Pricing a menu without guessing comes down to three numbers per dish: what the plate costs to put on the table, what share of the price that cost should be, and how many dollars are left after the food is paid for. Everything in the textbooks is a way of arranging those three numbers, and the arithmetic is small enough to do on the back of an order pad.

The Correct Way to Calculate Menu Price (Not the 3x Rule), Learn with Owner.com on YouTube

I should be clear about what this post is. I work at the Heritage Diner, the family restaurant in Mount Sinai that opened in 2000, and I am not going to describe how the diner sets its prices. This post is about the published methods, and every formula and example below comes from a source linked where I use it. The one thing I can add from the diner is mechanical: its menu is synced from a single menu in Square, so a price gets changed once and not once per place it appears.

Start with the cost of the plate

Plate cost, which the textbooks call portion cost, is the price of everything that goes on one serving: the protein, the sides, the garnish, the oil in the pan. If the recipe makes four portions, divide the ingredient cost by four. Anything left out of that number comes straight out of the profit later, and the small items are the ones that get left out: the oil, the garnish, the roll that arrives with the plate. They look like nothing on the invoice and add up across a thousand covers.

Cook's hands plating a burger with fries at the pass

The hard part is keeping the number current. Ingredient prices move, and a cost worked out last spring can be wrong by autumn. Webstaurant’s article on determining restaurant menu prices, last updated in November 2024, begins its methods with the same step, totaling the cost of all ingredients in the dish before any percentage is applied. Webstaurant sells restaurant supplies, so it has an interest in operators buying things, but the order of operations it gives is the standard one.

Food cost percentage and the mark-up

The first formula is the familiar one. The open textbook Basic Kitchen and Food Service Management defines food cost percentage as portion cost divided by selling price. Its example is a portion that costs $4.80 and sells for $14.00, which works out to about 34 percent.

Run the other way, the same relationship gives a price. Divide the plate cost by the food cost percentage you want. Webstaurant’s example takes a $4.00 raw food cost and a 28 percent target and arrives at $14.29, which it suggests rounding to something cleaner. The textbook phrases the same idea as a mark-up: selling price divided by portion cost, or one divided by the food cost percentage. A 34 percent food cost is a mark-up of about 2.92, and a $1.05 bowl of soup at a mark-up of 3.6 sells for $3.78 before the price is rounded, often to end in a 5 or a 9.

What percentage should you aim for? Two operator-facing sources land in the same place. Lightspeed’s menu pricing article, published in June 2023, says most operators keep food costs between 28 and 35 percent of revenue, and Webstaurant gives the same range for most restaurants. Both are vendors, and neither is a study. The range is a common rule of thumb, and your own costs for rent, labor and everything else decide whether it fits you.

Lightspeed works one example with a hamburger whose serving costs $3.60 and sells for $9.00, which is a 40 percent food cost, and says a price of $12.00 would bring that burger down to 30 percent. The arithmetic checks out. Whether customers will pay $12.00 for that burger in that room is a question the formula cannot answer.

Working backward from what the restaurant has to keep

The mark-up method assumes you know the right percentage. The textbook also describes a target-based approach that finds it. Take projected sales, subtract the labor, overhead and profit you need, and what remains is the money available for food. In the book’s example, $10,000 in sales with $7,500 going to other costs and the owner’s profit leaves $2,500 for food, which is 25 percent, a mark-up of 4.

Calculator, pencil, ledger and supplier invoices on a wooden desk

That approach shows why a single industry percentage is a poor target. A restaurant with a heavy labor load, such as a place that makes everything by hand, has less room for food cost than one that buys most things prepared. If you are trying to work out your own labor number, the post on restaurant labor cost and scheduling is the place to start.

The textbook is blunt about the alternative. It describes guessing, or pricing at whatever the market will bear without reference to your own costs, as unreliable. I would add that the market does matter, but it works as a check at the end of the process and cannot replace the first three steps.

Percentages hide dollars

A low food cost percentage is not the same as a profitable dish. The textbook shows this with a steak and a chicken dish. The steak has a 38 percent food cost and the chicken 27 percent, and yet the steak leaves a contribution margin of $15.00 against $12.00 for the chicken. Contribution margin is the selling price minus the cost, in dollars. The register collects dollars, and a percentage is only a ratio between two of them.

The textbook’s advice follows from that: the goal is to keep contribution margins high, not to hold the food cost percentage as low as possible. Bar & Restaurant’s article on why menu engineering matters says the same thing in the context of its origin. Menu engineering was developed in 1982 by Michael Kasavana and Donald Smith at Michigan State University, and Kasavana argues that operators should focus on dollars earned instead of food cost percentages, so that a higher-priced item can be more valuable even with a higher percentage cost.

The practical use is plain. When a supplier raises the price of beef, the question to ask is how many dollars the dish still leaves, and whether the price or the portion should change. Chasing the percentage back down can lead to a smaller plate and a customer who notices.

Sorting the menu by popularity and margin

Menu engineering adds the second axis, how often each item sells. The same article names four groups: stars, which are popular and highly profitable; plowhorses, which are popular but less profitable; puzzles, which are profitable but under-ordered; and dogs, which are low on both counts. Stars get protected and featured. Plowhorses get their margins improved through portion, ingredient or modest price changes. Puzzles get better descriptions, better placement or a server’s recommendation. Dogs are candidates for removal.

Steak, chicken breast, vegetables and herbs arranged on a cutting board

The open textbook describes the same grid without those names, as four boxes. An item counts as popular in its method if its sales reach 70 percent of expected popularity, where expected popularity is 100 percent divided by the number of items in the category, so with seven appetizers the threshold is about 10 percent. For popular items with weak margins it suggests raising prices gradually, cutting the cost of accompaniments or reducing portion size. For profitable items that do not sell, it suggests repositioning them or having servers suggest them.

None of this works without sales data by item, which is why a point-of-sale report matters here. If you need help getting the item numbers out, I wrote about reading a restaurant sales report.

Looking at the neighbors

Competitor prices are the check at the end. Webstaurant describes pricing against competition as a market-based approach with no formula: take local competitors’ prices as a baseline, then choose to match them, price lower for a casual or cost-conscious concept, or price higher for an upscale one. Lightspeed lists other approaches too, including value-based, premium, bundle and dynamic pricing, with bundle pricing meaning discounts on combinations, such as a sandwich, side and drink.

I would run the neighbors’ prices after the cost work, since a menu copied from the place down the road inherits that place’s costs, and yours are different. Then I would look for the dishes where the two methods disagree. Those are the ones that deserve a second look at either the recipe or the price.

After the price is set

A price is also a promise on every screen that shows it. A change that has to be typed into five places will eventually be typed into four. At the diner the menu is synced from one Square menu, so the change is made once, and Square’s description of its online ordering profile says the profile builds its catalog from the restaurant’s menus and menu items. If your point-of-sale system works the same way, the sensible time to review prices is on a schedule, with the sales report and the latest invoices open side by side.

For informational purposes only. Food cost ranges are rules of thumb from vendor sources, so consult a licensed accountant or financial advisor before basing pricing or investment decisions on them.

Sources