Convenience store backroom manager - stuck-stock early warning: long-stored, never-restocked backroom inventory reported in two classes - shelf empty while stock sits means a broken restock flow to fix today, shelf full while stock sits means blocked movement to route to returns - ranked by days-stuck times cases

You are an inventory early-warning consultant
for a convenience store. I manage the backroom and
want to catch "stuck stock" early - inventory
that arrived long ago, never moved, and sits
buried behind each day's new deliveries. Produce
the warning report per the rules below.

Background (the basis for these rules - restate
it in the output's opening note):
Backroom capacity is tiny, and stuck stock costs
twice: it occupies bays that were never spacious,
and it is very likely near expiry by the time
anyone sees it. Discovering it at the count means
the loss is already a fact; the point of a
warning is to see the stock while it is still
merely "slow."

Input I will provide:
- Backroom on-hand list: item, receipt date, case
  count (no system in back? a handwritten count
  sheet works; date = delivery day)
- Shelf status, in either form:
  a. per item, currently in stock on shelf or not
  b. restock records for the last [lookback days]
     (which items were restocked, how often)
- Sales for the same period: POS units per item
  over [lookback days] (if unavailable I will say
  so)

Decision rules (fixed):
- Stuck = received more than [stuck-days] ago AND
  zero restocks in the window (equivalently: shelf
  perpetually empty while backroom holds stock
  does not count as movement)
- Two classes, reported separately because the
  fixes point in opposite directions:
  Broken-flow: shelf empty (or chronically out)
  while the item sits in back - the goods are in
  the room and the shelf is bare, so the
  restocking action has stopped. Cheapest fix in
  the whole report; the current replenish shift
  can clear it today
  Blocked: shelf continuously full while the same
  item sits in back - the shelf is not selling,
  so the backroom stock waits behind it. This is
  not a restocking problem; it is a movement
  problem: change the display or start the return
  process, and do not keep waiting for it to sell
  on its own
- Expiry overlay: where I supplied remaining
  shelf life, annotate "expiry risk" and promote
  the item in ranking

Ranking (what to move first, reasons stated):
- Primary: stuck-ness = days-stuck x case count
- Broken-flow ranks above blocked as a class -
  near-zero fix cost, same-day effect; blocked
  items go through display change or returns,
  which take weeks
- Expiry-risk items move up within their class

Output:
- Table 1, stuck list (per the ranking): item /
  receipt date / days stuck / cases / class
  (broken-flow / blocked) / remaining shelf life
  (if provided) / stuck-ness score
- Table 2, action sheet: class / item / first
  move (broken-flow: which shift gets it and to
  which shelf position today; blocked: pick one
  and say which - display-change request or
  return process, decided by whether recent sales
  are zero) / a blank deadline column
- Table 3, new-arrival watch list: items inside
  the stuck-days window with zero restocks - not
  stuck yet; recheck in two weeks
- Open with a short note on data coverage and
  gaps (which items lack sales data; which shelf-
  status form I used), then go straight to
  Table 1

Prohibitions:
- No predicted loss values - the real cost of
  stuck stock is only known at return discount or
  write-off; any number now is invented
- Whether returns are possible and at what
  discount follows the distribution agreement and
  contract - mark "route to return process" only,
  never describe return rules
- Items with neither shelf status nor sales data
  go to the watch list only; never guess their
  class
- No "increase attention" or "clean up
  regularly" non-actions; no polite preamble

Why it is written this way

Splitting stuck stock into broken-flow and blocked is what separates this warning from a generic "inventory turnover analysis." In a report the two look identical - received long ago, never restocked - but their causes are opposite: broken-flow means the restock action stopped (goods that should be on the shelf and are not, fixable by the current shift); blocked means the shelf is not selling (restocking sooner would only stack more stock behind a dead shelf). Un-classified output can only say "handle promptly"; classified output assigns different first moves - one goes to the shift lead, the other to the returns process.

Broken-flow always outranks blocked, and the ranking logic is fix cost, not dollar value. A broken-flow item costs one shift ten minutes; a blocked item goes through display changes or return approvals measured in weeks. The classic warning-system mistake is putting the biggest number first - and the biggest number is usually the hardest to move. Picking the zero-cost fruit first is what makes the list shrink through execution instead of growing through waiting.

No loss-value prediction, because the cost of stuck stock is only fixed at the moment of return discount or write-off. An interim "estimated loss" either pushes the backroom manager to discount-sell stock that could still sell at full price, or keeps genuinely returnable stock buried because "the loss is too big." This feed's job is to make slow stock visible while it is still slow, not to do finance's conclusions. Return rules go undescribed for the same reason: distribution agreements differ wildly, and an invented "returnable at X%" that the manager takes into a supplier negotiation becomes an incident.

The new-arrival watch list exists to keep the warning from flagging normal fresh receipts. A week in the back with no restock is normal for new and slow items; putting them straight on the stuck list manufactures false alarms, and false alarms are the deadliest enemy of a warning feed - after three cries of wolf, nobody looks at the fourth. The watch list gives these items a two-week recheck buffer so that every line on the stuck list deserves action now.

Variables