Guide
Costing & Menu Engineering: Pricing a Menu That Actually Makes Money
By David Holden · Stock Auditing since 1992 – Allied Domecq, Hilton, Marco Pierre White, RedCat Hospitality Ltd.
A lot of menu margin is lost before the first plate goes out. VAT left in the price, meat costed at raw weight, mark-up mistaken for margin – each one makes a GP target look achievable when it never was.
A 3% creep in food cost can wipe out the entire net profit of a venue. This guide covers how to cost a dish properly, how to price it, how to judge what each dish is really earning, and how to keep the numbers true once the menu is live. It brings together everything I’ve written on it.
PART ONE - TAKE OUT THE VAT FIRST
Should GP be worked out with or without VAT?
Without. Take VAT off the selling price before you calculate GP. Leave it in and every margin target that follows is overstated.
A chef sent me his dish costs for a menu engineering report. One steak showed 75.7% GP. He’d left VAT in the £24.00 selling price. Take it out and the revenue is £20.00, the cost is still £5.83, and the GP is 70.9%. Nearly five points lower and nothing had changed in the kitchen.
You’re a tax collector, not the owner of that VAT. The moment money passes through the till, part of it belongs to HMRC. Any GP worked out with VAT inside the price is measured against revenue that was never yours.
PART TWO - WHAT'S ON THE PLATE
Why does cooked weight change the cost of a dish?
Because you pay for the raw weight but serve the cooked weight. The cost of the whole joint has to be recovered from what’s left after cooking.
A 9kg topside joint at £12.00/kg costs £108. Roasting loses around 30% of the weight, so you end up with 6.3kg.
The £108 value hasn’t shrunk, so the true cost is £17.14 per cooked kg.
A 200g portion: £2.40 costed at raw weight, £3.43 at cooked weight. That’s £1.03 a plate. On a roast selling at £18.50, a dish you think is making 70% GP is running closer to 63%, before the vegetables or the gravy.
The same applies to anything that loses weight in prep or cooking: trim on veg, bone and fat on meat, shrink on fish. Dish specs costed at a desk from invoice weights and never checked in the kitchen are a recipe for disappointment.
PART THREE - SETTING THE PRICE
What do you multiply your cost price by to hit your GP?
For 72% GP, multiply the cost ex VAT by 4.29.
That gives you the selling price with VAT included. For any target: 120 ÷ (100 − target GP%).
Two rules: use the cost price ex VAT, and remember the 120 is 100 plus 20% VAT, if the rate ever changes, so does the 120.
Mark-up isn’t margin. Add 72% on top of a £1.29 cost and you get £2.22 ex VAT - A 42% GP. Price that way and you’ve given away 30 points before the first one is sold.
PART FOUR - PERCENTAGE OR CASH?
Is a higher GP percentage always better?
No. You can’t bank a percentage.
A lower GP on a higher-priced item can put more cash in the till than a high GP on a cheap one.
Champagne typically runs at 30–50% GP, well below most of the drinks list. On percentage alone it looks weak. Yet the cash profit per bottle is often two or three times that of a standard bottle of wine, and because volume is low, the lower margin percentage barely moves the overall blend.
Look at both measures together: GP % shows how efficiently an item converts its price into profit; cash profit (price ex VAT minus cost) shows what it actually contributes. What matters in the end is the blend across everything you sell.
PART FIVE - MENU ENGINEERING
How do you know which dishes to keep, change or cut?
Plot every dish on two measures. How much it sells and how much it makes. Each one falls into one of four groups, and each group needs a different response.
“Makes” means cash, not percentage. Contribution = selling price ex VAT − food cost. A dish makes “high” if its contribution is above the average across everything sold: total contribution ÷ total dishes sold.
For example, a burger at £15.00 costing £4.50 makes £8.00 at 64% GP. A steak at £28.00 costing £11.20 makes £12.13 at 52% GP. The steak has the lower percentage, but it puts more cash in the till.
