The weekly food cost report worth ten minutes

Most kitchens either have no food cost report at all or have one nobody reads. The first kind finds out about food cost when the accountant sends the monthly figures, weeks after anything could have been done. The second kind produces a spreadsheet with forty rows every week, and the head chef glances at the percentage in the corner and moves on.

A useful weekly food cost report sits between the two. It fits on one page. It is ready by Monday morning. It takes ten minutes to read and discuss, and it answers two questions: what was last week's food cost, and can we believe it? The second question matters as much as the first, because a figure built on a missed shelf or a misplaced delivery will send the kitchen chasing a problem that does not exist.

This guide sets out what goes on the page, in what order, with the four lines that tell you whether the week was real. It then shows how to build the report, read it and turn it into one action for the week ahead. Every figure in it is an example.

Monthly food cost is what the accounts need, but it arrives too late to manage a kitchen with. By the time a month's figure is known, the habits that produced it are four weeks old, the people involved have forgotten the details, and the next month is already a week in. A weekly figure shows a change within days, while it can still be explained.

One page matters because of who reads it. A head chef has a kitchen to run, and a report that takes half an hour to understand will not be read every week. One page forces choices: only the numbers that lead to a decision belong on it. Everything else can live in the spreadsheet behind it, for the weeks when someone needs to dig.

Decide three things before the first report goes out.

The meeting matters more than the document. The page is there to make a ten-minute conversation possible.

The order below puts the headline first, then the evidence that the headline is real, then the detail behind it, then the action.

The week, its trading days and anything unusual about it: a bank holiday, a private party, a menu change, a closed day. Food sales for the week, net of tax and service charge, and the number of covers. These give the reader the context for everything below.

The week's food cost percentage, with the previous four weeks beside it so the trend is visible at a glance. If you calculate a theoretical food cost from your recipes, put it next to the actual figure. Our guide to calculating food cost percentage covers the calculation itself.

These come straight after the headline, because they decide how much weight it can carry.

If all four lines are clean, the headline can be believed. If one of them is not, mark the week's figure as provisional, and say why. A provisional figure is still useful, but it should not drive changes in the kitchen until the problem behind it is fixed.

The five lines with the largest variance by value, or the five ingredients whose usage changed most compared with sales. Five is enough. This is where the conversation about the kitchen starts, and our guide to reading food cost variance explains the order in which to look at them.

Two short figures: the week's recorded waste by value, against a usual week, and any significant price changes on invoices this week. Waste is the kitchen's to manage; prices are the buyer's. Keeping them on separate lines keeps the conversation fair.

One action for the coming week, with an owner. Not five; one. Last week's action, and whether it was done, goes on the line above it.

An invented week, as the page might read:

Read in order, the page tells a clear story. The number is up 1.2 points, but the theoretical figure rose by 0.4 because of the party's menu, the salmon credit accounts for about another 0.35, and the extra comps for about 0.5. What remains is small. The week was real, mostly explained, and the action is aimed at the part that is not.

The report draws on four sources, and most venues already have all of them:

A spreadsheet with one tab per source and one summary tab is enough for most venues. Our guide to building a food cost calculator in a spreadsheet shows a layout that works.

Be realistic about the time. Reading the report takes ten minutes. Preparing it takes longer, perhaps 30 to 45 minutes once the count and the invoices are in, as an example for a single venue. Most of that is matching deliveries to invoices. The time falls as the routine settles and as the same person does it each week.

The order in which the report is prepared matters as much as the order in which it is read. Done in the wrong order, the preparer spends time on a figure that turns out to be wrong.

Working this way, a problem in the count or the purchases is found before anybody has seen a headline figure that might need retracting.

The meeting is short, so it needs a shape. A useful pattern is to ask the same four questions every week, in the same order.

Keep the questions the same even when the figure is good. A good week explained is as useful as a bad week explained, because it shows what to keep doing.

A group with several venues can use the same page for each venue and add one summary page on top: each venue's headline, its four lines as ticks or crosses, and its action. The summary shows at a glance which venues' figures can be trusted this week and which cannot, which is more useful than a league table of percentages.

Resist comparing venues' food costs directly unless their menus are the same. Different menus run different figures for sound reasons, and a league table pushes managers towards the cheapest dishes rather than the best results. Compare each venue with itself, week by week, and with its own theoretical figure.

