Food and beverage cost control in hotels: a working guide

Food and beverage cost control in a hotel is a daily loop of records: specification, purchasing, receiving, storage, standard recipes, production, sales and a physical count. It ends in one figure, cost of sales divided by F&B revenue for the same period, and that figure must be explainable line by line.

The month-end food cost lands as a single number, and by then the delivery behind it is three weeks old, the cook who over-portioned has changed shift, and nobody can say which of the seven stages moved. The fix is not a better report. It is a loop of records that runs every day. After reading this guide you will be able to set up hotel food and beverage cost control stage by stage, calculate the monthly figure correctly, and know exactly where to look when the number moves.

What does food and beverage cost control mean in a hotel?#

It is a loop of records that runs from the purchase order to the month-end count, so that the cost percentage you read at month end can be traced back to a stage, a person and a document.

Start by separating purchases from cost of sales. Purchases are cash leaving the bank; cost of sales is what actually left the storeroom and was consumed in the period. A heavy purchase month is not a bad cost month, and treating the two as the same thing is the most common error in hotel F&B reporting.

The loop has seven stages: purchase, receive, store, issue, produce, sell, count. Each stage must leave exactly one record, and the loop is only as strong as its weakest stage. The cost percentage itself equals cost of sales divided by F&B revenue for the same dates, on the same basis on both sides. Mixing a cost figure from one period with revenue from another produces a number that cannot be acted on.

Remember that the revenue side moves the percentage as much as the cost side. A percentage can fall because the sales mix shifted toward cheaper items, not because the kitchen improved. Control therefore means comparing actual cost from the count against standard cost from the recipes; comparing only against last month tells you that something moved, never what. No external benchmark belongs in this loop. A hotel's own standard cost and its own history are the first comparison that means anything.

The loop also belongs to the whole business day, not to the kitchen alone. Night audit closes the day, the count closes the month, and the two must agree on dates, the same way one hotel business day fits together.

What has to be written down before the first purchase order?#

A specification: the written description that makes two prices comparable. Without it, a cheaper supplier is not cheaper, it is a different product.

A usable specification states what the item is, its origin or grade, cut or size, pack size, trim or fat level, temperature at delivery, shelf life, and the unit of purchase against the unit of use. A case of twelve one-liter bottles and one liter are different units, and the specification must say which one the price refers to.

Yield connects the two units. An untrimmed item at a lower kilo price can cost more per usable kilo once trim and cooking loss are counted. For example, a case of untrimmed beef at a lower price per kilo that loses a quarter of its weight in trim is not the cheaper case.

Order against a par level, from an approved supplier list, with one person ordering and a different person receiving. This separation is the cheapest control a hotel can install. Receiving is a measurement, not a signature: count, weigh, check temperature, check against the order and the specification, then sign and date. Rejections are recorded at the door, not settled later with a credit note.

Direct purchases such as fresh bread and produce go straight to the kitchen, but they still pass through receiving and are still recorded. Unrecorded direct deliveries are the most common hole in the loop. Storage needs date marking, first in first out, a locked storeroom with one keyholder, and a temperature log that someone actually reads. Issuing is a requisition signed by the department; anything that leaves the storeroom without one is invisible cost, and it will surface at the count as a shortage with no explanation. The full sequence is covered in more detail in hotel purchasing and receiving: a procedure from order to storeroom.

How do you know what a plate actually costs?#

From a standard recipe: fixed ingredients, fixed quantities, a stated number of portions and a stated portion weight, costed from current supplier prices and updated on a trigger rather than a feeling.

A recipe card carries the ingredients with unit and quantity, the yield in portions, the portion weight, the method, the cost per portion, the selling price and the date it was last costed. The date is the part most kitchens leave out. Cost per portion is the sum of each ingredient quantity multiplied by its unit cost, adjusted for trim and yield, divided by the number of portions the recipe yields.

Put the trim inside the recipe, not in the cook's head. One kilo of onions trimmed to 850 grams changes the cost per portion, and the recipe should say so. A portion is only standard if it is measured: scales at the pass, fixed-size ladles and scoops, pre-portioned items for expensive proteins. An unweighed portion is a range, and a range cannot be controlled.

Update unit costs on a trigger, for example when a supplier price moves beyond an agreed threshold or when a contract is renewed, not every morning. Daily re-costing produces noise and abandoned recipe cards. Recipe cost is the standard; the stock count gives the actual. The gap between the two is where over-portioning, waste, unrecorded complimentary items and pilferage live, and it is the only gap worth investigating. Menu decisions use cost per portion against units sold, not cost alone: the dish that sells most and costs most deserves attention before the dish nobody orders. The mechanics are laid out in standard recipes and portion cost: how hotel kitchens price a plate.

