Which menu items should you promote, reprice, or cut after a menu engineering review?

What menu engineering actually measures

Menu engineering is a simple idea with a long history in the restaurant business. Take every item in a category, say entrees, and plot it on two axes: how many you sold over a period (popularity) and how many dollars each sale contributed after plate cost (contribution margin). The average of each axis splits the chart into four quadrants, which most people know as stars, plowhorses, puzzles, and dogs. Related: How do you cost a recipe accurately down to the true plate cost?

The important words are contribution margin in dollars, not food cost percentage. A dish with a low food cost percentage and a low price can put fewer dollars in the register than a dish with a higher percentage and a higher price. Percentages tell you how efficiently you buy; dollars tell you what pays the rent. Menu engineering forces the conversation onto dollars, which is why it changes decisions that percentage-only reviews miss. Related: What is food cost percentage and why does it matter for a menu?

Keep reading: How do you cost a recipe accurately down to the true plate cost?, What is food cost percentage and why does it matter for a menu?, How should you price a menu using both plate cost and target margin?. See how PlateCostr helps you recipe food-cost and menu margin calculator.

Stars and plowhorses: protect and adjust

Stars are popular and high margin. Do not touch the recipe, do not shave the portion, and do not bury them on the page. Put them where eyes land first, make sure the line can execute them consistently on a busy night, and train servers to describe them well. The main risk with stars is complacency: a supplier price creep on a key ingredient can quietly drop a star into the plowhorse quadrant if nobody re-costs it. Related: How should you price a menu using both plate cost and target margin?

Plowhorses are popular but low margin. Guests love them, which means you have some room to move. Options include a modest price increase, a small portion or garnish adjustment that guests will not notice, a cheaper sub-recipe swap that preserves the flavor, or pairing them with a high-margin side or drink. Move one variable at a time and watch the sales count for a few weeks. If popularity holds, the change worked.

Puzzles and dogs: fix or remove

Puzzles are high margin but unpopular. The dish makes money when it sells, it just does not sell. Before cutting it, ask whether the name, description, or placement is the problem. Renaming, adding a server recommendation, moving it near the top of the section, or lowering the price slightly can turn a puzzle into a star. Give it one honest attempt with a defined window, then decide.

Dogs are unpopular and low margin. These are the easiest calls and the ones operators most often avoid making. Every dog occupies inventory, prep time, and menu real estate that a better item could use. Unless a dog serves a real purpose, such as the only vegetarian option or an owner's signature, remove it or rebuild it from scratch. Cutting two or three dogs usually simplifies the walk-in more than any other single change.

Running the review so it sticks

A menu engineering review is only as good as the inputs. You need accurate plate costs for every item, which means current ingredient prices and realistic yields, and you need item-level sales counts from the point of sale for a period long enough to smooth out one-off events. A month is typical for a full-service menu; a quarter gives more stable results for items that sell only a few times a week. Related: How do you protect margins when ingredient prices keep rising fast?

Repeat the analysis on a schedule, quarterly for most independents, and compare quadrant moves over time rather than reading each report in isolation. When plate costs live in a costing tool that already tracks price changes, the review takes an afternoon instead of a week, and the interesting work becomes the decisions, not the arithmetic.