Stars, Plough horses, Puzzles and Dogs: the idea behind menu engineering

Ask most people why a dish is on the menu and you'll get an answer about taste, tradition or what sells well. All true — and all incomplete. The thing that decides whether a menu makes money isn't any single dish; it's the mix, and how each item balances two things at once: how often it sells, and how much it makes when it does.

Where the idea came from

In 1982, two academics at Michigan State University's School of Hospitality Business — Michael Kasavana and Donald Smith — published a short book called Menu Engineering: A Practical Guide to Menu Analysis.[1] They borrowed a trick from corporate strategy. A decade earlier, the Boston Consulting Group had taught big companies to sort their products into a grid of "Stars" and "Dogs" by growth and market share.[2] Kasavana and Smith pointed the same idea at the one thing every restaurant already has: a menu.

Their move was simple but clever. Stop judging dishes one number at a time. Plot every item by two measures at once — its contribution margin (the cash left after the food cost) and its popularity (how many sell) — and the menu sorts itself into four natural groups.

The four jobs a dish can do

• Stars — popular and profitable. These are your heroes. Protect them, give them the best spots on the page, and never change them by accident.
• Plough horses — popular but low-margin. They pull people in but don't pay their way. Nudge the price, shave the cost, or use them to sell something better alongside.
• Puzzles — profitable but overlooked. The money's there; the sales aren't. Rename them, describe them better, move them to where the eye lands.
• Dogs — unpopular and unprofitable. They take up space, prep time and attention. Rework them, or take them off.

Why it still works

Because it forces a decision. A long menu is easy to leave alone — everything seems to be earning its keep until you look closely. The matrix makes the trade-offs impossible to ignore: it tells you which dishes to push, which to fix, and which are simply costing you money. It isn't about cutting the cheap dishes or pushing the expensive ones. It's about knowing the job each item does and acting on it. The maths is straightforward; the discipline of actually doing it is where the value sits. Four decades of menu research have refined the details — costing variants, layout and design studies — but reviews of the field still treat Kasavana and Smith's matrix as the starting point for menu analysis.[3]

More from the menu you already have

You don't need to sell a single extra cover to make more money from the menu you already have. Feature a Star, reprice a Plough horse, fix a Puzzle, drop a Dog — a handful of small moves, repeated across a full menu, compound into real profit. And once you've seen your menu this way, you tend to keep seeing it: every new dish gets the same simple test before it earns a place on the page.