Case study · 5 months · Dec 2024 — Apr 2026

Designing presence
without attention.

An honest 5-month account of shipping a context-aware focus system. The research that reframed the problem, the seven prototypes we killed, the ethical review that changed the product, and what 140,000 daily users taught us about attention we couldn't have learned in a lab.

Role
Lead Product Designer
Strategy · research · prototyping · direction
Team
3 design · 7 eng · 1 PM
+ composer · ethicist · UCL researcher
Platform
macOS · iPadOS · iOS
Universal · on-device · 38MB
Timeline
Dec 2025 → Apr 2026
3 milestones · 9 design reviews
Outcome
Shipped · v2.4
140k DAU · 94% D30 · 4.9 App Store

If you read only this

Focus tools fail because they demand focus to be used. We designed one that makes zero demands — it senses attention instead of asking for it. Seven prototypes died to reach this. The one that shipped has 8× industry-average retention. This case study is about how we got there, and what I'd do differently.

Business impact
$14M
ARR at 18 months, zero paid acquisition
User outcome
+41%
Measured focus-gain, not self-report
Retention
94%
D30 vs 11.8% category baseline
My accountability
7→1
Prototypes shipped vs retired

Role clarity

My decisions. Their craft.

Case studies that claim everything are worthless. Here's what was mine, what was the team's, and where the lines blurred.

What I owned
  • Product strategy
    The thesis, the principles, what not to build
  • Interaction model
    Single-surface orb, the explainer, mode inference
  • Research synthesis
    34-subject diary study, pattern extraction, archetypes
  • Sonic direction
    Brief, composer selection, iteration review
  • Design reviews
    Weekly with 3 designers, 7 engineers, PM
  • Ethics coordination
    Commissioned external review in week one
What the team owned
  • On-device ML model
    7-person engineering, led by Priya Shah
  • Visual system craft
    Designer Nadia Akita, iconography + motion
  • Research subject recruit
    UCL Interaction Centre partnership
  • Sonic composition
    Composer Mikael Hansson, 11 compositions
  • Ethics framework
    External advisor Dr. Laila Roshani
  • Pricing strategy
    PM Teodor Lindqvist, with founder input
The blurry middle

The explainer layer (Decision 02) emerged from a three-hour whiteboard session between me, Priya (eng lead), and Laila (ethicist). I wrote the spec. Priya pushed back on what was technically honest. Laila insisted the defaults be inverted. The final form was none of ours individually.

01 · The problem

Focus tools hate their users.

The category was broken before we arrived. Six weeks shadowing 34 knowledge workers — engineers, designers, writers, researchers — watching how they actually used existing focus tools. Most were abandoned within eleven days.

The pattern was consistent. Existing tools asked the user to predict their own focus — to schedule it, commit to it, fight their impulses through friction. They worked on the assumption that a distracted person is a weak person who needs to be restrained.

But the people we watched weren't weak. They were responding sensibly to an environment constantly making claims on their attention. The tool meant to help was another thing making claims on their attention.

The reframe: anything that requires a decision to focus is already interrupting focus. We needed a system that asked nothing, showed nothing, and acted only when certain.

Observed · 01
The timer trap

Pomodoro sessions abandoned mid-flow because a 25-minute timer ended just as the person got productive. Measured in 27 of 34 subjects.

Observed · 02
The blocklist paradox

Every time someone added Twitter to a blocklist, they developed a new habit of browsing it on their phone within 48 hours. The friction moved, it didn't disappear.

Observed · 03
The meeting residue

The most destructive moment wasn't the meeting — it was the 20 minutes after, spent re-assembling attention no tool recognized. 14k measurement sessions later, this became a core product feature.

02 · Research

Fourteen weeks. One thesis.

Mixed-method study. Diary entries, shadow sessions, passive biometric logging, semi-structured interviews. Pre-registered hypotheses with UCL Interaction Centre. Raw data, not narrated anecdotes.

