Store-carousel study of the top 3 Android and top 3 iOS alarm apps — every carousel frame saved locally, descriptions read in full — compared honestly against the Universal Timer session that scored 7/200. The uncomfortable finding: the carousels of the #1 apps are literally Pawan’s wizard spec — one idea per screen, benefit-voiced headline, compact controls, accent-discipline — the exact instructions the session ignored.






One frame = one idea, six frames = an emotional arc: promise → gate → sound → personal → trust → morning-after. This is, almost beat for beat, the Intention → Confidence → Trust → Gentle Awakening arc Pawan specified after hop 6. The #1 alarm app on earth structures its story the way Pawan structured his wizard.






Breathing room is the aesthetic. Every frame is ~40% empty sky. Nothing touches the edges; the phone itself floats at an angle with generous margin — the visual equivalent of the “each screen ≤ 75% of the viewport” constraint the timer build violated by cramming.





The product is a stage, not a form. Its hero surface is a full-screen nightstand clock — an app that fills one viewport and never scrolls. The oldest app of the six understood “it’s an app, not an article” a decade ago.








The sound picker gets two whole frames. Choosing and previewing your alarm sound is treated as a headline feature — the same feature Pawan had to demand (“let the user PICK the sound”) before the timer build included it.














Even 100s of features are sold one per frame. The most feature-dense alarm app in existence still never shows two ideas in one frame — 14 frames instead of one crowded frame. Feature density is expressed as MORE SCREENS, never as cramming.














The cautionary tale of the six. Its first frame is a stacked text-card feature list — the closest thing in this study to the timer session’s “one giant crammed card”. It is also the least-loved app of the six. Cramming correlates with losing.
| What the session did (POST_MORTEM.md) | What every top store carousel does | Verdict |
|---|---|---|
| Crammed the whole wizard on ONE page, Set button pushed off-screen; ignored “multiple screens” twice. | One idea per frame, always — even 100-feature Sleep as Android uses 14 frames rather than one dense one. Only weak AMdroid stacks a text list, and it trails the pack. | Pawan’s spec = the industry grammar. The session fought the market standard, not just the client. |
| Train-size buttons and an orange slab CTA. | Compact pills and small toggles inside the phone; Sleep Cycle keeps orange to a hairline ring; Alarmy’s red exists on exactly one button. | “Orange is an accent” isn’t taste — it’s how the category leaders paint. |
| Invented a palette (--paper/--amber/--hair) → invisible borders, strange text. | Each app has ONE locked brand system used with total consistency across every frame (Alarmy blue, Sleep Cycle sky+dark, SaA indigo). | Brand systems are frozen assets — same as “gold means gold”. |
| App scrolled like an article; full mega-nav header left on. | Every screen shown is a single-viewport stage — bedside clocks, mission screens, score rings. No alarm app screen scrolls; no chrome beyond a slim bar. | “Apps never scroll” is visible in all 53 frames. |
| Sound choice arrived only after Pawan demanded it. | Alarmy devotes 2 of 8 Android frames to sound browsing; ACFM leads with “Wake up to your favorite music”. | The user’s first request was the market’s headline feature. |
| Machine copy: “Set timer / New timer”. | “You’ll wake up. Always.” · “Prove you’re up or make it ring again” · “Wake up refreshed”. | Felt copy is not a nicety; it is the category’s selling voice. |
| 2-minute timer jumped to tomorrow (time-model bug). | Alarmy’s home shows “Ring in 7hr 30min” — countdown-to-ring is the primary reassurance line of the whole category. | Duration-first thinking is the standard mental model; the fix was the right call. |
| Engine: 3 gates, R1→R2→R3 ladder, anti-relapse Wake-Up guard, SSE sync, offline firing — verified. | Missions, snooze-limits and Wake Up Check are precisely the features the leaders lead with. | The one part built without ego is the one part that is genuinely competitive. |
The 7/200 session didn’t fail because the product idea was hard. It failed because the build ignored a presentation grammar so universal that every single one of the six market leaders follows it: one idea per screen, felt headline, compact controls, one disciplined accent, 40% air, sound as a hero, proof over adjectives.
Pawan’s wizard spec — multiple screens on an emotional arc, ≤ 75% viewport, no train buttons, pick-your-sound, amber as accent — wasn’t a personal quirk to be second-guessed. It was a market-accurate description of how winning alarm experiences are actually composed. The instinct that overrode it was wrong twice: wrong against the client, and wrong against the category.
What survives with pride: the engine. Gates, ladder, anti-relapse, SSE, offline firing map one-to-one onto the differentiators Alarmy and Sleep as Android lead their carousels with. Fix the P0 stylesheet, then rebuild the wizard to the grammar above — which is simply Pawan’s spec, executed literally.