The reviews say it is the other way round. Three pain points, each with the number underneath it.

<50%of participants in a twelve-week dietary self-monitoring study were still logging by week ten

1. The logging is the habit people fail first

A tracker that needs a daily entry is measuring compliance, not health. The people who most need one are the least likely to make it, so the app quietly selects for the users who needed it least.

Who wants to spend 10 mins or longer per day logging their sleep…? They have a bunch of stuff you have to click through daily it’s just obnoxious and too active.

77%of health and fitness app installs never come back after day one, on the published category benchmark

2. Setup is where the drop happens, not week six

Most trackers open on an empty state and ask the user to build the thing they came for. That is a configuration task handed to someone who arrived with a problem.

You have to plug in all your own habits. The available options are limited. It’s basically a blank slate.

298days in, and the character had become another thing to keep up with

3. A character fixes motivation, not effort

Pet mechanics hold attention until the pet needs feeding. At that point the character is one more daily task, and the thing that was meant to carry the habit becomes the habit.

…it can feel like a chore after I while. Im on my 298 day and it’s a lot of work to keep up

Sources: dietary self-monitoring adherence, Journal of the Academy of Nutrition and Dietetics, 2019. Day-one retention is an industry benchmark rather than a measured figure and is labelled that way on purpose. Review counts are my own pull from Apple's feed on 2026-09-03.

Framing the solution

Mission

Nularo asks for nothing daily. The data already exists in Apple Health, so the problem was never collection. It was consequence.

From that, three questions the product had to answer. Every screen got argued against them rather than against a feature list.

How
might
we

How might we build a health app that still works for someone who never opens it?

How might we turn numbers nobody reads into a consequence nobody can ignore?

How might we judge someone honestly without ever telling them what to do?

The mechanic

Your creature has a biological age. It is usually not the age you are.

It is a biological age, not a birthday. The creature starts at your real age, then five Apple Health inputs are each scored against a published guideline: sleep duration, bedtime consistency, daily movement, active minutes and heart. The gap on each one becomes years added or taken off.

Mine reads 27.9 against my actual age, and every one of those years is attributed rather than asserted.

It moves in both directions. A plus adds years to the creature and a minus takes them off, which is what makes it a live reading rather than a score.

The number, and the gap. It settles on a seven day cycle.

Every year is attributed, so the verdict is arguable.

This is the answer to theme 01 and theme 03 at once. There is nothing to log, so there is nothing to fall off, and there is no streak to break because the number moves in both directions on its own.

Design decisions

Three decisions did most of the work. Each one started as a constraint, not a preference.

These are the constraints the design ran into and what shipped because of them. They are build decisions rather than usability-test findings, and they are labelled that way because no test was run.

The shipped creature, petal crown carrying the mood

The silhouette took twenty four tries

The constraint. The creature had to be recognisable at tab bar size and still able to carry six moods. Blobs read as generic, wisps read as sad, and most explorations lost their face entirely once they were 28 pixels tall.

What shipped. A crown of petals that droop or perk with mood. The petals carry the expression, so the body stays simple enough to survive being small.

One number is not an argument

The constraint. A single wellness score is unarguable, and an unarguable number is one you stop reading. The app needed a verdict the user could push back on.

What shipped. Every year is attributed to a habit. Daily movement +2.0y is a claim you can check against your own week. A composite score is not.

Apple Health settles sleep a day late

The constraint. Sleep is not final until well after you wake. A live reaction would either be wrong or would have to sit there waiting, and both read as broken.

What shipped. One verdict a morning, filed under a mood and kept. The lag became the format instead of a bug to apologise for, and the log is what makes the creature feel like it has been paying attention.

Product judgment

A character with a real opinion, and the judgment to know exactly where that opinion stops.

A creature that reads your heart rate and your sleep could say almost anything. Deciding what it would never say is what gave it a voice people trust, and I set that line before I designed the first screen.

A personality, not a prescription
It can tell you that you slept badly and be funny about it. Keeping it out of advice is what keeps it a character you want to open, rather than one more app telling you to go to bed earlier.
One verdict, on a night that has finished
The morning read lands on settled data. Committing to that turned the product into a daily ritual instead of a number that keeps moving while you watch it.
Privacy as a feature, not a policy
It reads Apple Health and nothing leaves the phone but the few numbers behind each verdict. For a product that knows how you slept, that is a reason people pick it.

The permission screen, where the product says what it reads and what never leaves the phone.

Reflection

I designed, wrote, built and shipped a live iOS product end to end, from the research through the character system to the subscription. It is on the App Store, it is priced, and every decision in it is one I can account for.

What it taught me:

Build a product people actually want, solving a real need. Thirty two interviews told me what breaks, and eight competitors had 1,692 people writing the same thing down in public, for free. Between them there was no guessing left about whether the need was real before I built anything.

The constraint was the feature. Apple Health being a day late looked like a platform problem to work around. Turning it into one verdict a morning gave the product its format, its ritual, and the reason it never nags.

A character earns attention a dashboard has to beg for. Version 1.0 took its first five-star review before a single piece of marketing existed. People came back for the creature, which is exactly the bet the whole product was built on.

Next project

Canary

All work