# M2 v2: Her alarm got her sister up

Disposition: ready for disclosed illustrative draft; exact-product outcomes not claimed; launch approval separate.

## Decision and writing contract

One roughly 1,000-word problem-aware native later, not an ad body now. Preserve original master brief, gold-standard cadence and Alevia/SOP functions with the current evidence and disclosed-fiction authorization. Three hook directions, two headline directions and five image directions belong to this one architecture.

One desire: calmer mornings, specifically a waking attempt that does not enlist the other child. Experience: the intended wearer's alarm wakes her sister down the hall; that sister goes in to wake her. Emotion: frustration/protectiveness is a strategic inference, not the respondent's literal free-text emotion. Behavior: the household supplies human help after the alarm wakes the wrong person. The sister's intervention is explicit; Mom's exact reaction is proposed fiction. One buying argument: try delivering the scheduled cue against the intended wearer's wrist rather than broadcasting it down the hall.

Angle: relationship-wound, sibling spillover. Delivery: disclosed illustrative personal story. L2 problem-aware, proposed eventual cold Meta. S3 comparison depth is a creative hypothesis, not measured market sophistication. This is the sibling-tax reframe, not a claim of fresh performance selection. Do not make the school-ready sister wait during dressing; the actual supported cost is interruption of her sleep and recruitment into waking her sister. No diagnosis, sibling age gap, school schedules or family structure beyond the source-owned sister relationship.

## Exact evidence spine and context

Primary **survey-R**, Q1, August 14, `revision-v2/private/user-survey-source.txt`, line 249:

> She set an alarm that woke up her sister in the room down the hall. Her sister came in to wake her, and she still didn’t make out of bed for 2 more hours. Everyone was up because of her alarm except her.

This is one coherent account, not a collage. Keep the awkward “make out of bed” intact when quoting. The source reports an alarm heard elsewhere, an actual sibling intervention and two more hours before getting out of bed. It does not establish that she stayed unconscious for those two hours, that a band would keep her up, or that her sister missed school. The dropdown says “him”; the free text says “she/her.” Free text controls. Source is an owner-provided postpurchase export, not independently database authenticated; dropdown classifications are not quotes, diagnoses or proof of emotion.

Separate **survey-C**, Q1, August 20, line 221:

> Want to start the school year off with out the fighting

This corroborates the desire only. It is not survey-R's speech or evidence that her children fought that morning.

Separate category counterexample **`iu7et9x`**, `research/lane-index.json`, self-consideration, public https://www.reddit.com/r/ADHD/comments/yfxmda/advice_on_how_to_wake_up_earlier/iu7et9x/:

> I got a Fitbit to wear at night. Not a fancy one, just a Charge. I set it to start going off an hour before my alarm does. It just vibrates so it doesn’t bother my wife.

This archived adult self-report supports an existing wearable alternative and reported household fit, not teen or Dawn efficacy. Full mapped account adds other tricks have been outsmarted; do not claim every alarm worked permanently or import a spouse into the story. Original receipt: `knowledge/brands/dawnbands/research/adhd-self-buyer-voc-2026-09-05/threads/thread_yfxmda.json`.

Separate after-waking boundary **`1j429m9`**, mom-sa2, https://www.reddit.com/r/ADHDparenting/comments/1j429m9/morning_routine_despair/:

> Each step requires at least twenty repeated instructions, and no step gets done until my husband or I raise our voice.

The raw `knowledge/brands/dawnbands/research/reddit-mornings-2026-09-03/thread_1j429m9.txt` names breakfast, teeth, socks, shoes and coat, and a younger child. This is not waking failure and must not become M2's rupture. Its richer family story remains valid research without being used to manufacture a product bridge.

## Proposed narrator and physical climax

Introduce fictional **Nora** as a mother of two daughters whose bedrooms are down the hall from each other. Proposed natural introduction: “I'm Nora. My daughter's alarm woke her sister down the hall.” The family relationship and layout follow R; the identity and narrator voice are illustrative, not a renamed customer. No job, town, age, clinician or partner added.

Required prominent disclosure above a later first-person draft: “Illustrative story inspired by a parent survey response. Names, dialogue and connective scenes are fictional. This is not a customer results testimonial.” Do not place it only in hidden review metadata.

