NoctIQ
They sent me a reference app to copy. I sent back a teardown of where it loses users, and won the project with it.
NoctIQ Inc. · founder Brian Schwartz

The artifact that won the project. Before I was hired, I redrew their reference app's onboarding as Current Flow against Improved Flow — a forced login with no explanation, against explanation first, skippable auth, and a home screen that shows what's locked instead of showing nothing at all.
- Role
- Sole product designer — research, IA, UX, UI, UX copy
- Timeline
- May — June 2026, refined through the July launch
- Platform
- iOS & Android · ~200 screens, 15 flows, 8 report types
- Status
- Shipped July 2026 · delivered in full
- Team
- Brian Schwartz (founder, product direction) · Nicola (product manager, formula validation) · me
They already had this product
Brian messaged me on LinkedIn. He had working hardware and a competitor's companion app his team was treating as the template. Instead of replying with a rate, I went through that app and wrote up what was wrong: it asked for a login before telling you what it was, and handed an empty dashboard to anyone who hadn't paired or paid. Both are places people quietly stop opening an app. I framed it around drop-off and what that costs a hardware company. That won the project — he wasn't shopping for screens, he was trying to work out why people wouldn't stick.
Snoring is invisible to the person doing it. You can't feel it, you don't remember it, and unless someone tells you, you've no idea whether last night beat the night before. So the app answers one question every morning: is this thing helping me?
The reframe:
You can't prove the pillow works until you've measured the user without it.
Three decisions
- 01
Made the five-night baseline visible — a countdown that only moves on nights that produced data, and holds on silent ones.
- 02
Let people skip the wait with their eyes open: force-connect is allowed, intervention still defaults off, and two plain-language confirmations stop it happening by accident.
- 03
Built one reading pattern for all six metrics: the answer, what it means, the evidence, and the maths folded into an accordion.
Detail




What got easier
Six metrics, three time windows, two intervention states. Designed screen by screen, that's eight unrelated dashboards you have to decode before breakfast.
- One number and a plain sentence, first. 66% — less snoring compared to your Baseline. Nothing above it, nothing competing with it.
- The sentence never changes, only the number does. You read it once, learn it, and after that you only need to glance at the figure.
- Locked metrics say what they need. A result that hasn't unlocked yet doesn't vanish and doesn't sit at the top looking broken — it moves down and states what's still missing.
- The maths is there if you want it. Folded into an accordion, invisible to everyone who doesn't.
What changed
“So grateful for your pivotal role in shaping the consumer app experience so far. Now onto perfecting it and bringing those membership features to life!”
Shipped July 2026 and rolling out to customers. Contracted scope delivered in full; membership features are version two. Brian shipped me a pillow so I could test against the hardware.
Want the reasoning behind each decision, and what I turned down?