Menu Engineering 101: Stars, Plowhorses, Puzzles and Dogs

Menu engineering, formalized by Michael Kasavana and Donald Smith in the 1980s, is the discipline of looking at every dish through two lenses at once: popularity and contribution margin. It sounds academic; it is the highest-leverage spreadsheet a restaurant can build.

The four quadrants

Plot every dish by how often it sells and how much margin each sale contributes, and each lands in one of four boxes:

• •Stars — popular and profitable. Protect them: keep quality consistent, feature them visually, never let a cost increase quietly erode the margin.
• •Plowhorses — popular but low-margin. The crowd-pleasers. Options: re-engineer the recipe cost, nudge the price carefully, or pair them with high-margin sides and drinks.
• •Puzzles — profitable but rarely ordered. Usually a visibility or naming problem. Rename, describe better, photograph, move up the category, or feature them for a week and watch.
• •Dogs — neither popular nor profitable. Retire them. Every dog on the menu costs prep complexity, inventory and attention.

Margin, not margin percentage

The classic mistake is ranking dishes by food-cost percentage. A dish with 20% food cost that contributes 80 Kč per sale is worse for you than one with 35% food cost contributing 140 Kč. You bank crowns, not percentages. Always work with contribution in currency.

Where the data comes from

The popularity axis used to require exporting POS reports and cleaning them by hand — which is why most operators did the exercise once and never again. A digital menu adds a second, earlier signal: views. When a dish is seen often but ordered rarely, that gap is diagnostic — the description or price is losing the sale at the last moment. ProMenu's analytics show exactly this per-dish view data, alongside likes, so you can spot puzzles before the quarter's sales report.

Act on one quadrant per month

The matrix is only useful if it changes the menu. A sustainable rhythm: each month, retire one dog, promote one puzzle (better photo, better description, better position), and re-cost one plowhorse. Twelve months of that discipline transforms a menu's economics without a single dramatic decision.

Takeaway

Menu engineering is not a one-off project but a monthly habit: two axes, four boxes, one action each. With a digital menu the data collection is passive and the changes are instant — which removes the two excuses that kept most menus unexamined.