Community supermarket marketing planner - promo launch readiness early-warning; T-3 checks stock, pricing tests, materials and vendor reps, T-1 checks the changeover and the staff briefing, T-0 opens with one real test purchase - and every level lists only what is not ready, plus the fallback if it stays that way

You are the marketing planner of a community
supermarket. What sinks a promo week is rarely
the plan - it is launch day: prices not
changed, stock not arrived, posters hanging
over empty shelves. Build me a three-stage
readiness early-warning that rolls through the
days before launch and reports only what is
not ready.

I will provide:
1. The promo plan: categories, items, prices,
   depth, display locations
2. Current status: what is confirmed, what has
   not answered
3. Promo stock and inbound data
4. Store scheduling and staffing

Three check stages (each run that morning, same
output format):

T-3 (three days out) - four checks:
1. Arrival rate: promo items against purchase
   orders - low-arrival items flagged "high
   risk"; get a firm arrival date from
   purchasing for each
2. Pricing test: run test receipts on 3 promo
   items; the scan must ring the promo price -
   "it's set in the system" does not count,
  the scan counts
3. Materials: posters, display racks, shelf
   talkers, flyers - counts received at the
   store
4. Vendor reps: confirm the roster name by
   name; chase anyone unconfirmed that day

T-1 (day before) - three checks:
1. Changeover prep: old promo teardown
   materials staged, night-changeover staff
   scheduled
2. Staff briefing: did the morning meeting
   cover the promo talking points (hero items,
   limits, phrasing) - if not, it happens
   before open tomorrow
3. Digital content: blast, community group,
   and social copy final, send times locked -
   sends go out before open, not tossed off
   at noon

T-0 (30 minutes before open) - two checks:
1. First real purchase: buy one promo item for
   real; receipt price, giveaway, and points
   all correct = launch verified
2. Floor spot-check: displays photographed
   complete; 5 random promo items checked,
   shelf tag vs. system price

Output format (identical all three stages):
- Only not-ready items: item / owner (a
  specific name or role, never "the relevant
  department") / "must finish today" vs.
  "can wait until T-0" / how completion gets
  reported back
- Fixed closing section, "if still broken at
  T-0":
  · Out-of-stock items: pull the poster,
    remove the talker, put up a sold-out
    apology sign - promo signage over an
    empty shelf becomes a photographed false-
    advertising complaint
  · Items where the price never activated:
    hold the promo signage, sell at regular
    price, put the promo up once pricing is
    confirmed live
  · Displays whose vendor rep never
    confirmed: staff them from the store's
    own crew - no display left unattended

Constraints:
- Ready items never appear - the warning
  reports deviations only; good news buries
  the red items
- No sales forecasts, no "expect big crowds"
- Owners are names or roles, never
  "departments"
- If data for a stage is missing (no schedule
  yet), mark that check "no data" - do not
  quietly skip it

Why it is written this way

"The scan counts" is the standard for the whole pricing check and the highest-frequency launch failure. Between a price set in the back office and the register actually charging it sit several layers - effective times, store-level sync, coupon priority - and any layer can turn "it's definitely set" into "it rang the original price." Three test receipts, five minutes, retires the risk of the first real customer hitting the wrong price three days before it happens.

The three stages each own a window: T-3 owns what is still fixable (swap items, chase stock, chase people), T-1 owns the final approach (changeover crew, briefing, sends), T-0 owns acceptance (first purchase, tag spot-check). The logic underneath: the later an error surfaces, the fewer options remain - a stockout found at T-3 still allows swapping the item and reprinting the poster; found at T-0, only the apology sign is left. The entire value of an early warning is buying back that time gap.

The fallback block appears at the end of every stage by design, so "still broken at T-0" has an answer that does not depend on improvisation. Launch morning is a high-pressure environment - stockouts, queues, customers asking questions - and decision quality bottoms out exactly there; writing "pull the poster, put up the apology sign" in advance reduces the day to execution. Especially the poster-over-empty-shelf case: it looks trivial and photographs as textbook false advertising, and the complaint costs far more than an apology sign.

"Never list what is ready" is counterintuitive and non-negotiable: let "confirmed" items into the warning and the visibility of the exceptions dies - by day three the real page is a wall of checkmarks nobody reads for the red line. A warning reports deviations, the way a dashboard shows warning lamps, not the lamps that are off.

Variables