One week is noise; several weeks are a trend. Food cost is a performance indicator, and like any indicator its direction over several readings says more than any one reading. The four previous weeks on the headline line are there so nobody reacts to a single number.

Two simple habits help. Look at a rolling four-week figure as well as the week itself, because it smooths the effect of deliveries landing unevenly. And, once you have a year of reports, compare with the same week last year, because seasonal menus and seasonal prices make some weeks predictably higher than others.

Watch for direction as well as level. Four weeks each a little higher than the last say more than one high week followed by a normal one.

A one page report is defined as much by what it leaves out as by what it contains.

If the page starts to grow, ask of each new line what decision it leads to. Lines that lead to no decision come off.

The weekly reports for a month should add up to something close to the month's food cost in the accounts. They will rarely match exactly, because the accountant's month and your weeks do not line up, and because invoices that arrive late are corrected in the accounts but not in last week's report.

Check the two against each other each month. A small difference is normal. A large one points at the count or the cut-off, and is worth finding before it becomes a habit. Our guide to actual vs theoretical food cost and our guide to running a stock count properly cover the two most common sources of the gap.

The head chef reads the report; the brigade should hear about it. A short mention at the start of the week's first briefing, saying what last week's number was, whether it was real and what the one action is, does more for food cost tracking than any amount of reporting that stays in the office.

Share the good weeks too. A team that only hears about food cost when it rises learns to fear the report rather than to use it. When a portion check worked or the waste fell, say so.

A week with a low food cost deserves the same scrutiny as a week with a high one. A figure that drops suddenly is just as likely to be a counting error as a triumph: a shelf counted twice, a delivery booked to the following week, a credit entered twice. Run the four lines before celebrating.

When the four lines are clean and the number really is better, find out why and write it down. A portion check that worked, a change of supplier, a waste habit that stuck: each is something to keep doing, and the report is where the kitchen's good decisions are remembered as well as its problems.

Food cost is one of a small number of reports a venue needs. Our guide to restaurant reporting with six reports and no more sets out the others, and our guide to a weekly review that takes thirty minutes shows how to read them together in one sitting. Keeping the formats consistent, one page each with the headline first, makes the whole set quicker to read.

A food cost report needs counts, purchases, sales and waste. servearound covers the sales side, and the counts, invoices and waste log stay in your own sheet today.

The menu report in Analytics gives units sold, revenue and share of sales for every dish and category for any week, which is the sales column and the mix line of the report. The payments report gives discounts, voids and comps by member of staff, which is the sales adjustments line. Covers for each day are in the reservations reports, booked and walk-in, and bills for a date range can be downloaded as a spreadsheet to pull straight into the report's sales tab.

servearound does not track stock, purchases or waste today, so those lines come from your own records. Inventory tracking, for the stock behind the menu, is coming. If you would like to see the sales side for yourself, get in touch.

The week's context, the food cost percentage with the previous four weeks, four lines that check whether the figure is real (the count, purchases, sales adjustments and mix), the top five movers, waste and price changes, and one action with an owner.

Reading it should take about ten minutes. Preparing it might take 30 to 45 minutes for a single venue once the count and invoices are in, and less as the routine settles.

At least the head chef and the general manager, and the owner if they are involved weekly. The brigade should hear the headline and the action at a briefing.

A variance report lists every item with its expected and actual usage. A food cost report summarises the week on one page and shows only the top movers from the variance report.

Yes: the same definition of food sales, the same valuation of stock and the same treatment of comps and staff meals. Otherwise the weeks cannot be reconciled with the month.

Mark the week's figure as provisional, say what was missing, and do not act on it in the kitchen. Fix the count next week and look at the two weeks together.

It is not essential, but it helps a great deal, because it shows how much of a change the menu mix explains. Without it, write a short note about the mix on the fourth line.

A weekly food cost report worth reading fits on one page and is discussed for ten minutes at a fixed time. Put the context and the headline first, with four weeks of trend, then the four lines that say whether the week was real: a complete count, purchases matched to invoices, sales adjustments, and the mix. If any of them fails, mark the figure as provisional.

Below that, show the top five movers, waste and price changes, and one action with an owner, with last week's action checked off. Reconcile the weeks with the month, share the headline with the team, and keep the page short enough that it gets read every week, not only the bad ones.