Inventory is the least popular topic in this course and the one that changes the most. It is where the money you already spent is sitting, it is the only way to know what your food cost actually was, and it is the place where cost control stops being a set of good intentions and becomes a number. It also has a bad reputation it half deserves: done the way most books describe it, it takes all night. Done the way that works, it takes forty minutes a month.
🧮 What you bought is not what you used
Start with the single most useful piece of arithmetic in the topic. Your food cost for a month is not what you paid suppliers — it is what actually left the shelves:
Opening stock + purchases − closing stock = what you used.
In this restaurant last month: $9,000 of stock at the start, $15,500 of purchases, $8,500 left at the end, which is $16,000 used. If you had judged the month by invoices you would have said $15,500, understated by $500, and concluded that food cost was 31% when it was 32%. In a month where you stock up for a busy weekend the error goes the other way and is much larger, which is exactly the trap the financial reporting course warned about: a statement built on purchases makes a stocking-up month look like a disaster and the next one look brilliant, when nothing whatsoever has changed.
Two counts a year are enough to notice something is wrong. Twelve are enough to manage. This is the reason the count exists at all — not for the accountant, but because without it you do not know your own food cost, and every percentage in the budget course is a guess.
📦 Par levels: how much is enough
A par level is the amount of an item you want to have on hand, set once so that ordering stops being a judgement call at seven in the morning. The arithmetic is: how much you use in a day, times the days until the next delivery, plus a little for a bad estimate.
Buns, in this kitchen: 500 burgers a month is about 17 a day, deliveries come twice a week, so four days of cover is 68 — call it 72, which is six packs of twelve. Below 72 on delivery day, order. That is the whole method, and its value is that the decision was made once, calmly, by somebody who did the division, rather than every week by whoever is standing there.
Set pars for your main items only, review them when the menu or the volume changes, and resist the two failure modes. Too high and you are financing your supplier's warehouse in your own walk-in, with spoilage on top. Too low and you run out on a Saturday, which costs far more than the stock would have — a dish unavailable at peak is a sale lost and sometimes a customer lost. The point of a par level is not to be lean. It is to stop the number being invented under pressure.
🔄 Rotation, and money asleep in the walk-in
Divide what you used by what you hold and you get how many times your stock turns over in a month: $16,000 divided by $9,000 is 1.8 turns, which is another way of saying this restaurant is carrying about seventeen days of food. For a kitchen that gets deliveries twice a week, that is a lot — a week of cover is a reasonable target for fresh-focused restaurants, and it would put the same shelves at around $4,000.
The $5,000 difference is not an accounting nicety. It is money in the bank instead of money in a fridge, and it is money that stops expiring: every peso of stock beyond what you need carries a quiet risk of becoming waste, which is the next topic. High stock also hides problems, because a well-stocked shelf never forces anybody to notice that a product is not selling.
Getting it down is mostly two habits. First in, first out — new deliveries go behind, not in front, everything is dated on arrival, and the oldest is always the next one opened. It sounds trivial and it is the difference between a walk-in that loses nothing and one that throws out a tray a week. And second, order to the par rather than to the truck: buying an extra case because it was on offer is a decision to convert cash into stock, which is sometimes right and is never free.
📋 A count that actually happens
The reason most restaurants do not count is that they try to do it perfectly. Here is the version that survives contact with a real month.
- Same day every month, at the end, after the last delivery and before service. On the calendar, like payroll.
- Sheet in shelf order. The list follows the physical layout — walk-in, freezer, dry store, bar — so nobody walks in circles or skips a corner.
- Two people: one counts, one writes. It is twice as fast and half as wrong.
- Count what matters. The expensive and fast-moving items get counted properly; the vinegar gets estimated. Precision on cheap slow items buys nothing.
- Weekly for the top few. Meat, fish, alcohol — the items that are valuable, portionable and easy to lose. Four lines, five minutes, and they catch a problem within a week instead of within a quarter.
- Same method every time. An imperfect count repeated identically shows you the trend, which is what you are actually after. A perfect count done once tells you nothing.
What you do with it is a single comparison: what the count says you used, against what your recipes say you should have used for the dishes you sold. When those two numbers separate, the difference is trim, over-portioning, waste or something walking out of the building — and the following two topics are about which. That comparison is also what makes the budget course's variance sheet trustworthy, because until you count, the food cost line on your statement is an estimate wearing a suit.