Every menu price has two jobs. It has to cover what the plate really costs and leave enough to pay rent, staff and the owner. And it has to feel right to the guest reading it, next to the other items on the page, the room they are sitting in and the place down the street. A price that does only the first job loses guests. A price that does only the second loses money on every plate, often on the most popular ones.
Most menus are priced once, at opening, by a mix of cost-plus and a look at competitors, then raised across the board whenever ingredient prices jump. This article is a better routine: cost each dish properly, look at the menu as a whole with a well-tested method, check how guests read the prices, and change a few things on purpose instead of everything at once.
In short: Cost sets the floor, the guest's sense of value sets where you can sit above it, and the menu layout steers what gets ordered. Cost every plate fully, including waste and packaging. Judge each dish by the money it leaves (price minus plate cost), not only by its food-cost percentage. Sort the menu by popularity and margin, act differently on each group, and present prices so guests read the dishes before the numbers.
The cost side: what a plate really costs
A recipe cost card lists every ingredient by weight, at its current purchase price, adjusted for yield: the meat lost in trimming, the vegetables peeled, the oil absorbed. Add garnish, sauces, bread on the side, and for delivery the box, bag and cutlery. Most owners who do this for the first time find that some popular dishes cost noticeably more than they thought, because yield and extras were never counted.
Many operators watch food cost as a percentage of the price, and it is a useful alarm. But it can mislead when used alone. Illustratively: a pasta dish that costs 90 and sells for 260 has a food cost of about 35% and leaves 170. A salad that costs 30 and sells for 120 has a "better" 25% food cost but leaves only 90. The rent is paid with the 170 and the 90, not with the percentages. Look at both, and decide with the money.
The feeling side: how guests read a price
Guests rarely know what a dish costs to make. They judge a price against references: the other prices on the same page, what similar places nearby charge, the room, the service, the portion that arrives. A few practical points follow from that.
- The page sets the scale. A high-priced item on a section makes the others look reasonable next to it, the same anchoring and decoy effects Dan Ariely describes in Predictably Irrational (2008). One premium dish can make the dish you most want to sell feel like the sensible choice.
- The format matters. Research from Cornell's Center for Hospitality Research (Sybil Yang, Sheryl Kimes and Mauro Sessarego, "$ or Dollars: Effects of Menu-price Formats on Restaurant Checks", 2009) found that guests given menus with plain numbers, without a currency sign, spent more than those whose menus showed the currency. A price that does not shout "money" lets the dish be read first.
- A price column invites shopping by price. Prices lined up on the right, often joined to the dish by dotted lines, make it easy to scan for the cheapest. Placing the price after the description, in the same type size, keeps attention on the food.
- Words carry value. A clear description (where the ingredient comes from, how it is cooked, what makes it the house version) helps a guest understand a higher price without being told it is worth it.
The tool: the menu engineering grid
Michael Kasavana and Donald Smith's Menu Engineering (1982) gave restaurants a simple way to look at the whole menu at once. Each item is judged on two things: how popular it is compared with the other items in its section, and how much money it leaves per plate compared with the section's average. A common rule in the method counts an item as popular if it sells at least 70% of an equal share of its section. That gives four groups, each with its own action.
| Group | Popularity / money left | What to do |
|---|---|---|
| Star | High / high | Keep it exactly as it is, feature it, protect quality |
| Plowhorse | High / low | Re-cost, adjust portion or sides, small price step, add a paid upgrade |
| Puzzle | Low / high | Move it up the page, rename, describe, photograph, let staff suggest it |
| Dog | Low / low | Remove or rework, unless it serves a clear purpose |
The method, step by step
- Cost cards for every dish. Current prices, real yields, extras and delivery packaging. Output: a costed menu sheet with plate cost and money left for each item, dine-in and delivery separately.
- The sales mix. From the till or the delivery apps, units sold per item over the last four to eight normal weeks (not Ramadan or a holiday). Output: popularity per item within each section.
- The grid. Place every item in one of the four groups. Output: a one-page map of the menu.
- A perception check. Read three nearby competitors' menus for comparable dishes, and ask a handful of regular guests and the floor staff which items people call expensive and which feel like good value. Output: notes on where the price feels wrong, item by item.
