Setting PAR levels that stop stockouts

What PAR level actually is

Not a forecast, not a formula — the number you already carry in your head for each item, made explicit. If you know a full house burns through six gas canisters a week and you never want fewer than two in reserve, your PAR level is two. The mechanism is entirely that simple: set the floor, get told when stock crosses it.

Setting it honestly, item by item

A useful PAR level accounts for how fast the item actually moves and how long a reorder takes to arrive — not a round number picked because it looked tidy. An item with a same-day local supplier can run a lower PAR than one that takes a week to restock from off-island, even if both move at a similar pace.

Reorder point, related but distinct, is the level that should actually trigger an order — usually PAR level minus what you expect to use before the next delivery arrives. For an item delivered weekly the two numbers sit close together; for one delivered monthly, reorder point needs real headroom above PAR.

Organised the way your store actually is

PAR levels are only useful attached to items that are actually tracked as a ledger — purchases, usage, waste and counts, categorised by area (hotel, restaurant, bar) so the kitchen store and the housekeeping store are separate lists rather than one long scroll nobody scans properly. An item without a supplier and storage location attached turns a low-stock warning into a fact nobody can act on quickly.

A worked example

A 20-room property runs through roughly two gas canisters a week at typical occupancy, three during a full house. The property never wants fewer than one full canister in reserve at any point — so PAR level is set at three. The supplier delivers within two days of an order, and the property burns through about one canister every three to four days at busy times, so reorder point is set at two: order when the count hits two, and the delivery arrives well before it drops toward the PAR floor of one.

That is the whole calculation, done once. From there the item simply tells the person doing the ordering when the count crosses two, instead of that person having to remember to check.

The mistake in the other direction

PAR levels set too low cause stockouts. PAR levels set too high cause a different, quieter problem: cash tied up in stock sitting on a shelf, and for perishables, spoilage that shows up as waste a month later. A PAR level padded 'to be safe' on an item that barely moves is not actually safety — it is money sitting in a storeroom that could be doing something else.

The honest fix in both directions is the same: set the level from what the item has actually done over the last few months, not from a guess made once and never revisited.

Adjusting for season, not setting it once forever

A PAR level that was right for low season is usually wrong for a fully booked long weekend, and a level sized for peak season sits needlessly high through a quiet month. Properties with a clear seasonal pattern are usually better served by two PAR levels per item — a peak-season floor and a low-season floor — reviewed at the point the season actually turns, rather than one number stretched to cover both.

The items worth setting first

Not every item needs a PAR level on day one. Start with what has actually caused a stockout in the last few months — gas, a specific popular dish's key ingredient, the linen count for a full house — rather than trying to set a floor for every SKU in the store at once. The list that matters is usually shorter than it feels.

Whether it is working

The honest measure is simple: did the property run out of anything this month that a PAR level should have caught. A stockout on an item that already had a PAR level set means the level itself was wrong, not that the mechanism failed — and that is a five-minute fix, adjusting one number rather than rebuilding the system.

Who should actually own the number

PAR levels drift out of date fastest when nobody is specifically responsible for them — set once at opening, then never revisited as the property's occupancy pattern or menu changes. Assigning ownership of the PAR list to whoever already does the ordering, rather than leaving it as a shared responsibility nobody quite owns, is what keeps the numbers matching reality a year in rather than reflecting what the property looked like when it opened.

PAR levels and menu changes

Adding a new dish to the menu, or dropping one, changes what the kitchen actually consumes — and a PAR level set for the old menu can either run needlessly high on an ingredient a dropped dish used to need, or too low on one a new popular dish suddenly depends on. Treating a menu change as a trigger to review the affected items' PAR levels, rather than only reviewing PAR levels on some unrelated schedule, catches the gap before it becomes a stockout.

A note on non-food items

Linen, toiletries and welcome-kit consumables benefit from the same mechanism even though they never touch a kitchen. A property that runs its housekeeping stock the same disciplined way it runs its bar stock — a floor per item, tied to actual occupancy patterns rather than a monthly guess — closes the exact same gap that causes a full house to discover a shortage of fresh towels at the worst possible moment. The mechanism does not care what the item is; it only cares that someone wrote down the floor and the ledger stays current.