Par levels: how much stock to hold, and why

Par levels answer a deceptively simple question: how much of each ingredient should be on the shelf right now? Too little, and the kitchen runs out in the middle of a Saturday. Too much, and cash sits on shelves, stock spoils before it is used, and the walk-in becomes so crowded that rotation breaks down. Most kitchens solve the problem by instinct, and most of them err on the side of too much, because running out feels worse than wasting.

A par level replaces that instinct with arithmetic. It is the quantity of an item you need on hand to get through to your next delivery, plus a small buffer for the unexpected. When stock is counted, whoever orders simply tops each line up to its par. The decision about how much to order stops being a judgement made in a hurry and becomes a number that anyone trained can follow.

This guide explains how to set par levels from real usage rather than habit, how to build them around your delivery days, how to adjust them for busy and quiet periods, and the two kinds of ingredient where a par never quite works and something else is needed instead.

A par level is a target quantity for a single item, held in the stores at the point just after an order arrives. For a case of tinned tomatoes, it might be eight tins. For fresh cream, it might be twelve litres. For a premium cut of beef, it might be a precise number of portions. Each par is specific to one item, one venue and one stretch of the week.

The par is not the amount you order. The amount you order is the par minus what is already on the shelf. If the par for double cream is twelve litres and four are left when you count, you order eight. That simple subtraction is the whole system, and it only works if the par itself is set well and the count is honest.

Pars sit inside a wider cycle of ordering, receiving and checking stock. Our guide to restaurant supplier management covers that cycle from start to finish, and the par is the piece of it that decides quantities.

Without par levels, ordering drifts. It drifts upward in the weeks after somebody ran out, because nobody wants that to happen again, and it rarely drifts back down. The result is over ordering that nobody decided on, which shows up in three places.

The first is spoilage. Stock that is bought before it is needed has more time to go off. Fresh produce, dairy, bread and proteins all carry a limited life, and every extra day on the shelf is a day closer to the bin. Our guide on how to reduce food waste in a restaurant kitchen shows how often waste traces back to buying too much too early.

The second is cash. Stock on shelves is money that has left the bank and not yet come back as sales. For a small business, a walk-in holding several days more stock than it needs is cash that could be paying a supplier, a wage or a bill.

The third is control. A crowded store is a store where stock gets pushed to the back and forgotten, where rotation fails, and where counts take longer and are less accurate. Lower, steadier stock levels make every other kitchen habit easier.

Here is a reliable method for setting a par from real data. It takes an afternoon for your main lines and gets quicker once you have done it once.

Notice that the par in the example applies to the Friday delivery. The Tuesday delivery only needs to cover Tuesday, Wednesday and Thursday, three days, so its par is lower: three litres times three days plus a buffer, or about twelve litres. Many kitchens use a single par for each item and accept the slight over-stocking on shorter gaps. Better-run kitchens use a par for each delivery day, because the difference adds up across dozens of lines.

Daily usage works well when trade is steady. When it is not, usage per cover is more accurate. Instead of asking how much cream the kitchen uses per day, ask how much it uses per guest served.

Take the same four weeks. If the kitchen used 84 litres of cream while serving 2,800 covers, usage per cover is 0.03 litres, or 30 millilitres. Now the par can follow the bookings. If the next four days are forecast at 450 covers, the kitchen needs about 13.5 litres to cover them, plus a buffer.

This approach shines in venues with uneven weeks: a restaurant that does three times the covers on Saturday that it does on Tuesday, a hotel restaurant whose trade follows occupancy, a venue near a stadium or a theatre. It also makes pars self-adjusting for busy and quiet periods, because the covers forecast does the work.

The forecast itself can come from your bookings, plus an allowance for walk-ins based on the same days in recent weeks. A restaurant that takes most of its trade by reservation already has much of the forecast written down.

Here is an illustrative par calculation for a restaurant that receives produce and dairy on Tuesdays and Fridays. The figures are examples to show the method, not suggested quantities.

The third line is the important one. A par is a quantity, but it has to respect shelf life. When the days to cover are longer than the life of the product, no par will work without causing spoilage, and the answer is a different delivery pattern rather than a bigger number.

Pars and the ordering schedule are two halves of the same thing. The schedule decides how many days each delivery must cover; the par decides how much stock that requires. Change one and you must change the other.

Start by writing your ordering schedule down. For each supplier, note the delivery days and the order cut-off times. Then, for each delivery, note how many trading days it must cover until the next one. That number is what you multiply daily usage by.

If the schedule creates long gaps for perishable lines, consider whether a second, smaller delivery is worth it. An extra delivery has a cost, sometimes a minimum order or a delivery charge, but it may save more than it costs in spoilage on short-life products.

The buffer is what protects you from running out when usage is higher than average or a delivery is late. Too small, and you run out on busy nights. Too large, and you have rebuilt the over-stocking the par was meant to fix.

A sensible starting point is one day's usage for most lines. Raise it for items where running out would be serious and the supplier is unreliable. Lower it for items with short shelf lives, where a large buffer becomes waste. For expensive items with predictable demand, keep the buffer tight, because every extra unit is costly cash on a shelf.

Review buffers when you review supplier performance. A supplier who delivers late or short-ships often justifies a bigger buffer on their lines; a reliable one lets you run tighter.