World-building stays on the sound and the unintended helper. The alarm is set for one daughter. Her sister hears it in a different room. She leaves her room to wake its owner. Nora watches the waking task recruit another person. The watching is proposed staging; do not represent it as part of the survey. No invented history of daily sleep deprivation. Anticipation is proposed Nora listening to the alarm and hoping its owner gets up before her sister does. Before-state: the reasonable expectation that the daughter's alarm is her own wake-up cue, not a fabricated earlier happy household.

**One climax:** the sister stands at the intended wearer's bedside after getting up because of the alarm, while the intended wearer is still in bed. Nora tells the sister she does not have to keep doing this. Proposed line: “You can go back to your room. I'll help her.” No invented angry response, tears or humiliation. The source-owned escalation continues: intervention has happened, but the intended wearer still does not get out of bed for two more hours. In the illustrative body prefer “she was still in bed long after that” unless the source's exact two-hour detail earns its space and stays narrator-only. This delay is not proof of continued sleep.

End state: the other daughter has already had her sleep interrupted and helped. Nothing later can give that morning back. Nora's remaining wish is not to recruit her next time. This makes the PS callback coherent. No story about an unfinished drawing or sibling waiting with a backpack.

## Discovery and belief turns

Proposed discovery chain: later Nora recounts the incident in a Facebook post; **June**, a parent friend, asks which person actually responded to the alarm. June offers to compare alarm setup pages with her. On a call, they read the product listing and current on-band instructions together. June's access to specifications is the page open in front of her, not a medical job or an invented household success. This is an illustrative conversation, not a claim that survey-R posted publicly or received June's recommendation.

The dialogue does a different job from M1:

1. Nora assumes a loud enough alarm should solve the problem. June asks who heard this one. Nora names the sister. This source-owned contradiction distinguishes audible volume from a successful wake-up for the intended person, without asserting why her brain did not respond.
2. Nora considers having the sister wake her sooner. June acknowledges that she already tried to help. Nora realizes she is assigning the same interrupted sleep and added job again. This is a fairness decision, not clinical advice.
3. Nora asks if closing doors or reducing volume will fix it. Treat these as considered alternatives, never past failures. They might reduce spillover but could also leave the intended wearer without an effective cue; neither is established as a complete solution here. Avoid declaring doors acoustically useless.
4. Nora separates “the alarm didn't get her up” from “she was still in bed after help.” The first is a reasonable cue test; the second still needs observation and support. She need not abandon trying a different cue because it cannot settle every later action.
5. At the late reveal, June says the category aloud, proposed: “A wrist alarm. This one uses vibration.” Nora asks where the time is set; June names “Dawn Bands” while pointing her to the page. The product answers the person-specific delivery question already earned by the story.

Earliest meaningful sales reveal at least 70% of narrative-body words, target 75–80%; count any earlier recommended tactile alternative as reveal, not only brand. Plan around word 750 of roughly 1,000, with no fixed 70-word feature box. Include close/signature/PS in denominator, exclude disclosure/headlines/review notes. Use the pre-reveal space for the source-owned contradiction and Nora's changing assumptions, not an early stealth pitch or generic sleep lecture. Reserve the connected last section for useful comparison and offer rather than filler postscripts.

## Why this product is relevant and what it cannot do

Product receipts: `product/product.json`, price 3999 minor USD units and wearable-vibration description; `knowledge/brands/dawnbands/product-truth.md`, owner facts lines 33–41; `product/refund-policy.txt` for policy. The JSON's “wakes you, not the household” marketing line is not a universal outcome receipt.

Three differentiated factual reasons to give after the spoken reveal:

1. Scheduled vibration against the wearer's wrist instead of an audible alarm tone broadcast into the hall: directly relevant to the sister hearing the room alarm. Say it does not use an audible alarm tone; do not promise absolute inaudibility or that no sibling can ever notice vibration.
2. On-band setting with no app or pairing: the daughter can handle the scheduled alarm without someone else's phone connection. This is operation, not proof of responsibility or compliance.
3. $39.99 for a dedicated wrist alarm: a concrete purchase to compare against using a working watch already owned, not an unsupported “cheapest” claim. If an existing watch works, keep it. No invented savings or universal smartwatch failure.

The owner-confirmed stronger motor is available as a secondary factual point if a watch comparison needs it; it is an owner claim, not a measured wake-rate advantage. Do not say stronger vibration guarantees quietness. The vibration slowly gets stronger until dismissed, but dismissal does not mean the wearer is standing.

Causal chain: room alarm → sister hears it → sister supplies waking help; wearer-specific scheduled vibration → if effective for the wearer, an audible alarm and sibling intervention may be less necessary → calmer shared mornings remain desired. No claim of improved sleep, cortisol, mood regulation, school attendance or two-hour-delay cure.

