What should a chef consider when setting par levels for prepped items made in house daily?

Prep pars answer a different question than purchase pars

A purchase par asks how much of an ingredient should be on the shelf when the truck arrives, and it is governed by delivery frequency and vendor lead time. A prep par asks how much of a made item should exist at the start of service, and it is governed by labor hours, batch yield and shelf life. Using one kind of number for both is where kitchens get into trouble. Set a demi glace par the way you set a par for a case of onions and you will either dump quarts on Tuesday or scramble at six on Friday.

The second difference is who fixes it when the number is wrong. A short purchase par gets solved by a vendor, sometimes overnight and sometimes not at all. A short prep par gets solved by pulling a cook off the line in the middle of service, which costs ticket times and usually costs quality too. Overshooting has an equally direct price: prep labor, plus product, plus whatever gets scraped into the bin at close. That combination is why prep deserves its own review cycle instead of riding along with the order guide. Related: What is the best way to organize vendor order sheets across several suppliers?

Keep reading: How do you set par levels for restaurant inventory that actually hold up?, What is the fastest way to run accurate weekly inventory counts in a kitchen?, How do reorder alerts prevent stockouts before they hit the line?. See how ParLevelr helps you restaurant inventory par levels and reorder alerts.

Work backward from the menu, not from the walk-in

Start with the menu items each component feeds. A pan sauce base that goes into three entrees behaves nothing like a garnish used on one appetizer. List the components, then list the menu items that consume them and the portion each one takes. Yield matters here more than raw weight. If a batch recipe finishes at a certain number of portions after cooking loss and straining, that finished number is the unit the par should be written in, because that is the unit the line actually spends.

Then attach a forecast. Most kitchens already have a usable one sitting in the POS menu mix report: item counts by day of week over the last several weeks. Multiply the typical count for a given day by portion size, sum across every menu item that uses the component, and you have a demand figure instead of a hunch. Do this for the ten or fifteen components that consume the most prep labor first. The long tail can stay on feel until the big ones are stable and someone has time. Related: What is the fastest way to run accurate weekly inventory counts in a kitchen?

Cap every prep par at its shelf life, not at a week of demand

A purchased case can sit. A quart of blanched greens or a portioned fish fillet cannot. The practical rule is that a prep par should never exceed what the item still tastes right after, which for many refrigerated preps is a small number of days and for delicate items is less than one service. Write the hold window next to the par on the prep sheet so whoever sets the number sees the constraint in the same glance. When the day's demand runs past that window, the answer is two smaller batches, not one larger one.

Batch size is the constraint people forget second. If your stock pot, tilt skillet or vacuum bag size forces a fixed batch, the real choices are one batch, two batches and so on, not any number in between. Express the par as a count of batches and note the trigger point that makes the next batch worth firing. Kitchens that skip this step end up making half batches that tie up the same equipment and roughly the same labor as a full one, which is how a prep list quietly eats a morning.

Turn the prep par into a list that fills itself in

The prep sheet needs four columns: item, par, on hand and make. Someone walks the line and the reach-ins once, writes down what is there and subtracts. That single subtraction removes the guesswork that leads a cook to make a full batch of something with three quarts already sitting behind the door. If your inventory tool stores the pars, the same counting pass can generate the make quantities directly, which has the useful side effect of leaving a record of what was actually produced each day. Related: How do reorder alerts prevent stockouts before they hit the line?

Review the sheet weekly against two other records: the waste log and the 86 log. Items that keep appearing in waste have a par above real demand or above their hold window. Items that get 86'd late in service have a par below demand or a batch trigger set too late. Adjust in one direction at a time, in modest steps, and give each change two or three weeks before judging it. Prep demand moves with weather, events and menu changes, so a par that was right in February is not automatically right in June. Related: How does tracking usage help you right-size par levels over time?