Pars are not set once and left. They need reviewing whenever the things they are built on change:

A practical rhythm is to review pars for your top twenty lines every month, and the rest every quarter, plus an immediate review whenever the menu changes. If you use usage per cover, much of the seasonal adjustment happens automatically, because the forecast changes even when the per-cover figure does not.

Par levels suit ingredients with steady use and a reasonable shelf life. Two kinds of ingredient do not fit.

Fresh fish, artisan bread, some soft herbs and certain fresh pastries are best bought daily, or nearly daily, in exactly the quantity needed. A par that covers several days will either spoil or force the kitchen to sell product past its best. For these, order from tomorrow's forecast rather than to a standing par, and accept that ordering is a daily task.

A tin of truffle paste, a premium spirit used in one cocktail, an ingredient for a special that runs occasionally: these are used too rarely for an average to mean anything. A par based on their average usage would either sit unused for weeks or run out at the wrong moment. For these, order to specific needs, such as a planned special or a booked event, and keep a minimum of one unit only if it genuinely needs to be available.

The bar works on the same principle, with a twist: many bar lines have long shelf lives, so the risk is tied-up cash rather than spoilage. Bar pars can be set from usage per day or per cover in exactly the same way, with the buffer kept modest on expensive spirits and wines. Count the back bar and the cellar separately, because the bar itself holds a working stock that is topped up from the cellar, not from the supplier. Our guide to bar inventory management covers bar counts in detail.

A new venue has no four weeks of usage to work from, so the first pars have to be built from the menu instead. Take each dish, estimate how many will sell per cover from similar venues or from the opening forecast, multiply by the quantity of each ingredient in the recipe, and add the results across the menu. That gives an estimated usage per cover for every ingredient, which can be multiplied by the forecast covers in the same way as an established kitchen's data.

These first pars will be wrong, and that is expected. Keep the buffers slightly larger in the opening weeks, count weekly rather than monthly, and replace the estimates with real usage as soon as two or three weeks of trading have been recorded. By the end of the first month, most lines should be running on measured usage rather than guesses.

Par levels are one piece of a wider practice of keeping stock and records in step, often called inventory control. The other pieces are the count, the receiving check and the record of what moved between them. Pars decide how much to hold; the count tells you how much you actually have; the comparison between them, over time, tells you whether your ordering is working. A kitchen that is regularly well below par at the end of a delivery cycle is under-stocking, and one that is regularly well above it is over ordering. Both are easy to see once the numbers sit side by side, week after week.

A par is only as good as the count it is compared with. If the stock count is wrong, the order is wrong, however carefully the par was set. Count in shelf order, in consistent units, at the same point in the week, so the number on the sheet matches what is really on the shelf. Our guide on how to run a restaurant stock count properly sets out the routine.

The last point is worth stressing. A par is the most you should hold after a delivery, not a number the stores must always show. Stock below par at the end of a service is normal and expected; that is what the next order is for.

Pars become a habit when the order guide carries them. Put the par beside each line in the guide, along with the pack size and the supplier. Whoever orders counts the line, subtracts the count from the par, rounds to the pack and writes the order. Nobody has to remember anything, and two different people ordering on two different days will order the same amount. The guide is also where pars get reviewed: once a month, check the top lines against recent usage and adjust.

Par levels depend on knowing how busy the coming days will be. In servearound, the reservations diary shows the covers booked for each service, by party size, so the forecast behind usage-per-cover pars starts from real bookings rather than a guess. The Analytics module shows covers by day and hour, set against the period before, which helps you choose the right weeks to base usage on and spot the seasonal shifts that should move your pars.

The menu report shows how many of each dish actually sold, which tells you which ingredients the menu is really leaning on. If you would like to see how that reads for your venue, get in touch.

Every item you buy regularly. Start with the twenty or thirty lines that account for most of your spend, then extend to the rest. Occasional and specialist items are better ordered to need.

Only with a reason, such as a booked event or a forecast spike, and ideally with a note on the order. A par that is regularly exceeded is a par that needs changing, not ignoring.

Look at whether usage has risen, whether the buffer is too small for an unreliable supplier, or whether the count was wrong. Running out occasionally is acceptable; running out regularly means the par is out of date.

Partly. Staples used across many dishes still suit pars. Ingredients for the changing dishes are better ordered to each week's menu and forecast.

Directly. Lower, steadier stock means less spoilage and fewer emergency purchases at higher prices, both of which show up in food cost. Our guide to calculating food cost percentage explains how to measure the effect.

Treat it as a signal. Either usage has fallen and the par should come down, or the item is no longer earning its place. Use the surplus deliberately, through a special or the staff meal, before it reaches its date, and lower the par so the same thing does not happen again next month.

Rarely. Each venue's trade, menu mix and delivery days differ. Set pars per venue, even when they share suppliers and an order guide.

Par levels are the quantity of each item you need to get through to the next delivery, plus a small buffer. Set them from measured usage, either per day or per cover, multiply by the days each delivery must cover, add a buffer, and round to the pack size. Build them around your ordering schedule, respect shelf life, and review them when the menu, the trade or the suppliers change.

Use them for steady, regular lines, and order to need for very short-life items and for rare, expensive ones. Done this way, par levels quietly remove most over ordering, free up cash and make the whole kitchen easier to run.