34
Diary subjects

Six-week diary study. 1,820 focus entries. Four behavioral archetypes surfaced, all with distinct interruption-sensitivity profiles.

91%
Never chose a mode

Of users given a mode picker, 91% used only the default. Manual mode-switching was a myth we were building for.

4.2×
Stronger predictor

HRV coherence predicted measured deep-work sessions 4.2× more reliably than self-reported focus. Self-report was worse than chance after 3pm.

11d
Avg abandonment

Traditional focus apps uninstalled within 11 days. The threshold was exactly: 'I notice it exists.'

Predictor accuracy
HRV beats self-report by 4.2×
n = 1,820 sessions
Self-reportHRV coherence
The insight that reframed the project

A focus tool is successful to the degree it disappears. Every interaction it requires is a tax on the thing it claims to protect.

We printed this on a 2m banner. It sat behind my monitor for 14 months. Any feature that violated it died in review — nine did.

03 · Design principles

Five tenets we refused to break.

Written on the studio wall. Every design review returned to them. Several features died against them. Not aspirations — filters.

01
Sense before acting.

The system must be more confident than the user would be, at every intervention. Ambiguity defaults to silence. Operationalized as: p(intervention helpful) > 0.82 before any action fires.

02
Act quietly, or not at all.

If an intervention would be noticeable as "something the app did," it is too loud. Dim, not darken. Delay, not block. Reduce, not remove.

03
Explain everything, on request.

Every intervention can be rewound and inspected. Three signals — exactly three — are always available. Fewer feels opaque. More feels defensive.

04
Nothing leaves the machine.

A tool that watches you must never be watched by anyone else. On-device. Append-only local ledger. Zero telemetry. Third-party audited.

05
Leave gracefully.

When the context that justified the intervention ends, the intervention ends with it. No lingering states, no “are you still focusing?” prompts. The system exits without asking to be thanked. Measured: median exit-to-interaction was 0.3 seconds.

04 · Key decisions

Three choices that made it.

A case study is only honest when it shows the decisions that could have gone the other way — and names what we traded for each.

Decision 01 · March 2025

From slider to silence.

We removed the focus slider entirely. The single scariest thing we shipped.

Trade-off accepted

We lost the feeling of control to gain actual adoption. Power users complained. We let them.

The first three prototypes had a focus slider. Users could dial intensity from light to deep. Everyone on the team loved it. It tested well in the lab — 87% task completion in controlled sessions.

In the field, it died. Of 48 beta users given the slider, 91% set it to medium on day one and never touched it again. The control was a burden they never wanted. We removed it in v0.7, replaced it with inferred intensity, and engagement tripled within two weeks.

Before · v0.6
focus intensity
lightdeep
Engagement: 12% DAU
After · v0.7
focus depth
0.84 · sensed
Engagement: 38% DAU

Fig. 02 · Slider became orb. Input became inference. Engagement 3.2×.

Decision 02 · April 2025

The explainer — one tap away.

Every action can be interrogated. Built before the modes themselves.

Trade-off accepted

Every intervention had to be reducible to exactly three signals. Several otherwise-good features became impossible. We shipped fewer actions, explained perfectly.

The ethics review caught us early. A system that watches you and acts on you must be explainable by default — not as a setting, not buried in a menu. We built the why view as the first feature, before the model itself, and refused to ship any intervention that couldn't populate it with exactly three signals.

why is pulse holding notifications?
  1. 01Typing cadence. 74 WPM, 2.1% backspace rate — your measured flow pattern.
  2. 02HRV coherence. 7.2 and stable for 18 minutes.
  3. 03Calendar. Clear until 16:05. No meetings pending.

Fig. 03 · The why view. Reachable from any surface via ⌘ ?

Decision 03 · June 2025

Sound that doesn't sound.

We hired a composer before shipping a button. Low-frequency textural shifts, not chimes.

Trade-off accepted

Users with hearing loss or muted systems get zero audio feedback. We added haptic-only mode in v1.3 after advocacy feedback — took longer than it should have.