How should waste, staff meals and complimentary items be recorded?#

At the moment they happen, with a quantity, a reason and a name. Anything that leaves the kitchen without a sale still has a cost, and a cost with no record always lands in the food cost percentage with nothing to explain it.

Treat the three as separate categories, because they are controlled differently. Spoilage and waste covers expired, dropped and guest-returned items. Staff meals are a real cost with a real number of covers. Complimentary items cover welcome drinks, VIP amenities and a meal comped after a complaint.

A waste sheet records date, item, quantity, unit, reason and who recorded it. The reason codes matter more than the totals: repeated spoilage of the same item is a purchasing signal, not a kitchen discipline problem. A complimentary item should be posted to the guest's bill at zero value, or to a house account, so that the cost and the revenue it replaced meet in the same report. A comp that never appears anywhere makes the outlet look worse than it is.

Staff meals need a counted number of covers and a standard staff menu, so the cost per cover can be calculated. Left inside an "other" line, it becomes the explanation for every unexplained movement. Transfers between outlets, for example from the kitchen to the bar or to the staff canteen, are transfers, not losses: record them as transfers so the receiving outlet carries the cost. How these items reach the guest's bill and the fiscal document is a separate question, covered in fiscal devices in hotel outlets and how charges reach the bill.

How do you calculate the monthly food and beverage cost?#

Cost of sales equals opening stock plus purchases minus closing stock, adjusted for transfers and staff meals. The cost ratio equals cost of sales divided by the F&B revenue of exactly the same period.

Step one: fix the cut-off. State which deliveries and which invoices belong to the period, write the rule down, and apply the same rule every month. Step two: count everything at the same time, every storeroom, bar, kitchen and outlet, with two people, one from the operation and one from finance or cost control.

Step three: count blind. The counter should not see the expected figure, because a count that is expected to match usually does. Step four: value the count on one basis, either last purchase price or weighted average, and keep that basis. Switching basis moves the percentage without anything changing in the kitchen.

Step five: add purchases for the period, subtract the closing stock, then adjust for transfers in and out and for staff meals, so that each outlet carries what it actually consumed. Step six: divide by F&B revenue for the same dates. The result is one number for a whole month, which is why the standard cost comparison from the recipes is what turns it into action. The count is also an audit of the loop: a shortage that repeats in the same few high-value items points at receiving or issuing, not at the kitchen.

These figures are invented to show the arithmetic; they are not a target and not an industry average.

À la carte, buffet and all-inclusive: how does the loop change?#

The loop stays the same in all three. What changes is where the portion is fixed and where the revenue is measured, and those two changes decide which comparison is meaningful.

À la carte fixes the portion in the recipe and the revenue per item, so cost per portion against selling price is a direct comparison. The risks are over-portioning at the pass and complimentary items that never reach a bill.

A buffet fixes the portion at the guest, not in the recipe. Control moves to production planning from reservations and history, batch cooking, and cost per cover, calculated as total buffet cost divided by covers actually served. Covers must be counted, not estimated from revenue.

In all-inclusive, board is inside the room rate, so a food cost percentage measured against outlet revenue compares a cost with a revenue figure that does not contain it. The meaningful measure is cost per guest-night, and separating room and board revenue is a revenue-side exercise, not a kitchen one, as explained in separating room and board revenue in all-inclusive hotels. Banquets and groups are costed per event against the contract and the function sheet, which is a different comparison from a retail outlet serving walk-in guests. Room service and minibar carry small volumes and high waste, so par levels, pack sizes and expiry dates matter more there than anywhere else in the building.

One rule holds across every concept: same period, same costing basis, same definition of what counts as revenue. Without that, comparing two outlets or two months is comparing two different calculations.

When the cost percentage moves, where do you look first?#

Work backwards along the loop from the count, and check the records before you check the people. Most movements are explained by the calculation itself, not by the kitchen.

First, test the number: were all count sheets returned, was the same costing basis used, was the cut-off the same, and does the revenue cover exactly the same dates. Second, test the mix: did sales shift toward lower-margin items, or did a group, a package or a banquet change the revenue base under the percentage. Third, test the price: did supplier prices move while the recipe cards stayed on last quarter's costs.

Fourth, test the portion and the yield: weigh ten plates at the pass and check trim on the highest-cost proteins. Fifth, test the outflow: complimentary items, staff meals and transfers without a slip, all of which consume cost and produce no revenue. Sixth, test the stock: a shortage that repeats in the same few items is a receiving or issuing problem, and it will keep repeating until the door and the storeroom are fixed.

Only then look at people and process. A conversation about discipline that starts before the records are clean teaches the team that the numbers are unreliable, and the records get worse.

What to do next#

Pick one stage of the loop and close it this week: write the specification for your five highest-spend items, or start the waste sheet tomorrow morning. The loop only produces a number worth reading once every stage leaves a record, and the first record is always cheaper to install than the report it eventually explains.