top of page
Metafoodx Logo Horizontal.png

Menu Engineering: How to Use the Menu Engineering Matrix in a Commercial Kitchen

5 hours ago
5 min read
A restaurant menu displayed on a table

Menu engineering is a method for analyzing a menu by two measures at once: how often each item sells and how much money each item contributes after its food cost. Plotting every item on those two measures produces the menu engineering matrix, which sorts items into four groups, usually called stars, plowhorses, puzzles and dogs. Each group calls for a different action, from featuring an item more prominently to changing its price, portion or recipe, to taking it off the menu.


The method was developed by Michael Kasavana and Donald Smith at Michigan State University's School of Hospitality Business in the early 1980s, as the American Hotel and Lodging Educational Institute explains. It remains one of the most widely taught tools in hospitality management. This guide covers how to build the matrix, what to do with each group, and how to adapt it for hotels, college dining and other high-volume kitchens.


The two measures: popularity and contribution margin

Popularity is measured by menu mix, each item's share of total sales within its category. The standard threshold, described in the open textbook Basic Kitchen and Food Service Management/03%3A_Food_Costing/3.05%3A_The_Principles_of_Menu_Engineering), is the 70% rule. Expected popularity is 100% divided by the number of items in the category, and an item counts as popular if its share of sales is at least 70% of that expected figure. A 1995 paper in the FIU Hospitality Review describes the same rule, noting that in an eight-item category the threshold works out to 8.75% of sales.


Contribution margin is an item's selling price minus its food cost, in dollars. It shows how much each sale contributes toward labor, overhead and profit. The textbook compares each item's contribution margin with the average contribution margin for the category, total contribution margin divided by the number of items sold. Items at or above the average are classed as profitable. The textbook stresses that the goal is high contribution margins, not simply low food cost percentages, since an item with a higher food cost percentage can still bring in more money per sale. Our guide to how to price a menu using food cost explains contribution margin in more detail.


The four quadrants and what to do with each

Stars are popular and profitable. The textbook recommends keeping the recipe unchanged, featuring the item prominently, having staff suggest it, and testing a small price increase.


Plowhorses are popular but less profitable. Guests like them, so removing them is risky. The textbook suggests raising the price gradually, reducing the cost of accompaniments, or adjusting the portion size, and moving the item to a less prominent spot if it still underperforms.


Puzzles are profitable but not popular. These items earn well when they sell, so the aim is to sell more of them: reposition them on the menu, have staff recommend them, or add value, without cutting the price so far that they stop being profitable.


Dogs are neither popular nor profitable. The textbook calls them "obvious candidates to remove from the menu." The FIU paper adds a note of caution, warning that it would be improper to delete or change an item "based solely on its individual popularity," and extending the model to account for the labor each item requires.


How menu engineering connects to production

Menu engineering is usually framed as a pricing and marketing tool, but it also has a production side. Plowhorses and stars sell in volume, so they are the items where a production plan that runs high wastes the most food. Dogs are often overproduced because they are prepared in case someone orders them. Pairing the matrix with data on how much of each item is produced and left over shows which items cost the kitchen money twice: once in a low contribution margin and again in food that is never sold. Our guide to finding the five items driving your leftovers covers that side of the analysis, and our post on the 30/30/30, 60/40 and 80/20 rules explains why a short list of items usually matters most.


Adapting menu engineering for high-volume dining

The classic matrix assumes each item is sold separately at its own price. That fits restaurants, hotel outlets and campus retail, but not all-you-care-to-eat dining halls, buffets or banquets, where guests do not pay per item. In those settings, popularity can be measured by how much of each item guests actually take or eat, and cost can be measured per portion served, including the food that was produced and not eaten. The resulting picture answers similar questions: which items guests choose most, which ones cost the most per portion consumed, and which ones the kitchen keeps producing that guests pass over.


Cycle menus make this easier, because the same items come around on a regular schedule and results can be compared from one turn of the cycle to the next. Our guide to cycle menus covers how to use that rhythm for planning.


Where Metafoodx fits

The Metafoodx food operations intelligence platform measures what each commercial kitchen produces, serves and has left over by menu item, which gives high-volume kitchens the popularity and cost data menu engineering needs even where nothing is sold item by item. A weekly report ranks the five high-cost overproduction items for each venue, and forecasting and Automatic Production Planning adjust quantities for the next service. At Mayan Princess Beach & Dive Resort, the general manager said the platform helped the team see "which dishes were being overproduced and which ones were actually driving guest demand," as described in the resort's case study.


FAQ

What is menu engineering?

A method of analyzing menu items by popularity and contribution margin to decide which items to promote, reprice, change or remove. It was developed by Michael Kasavana and Donald Smith at Michigan State University.


What are the four categories of the menu engineering matrix?

Stars (popular and profitable), plowhorses (popular but less profitable), puzzles (profitable but not popular) and dogs (neither popular nor profitable).


What is the 70% rule in menu engineering?

An item counts as popular if its share of sales in its category is at least 70% of the expected share, which is 100% divided by the number of items in the category.


How is contribution margin used in menu engineering?

Each item's contribution margin, its price minus its food cost, is compared with the category's average. Items at or above the average are classed as profitable.


Can menu engineering work in a dining hall or buffet?

Yes, with adjustments. Popularity can be measured by how much of each item guests take or eat, and cost by cost per portion served, including food produced and not eaten.


Metafoodx is a food operations intelligence platform that captures the food journey from prep to plate, so foodservice teams can lower food cost, produce more accurately, and prevent waste before it happens. Ready to see how we can help transform your food service operation? Book a demo with our team today.

bottom of page