- A short list of changes. One action per group, chosen from the table, plus layout changes. Output: a draft menu with the changes marked, and the reason for each.
- Four weeks of measurement. Average check, total money left across the menu, and the sales mix, compared with the weeks before. Output: keep, adjust or reverse each change.
A worked example
This example is hypothetical, and all prices and costs are illustrative. A brunch café in Sheikh Zayed has seen ingredient costs rise. The owner's plan is to raise every price by 15%.
The cost cards show the problem is uneven. The classic burger, the best seller, now costs about 120 to make and sells for 220, leaving 100, below the mains average of about 110. The shakshuka sells for 150, costs 35, leaves 115 and sells strongly. The truffle pasta leaves 170 but few people order it. The cheesecake leaves 60 and barely sells.
The grid puts the shakshuka as a star, the burger as a plowhorse, the pasta as a puzzle and the cheesecake as a dog. The perception check adds one thing: guests describe the burger as fair and the coffee as the item that feels expensive compared with two nearby cafés.
The changes are targeted. The burger goes up a small step and gains a paid "double" upgrade. The pasta moves to the top of the mains with a short description and a photo. The cheesecake is replaced with a dessert that uses ingredients already in the kitchen. The shakshuka stays untouched. Coffee prices stay where they are for now, because the full 15% would have landed hardest on the item guests already find expensive. Currency signs and the right-hand price column are removed. The number to watch for four weeks is total money left per week, with average check alongside it.
Five mistakes we see most
- Pricing by the competitor's menu alone. It happens because it feels safe. But their costs, rent and portions are not yours, and copying their price can mean losing money on a dish they make a profit on. Start from your cost, then look at theirs.
- One mark-up for everything. Multiplying every cost by the same number ignores what guests will pay for each item. Some dishes can carry more; some are the reason guests come in and need to stay sharp.
- Forgetting yield, waste and packaging. The cost card is built from the purchase price of a whole kilo, not the usable part, and delivery boxes are left out. The dish looks profitable and is not.
- Raising everything at once. Guests notice a whole menu going up far more than a few targeted changes, and the stars, the dishes that bring them back, are hit as hard as the dogs.
- The same prices for delivery. Commission and packaging change the money left on every plate. Many restaurants keep a shorter delivery menu, priced separately, built around dishes that travel well.
The strongest objection
"In our area guests only care about price." They care about value, which is price compared with what they get and with what they expected. Even in price-sensitive areas, guests pay more for the dish they come for and notice the one item that feels out of line. And a popular dish priced below its true cost is not generosity. It is a loss that grows with every order.
Honest limits
Menu pricing cannot fix a cost structure that does not work: rent too high for the seats, a kitchen overstaffed for the volume, or too few guests coming in at all. If the room is empty, the brake is elsewhere, and our article on a slow-moving business is a better start. We are not accountants or chefs: recipe development, kitchen operations and your financial model need the right specialists, and we work alongside them rather than instead of them.
How OG helps with menu pricing
Through pricing strategy we work from your cost cards and sales mix to the menu engineering grid, then decide which items to raise, hold, promote or remove, and how dine-in and delivery should differ. Where we need to know how guests see value, customer research asks them. Then the menu has to do its job on the table: print design and production for the layout and the printed menu, and photo production for the dishes you want to move. If the restaurant is not open yet, the decisions before the decor come first. Scope follows the need; there are no fixed packages.
Built on: our pricing work with food and hospitality clients since 2013; Michael Kasavana and Donald Smith, Menu Engineering: A Practical Guide to Menu Analysis (1982); Sybil Yang, Sheryl Kimes and Mauro Sessarego, “$ or Dollars: Effects of Menu-price Formats on Restaurant Checks”, Cornell Hospitality Report (2009); Dan Ariely, Predictably Irrational (2008). The café is a hypothetical example and its numbers are illustrative.
Before closing this page
- When were the cost cards last updated with current ingredient prices and real yields?
- Which best-selling dish leaves the least money per plate?
- Which high-margin dish do guests rarely order, and where is it on the page?
- Which item do guests call expensive, and compared with what?
- Is the delivery menu priced for commission and packaging, or copied from the dine-in one?