Wear objection: ask the intended wearer to try the fit and set it herself; do not buy secretly and strap it onto a sleeping child. If she won't wear or charge it, it cannot give the scheduled cue. Efficacy objection: retain a reliable backup while testing and judge both waking and getting up. Don't appoint the sister as that backup. Longer delay after waking deserves separate attention; don't medicalize or dismiss it as laziness.

## Desired future, offer and close

Conditional future pace: if the daughter wakes to her own cue and gets up, her sister would not need to leave her room to help. On later ordinary mornings Nora would want not to hear that sister walking down the hall because of someone else's alarm. These are desired absences, not achieved first-morning or eight-week outcomes. No invented sister's thank-you.

Introduce Dawn Bands as the seller of the standalone wrist alarm; no fabricated founder, small-batch story, review count or expert endorsement. Price $39.99. Return direction: “You can request a return within 60 days of ordering; conditions and return shipping costs apply.” The material conditions remain available at https://dawnbands.com/policies/refund-policy: undamaged/same condition, original packaging, proof of purchase, inspection and approval; customer pays return shipping. The advertised 60-night label conflicts with the written order-date policy. Do not reopen Shopify, stall the brief, promise 60 nights after delivery or insert a policy lecture into June's speech.

CTA #1: inspect the actual setup together at https://dawnbands.com/products/wake-up-band-adhd. Proposed destination, claim congruence/launch approval separate. Close returns to the sister's interrupted morning, not a threat that failing to buy harms children. Signature Nora, explicitly fictional.

P.S. returns to the sister in the hall: the desired change is that she stays in her own room. P.P.S. answers financial risk using concise policy access, not another emotional callback. P.P.P.S. answers “she'll still lie in bed” by distinguishing a wake cue from the later decision to stay up, framed as what Nora will observe. Final reader-address coda speaks to an alarm waking someone else before its owner, then a second bare product link. No scarcity coda or fabricated urgency.

## Three hook directions

- H1 first-person Loop: one daughter's alarm gets the other daughter out of bed. Protective frustration; stakes are another child's interrupted sleep. R source, immediate family/waking identification.
- H2 symptom-flag Loop: an alarm heard down the hall while its owner stays in bed. Recognition; withhold the mother's next decision, not who the story is for. R source.
- H3 first-person Loop: Mom finds the sister doing the waking job. Source-owned action with disclosed Nora witnessing; tension about whom the alarm is serving. No invented child quotation in the hook.

Two headline directions: T1 reinforce wanting the alarm to reach its intended wearer; T2 reinforce not making a sibling do the waking. Both sentence case, desired/consideration framing, no achieved “everyone sleeps now.”

## Five image directions and metadata

I1 straight-rip ordinary hallway connecting bedroom doors, R layout. I2 help-post room photograph taken from another bedroom doorway, R spillover context. I3 generic alarm clock near the far bedroom opening, proposed staging: source does not establish clock model. I4 cropped adult view of two open bedroom doors, proposed Nora viewpoint, no people. I5 contradiction composition in one casual photo: alarm visible in one room while the empty doorway of the other occupies foreground, proposed interpretation, no added text. Each is a direction, not a sourced/cleared image.

All: rip first using ordinary queries such as “bedrooms down the hall”; real image rights and child privacy checks required before use. Phone-photo register; no product, overlays, fake Facebook UI, staged sadness, identifiably sleeping minors, coffee or clinical props. Reference-based generation only if no viable cleared rip, and only in a separately authorized image task. Do not generate now.

## Voice and acceptance

Voice: matter-of-fact protective parent, not a speech about fairness. Contractions and ordinary words, grade 3–5 target, average under 14 words and no sentence over 28 in the later body; 2–4-sentence Alevia paragraphs, dialogue isolated. VOC anchors “room down the hall,” “Her sister came in to wake her,” and “Everyone was up because of her alarm except her” are R's exact evidence. Proposed voice examples, not customer quotes: “Her sister had already gone in to help.” / “I didn't want to make that her job again.” Don't turn “sibling tax” into narrator jargon.

Narrow remaining constraints: keep visible fiction disclosure; preserve waking-versus-getting-up distinction; check actual reveal timing and readable sale in the draft; verify cleared images and destination claims at launch. The new source resolves the old after-wake incident mismatch without claiming a product outcome.