“Sells” uses the 70% rule. A dish counts as popular if it takes at least 70% of its fair share of sales. With 20 dishes on the menu, each one’s fair share is 5%, so the line is 3.5% of dishes sold.
HOW TO RUN IT
-
Pull 8–12 weeks of EPOS sales by dish.
-
Cost every dish properly – cooked weights, current supplier prices.
-
Work out each dish’s contribution and share of sales.
-
Sort every dish into the four groups.
-
Review menu layout, portions, supplier prices and add-ons – sides, desserts, drinks.
The hard calls are the dishes on the borderline. That’s where operational reality is the tiebreaker: If two dishes earn about the same, but one needs a specialist ingredient nothing else uses (ordered weekly, partly binned weekly) that one is costing you more than the report shows.
Smaller menus usually make more money: Less stock, less prep, less waste, faster decisions. Ask of every dish: If it disappeared tomorrow, would anyone notice?
The biggest gains often come not from cutting dishes, but from turning a few Puzzles into Stars - A better position on the menu, and staff who recommend them.
Put it next to the stock results. If a site has poor food GP and high waste, the Dogs and Plough-horses usually show where the problems are hiding. The menu tells you where to investigate next.
PART SIX - KEEPING COST PRICES UP TO DATE
Why does a properly costed menu still lose money?
Because prices go up, suppliers change and portions get bigger - All whilst the costings on file remain the same. The menu says one thing, the kitchen does another.
Supplier price rises
Your supplier puts 18p on a burger. Sell 1,500 a month and do nothing, and that’s £3,240 a year gone. Work out the new price before you change anything: on that burger, a 75p rise is worth £11,250 a year.
Read: Your Supplier Just Put 18p on Your Burger. Here's The £11,250 Decision →
Buying everything from one supplier
It’s easy, but no supplier is cheapest at everything. Trying a local butcher for one steak saved 38p a portion. At 350 a week, that’s £6,916 a year.
Portions creeping up
Chips 30g over on a 200g portion. At 1,500 portions a week, that’s £4,212 a year – on chips alone. Nobody decides to over-portion. It just creeps in.
Read: A 15% Variance On Chips Is Quietly Costing You £4,212 a Year →
Check the numbers before you change the menu. At Hilton and Marco Pierre White, every new dish, price change and cost rise went through rigorous costing and modelling first, before a single menu was printed.
PART SEVEN - THEORETICAL vs ACTUAL GP
What’s the difference between theoretical and actual GP?
Theoretical GP is what your costings say you should make on what you sold. Actual GP is what the stocktake says you actually made. The gap is where your missing profit is going.
A gap of 1–2% may be acceptable. Anything more needs investigating and the investigation has to find the operational cause, not just the number. Theoretical GP is only as accurate as the costings behind it: If the recipes are out of date, you have nothing reliable to measure against.
Watch the sales mix too. If GP drops but the mix has shifted toward lower-margin lines, that may explain it. If the mix hasn’t changed and GP has still dropped, the answer is obviously in the operation.
What’s it worth to you?
VAT left in the price, raw-weight costing, an absorbed supplier rise and 15% over on chips – on the examples in this guide, that’s thousands of pounds a year, and none of it shows up as waste or theft.
PART EIGHT
Common questions
What is a good food GP for a restaurant?
Most restaurants target 60–70%. Fine dining may reach 70% or more; high-volume casual dining often sits nearer 60–65%.
How do I calculate GP?
Opening stock + purchases − closing stock = cost of sales. Net sales (ex VAT) − cost of sales = GP in pounds. Divide by net sales × 100 for GP %.
What do I multiply my cost by for 70% GP?
4.00. That gives you the selling price including 20% VAT. Use the cost ex VAT.
How often should dishes be re-costed?
Whenever a supplier price changes on a main ingredient, and as a full review at least every quarter alongside your sales figures.
RELATED GUIDES
Understanding Your Stocktake Report
What the numbers mean, and when a variance is a problem.