What is the menu engineering matrix?
The Kasavana-Smith model evaluates each menu item with two questions:
- 1How often does it sell? (popularity)
- 2How much does it earn? (profitability)
Those axes produce four decision zones.
Four categories and the right action
1) Stars (high popularity + high profitability)
Place them in high-visibility areas, protect recipe consistency, and prioritize server recommendation scripts.
2) Plowhorses (high popularity + low profitability)
Improve margin through portion design, garnish strategy, add-ons, and production-cost optimization.
3) Puzzles (low popularity + high profitability)
Test naming, menu placement, description language, and visual framing. Often the issue is presentation, not the dish itself.
4) Dogs (low popularity + low profitability)
They consume prep capacity and inventory. Remove them unless there is a strategic reason to keep them.
30-day implementation plan
- 1Days 1-7: Clean item-level sales and cost data
- 2Days 8-14: Classify into the four groups
- 3Days 15-21: Update placement and pricing language
- 4Days 22-30: Measure and iterate
Use this with the core menu engineering guide and food cost guide.
Conclusion
The matrix creates value only when used in cadence. Teams that update menu decisions monthly grow margin by system, not by guesswork.