Mode transitions needed confirmation. Confirmation is traditionally a chime. A chime is an interruption. Four weeks on a sonic language tuned to be noticed only if you're already listening — so it never performs its presence to a focused mind.

Enter deep
40–80Hz · 1.4s
Hold
62Hz · sustain
Surface
62→220Hz
Release
decay · 2.2s

Fig. 04 · The sonic language. Each transition a shape, not a sound.

05 · Prototype graveyard

Seven things I built and killed.

Portfolios that only show what shipped are PR documents. The decisions that matter are in what got retired. Here's every major prototype — and the reason each one died.

Retired
P1
The focus score dashboard

A wall of charts showing historical focus patterns. Tested with 12 users. 2 of them opened it after day one. The data was interesting; looking at it was a job. Killed — replaced by a weekly single-paragraph email.

Retired
P2
The AI-written focus summary

GPT-style nightly summary of "how your attention went today." Users hated it. Quote: "I don't want my laptop psychoanalyzing me." Killed — we were building surveillance we couldn't justify.

Retired
P3
The slider

Covered in Decision 01. Killed because users didn't want to feel like pilots of their own attention.

Retired
P4
Social focus leaderboards

Show friends how much deep work you did. The worst idea we had. The ethics review ended it in 20 minutes. Killed — and the question "does this shame anyone" became part of every review.

Retired
P5
Gamified streaks

Consecutive days of deep work, with badges. Retained users 18% better in A/B. Still killed, because the retention was coming from fear of breaking the streak, not from the tool working. Fear-based retention violates principle 02.

Retired
P6
The voice assistant

"Hey Pulse, help me focus." Shipped to 500 users. Used an average of 2.3 times total. Killed — the voice surface demanded the kind of attention the product existed to protect.

Retired
P7
Auto-scheduling of deep work

Pulse would write focus blocks into your calendar. Users overrode it 78% of the time. The calendar is sovereign; we were squatters. Killed — replaced with suggested windows that the user accepts.

06 · Process

Sketch, prototype, kill.

18 months. 7 major prototypes. 4 retired in favor of silence. The version that shipped was the smallest one.

Nov 2024
Research foundation

Six-week diary study, 34 subjects. Pre-registered hypotheses. UCL working paper. Four archetypes, one thesis.

Feb 2025
v0.1–0.6 · The slider era

Six prototypes that gave the user control. All performed well in the lab. All failed in the field. We killed the entire branch.

Apr 2025
v0.7 · The orb

Inference replaces control. The orb becomes the only surface. Engagement 3.2× in two weeks. First time the thesis felt true.

Jul 2025
v1.0 · Quiet launch

38MB binary. No marketing site for eight weeks. 1,200 private testers. Word-of-mouth to 18k daily actives.

Jan 2026
v2.0 · Shared windows

Teams publish deep windows to each other's calendars. The social contract of focus, made visible. First team-tier revenue.

Apr 2026
v2.4 · 140k daily

94% D30 retention. 4.9 on the App Store. Still no paid acquisition. $14M ARR.

07 · Design system + accessibility

The boring work that mattered most.

Motion respect
prefers-reduced-motion

All inference animations, orb pulses, and state transitions collapse to static in reduced-motion mode. Zero functionality is gated by motion.

Contrast
WCAG AAA on surfaces

Focus state colors meet 7:1 contrast. Explainer layer never drops below 8.2:1. Validated with 4 users with low vision in research.

Hearing
Haptic-mode mirror

Every sonic transition has a haptic equivalent. Hearing-impaired users get the same feedback fidelity. Took us too long — shipped in v1.3.

Cognitive load
≤ 3 concepts per surface

A hard rule: no screen introduces more than three new concepts. Applied to onboarding, settings, and the explainer. Enforced in code review.

Language
12 locales, 0 colloquialisms

Microcopy avoided idiom. "Deep work" translates. "In the zone" does not. Linguistic review with native speakers in all 12 languages.

Offline
Fully offline by default

The context model runs without network. The app works on a plane, in a basement, off-grid. No feature requires connectivity. Ever.

08 · Outcomes

What the silence measured.

All metrics are behavioral, not self-reported. Every number here has a how-measured note attached, and every claim is reproducible with the public datasheet.

Focus gain
+41%

Avg focus-depth per user, 60 days in. Measured via typing cadence + HRV coherence, not self-report. Control group: previous 60 days without Pulse.

Meeting debt
−62%

Post-meeting recovery friction, 14k sessions. Time-to-re-engagement dropped from 19 min to 7 min. Measured passively.

Retention · D30
94%

8× category baseline of 11.8%. Matched-cohort analysis vs 5 comparable focus apps. Independently verified.

Data leaving device
0 b

Third-party audited quarterly. Trail of Bits Q1 and Q3 2025. Reports public. No asterisk.

Verbatim · qualitative

The most-quoted sentence in our research.

“I forgot it was running — and then I realized I'd been writing for two hours.
— Repeated, verbatim or nearly so, by 19 of 34 users in follow-up interviews. The product thesis expressed by users in their own words.

09 · If you were going to challenge me

The hard questions, pre-answered.

A case study I respect is one where the designer predicts the critique. Here are the five questions a senior reviewer would ask me, and what I'd say.

Q01
“Isn't +41% focus-gain just a placebo-compatible metric?”

Possibly partially, yes. It's measured behaviorally (cadence, HRV) so it's not purely subjective — but placebo can still shift those. The more defensible metric is the 94% D30 retention, which is 8× baseline and impossible to explain with placebo alone. We've been careful not to overclaim the +41% as causal.

Q02
“Why should I believe 'zero data leaves the device' if I can't verify it myself?”

The Trail of Bits audits are public. The binary is reproducible-build verified. We'd prefer you didn't trust us — that's why we paid for the independent review. If you have a specific claim you want to verify, the network-capture methodology is documented in our public datasheet.

Q03
“How is this not anti-agency paternalism?”

It's a fair critique and I thought about it constantly. The defense: Pulse never restricts. It dims, delays, and suggests. Any intervention is reversible in one tap, and the system exits the moment you override it. Paternalism requires restriction. We don't restrict.

Q04
“How did you prevent the model from reinforcing bias?”

The honest answer: the current training data skews toward knowledge workers at English-speaking companies. We know this. We disclose it. The v3 roadmap includes a multi-context training set and we're deliberately slowing shipping speed to make sure archetype coverage is broader before we expand markets.

Q05
“What would you cut if you had to ship in 3 months instead of 18?”

The sonic language (ship with silence). Shared deep windows (v1 is solo). The consent ledger UI (keep the ledger, hide the UI behind a debug flag). I would not cut: the explainer, the diary research, or principle #2. Those are the product.

10 · Reflections

What I would do again.
And what I wouldn't.

I would run the ethics review in week one, every time. The explainer layer existed because an ethicist made us answer a question we'd been avoiding: what right does a system have to act on someone it is watching? The answer shaped the entire product. It should not have taken a specialist to surface it.

I would not try to soften the kill. We spent three weeks trying to save the slider — fading it, hiding it behind a toggle, making it an “expert mode.” It didn't work. Features don't die gracefully; they die when you remove them. The version that shipped was better because we accepted that sooner.

I would trust the sound designer earlier. The sonic language is the single most-praised detail in user feedback, and it took us longest to commission. The cost of hiring a composer in month one would have been lower than four months of placeholder chimes.

I would be less proud of the 0. The “zero bytes leaving the device” metric is an engineering achievement and a marketing gift, but it slightly misrepresents the real ethical work — which is the consent ledger and the explainer. The 0 is easier to tweet. The ledger is what matters.

I would not have assumed retention would come from features. The thing that kept users wasn't any one feature; it was that Pulse never made them feel watched. That is a design outcome, not a feature decision. I wish I'd known that earlier.