The four-quadrant framework
The classic menu engineering matrix plots every dish on two axes: popularity (how often it is ordered) and profitability (how much gross profit it contributes). This produces four categories:
- Stars — high popularity, high profit. Protect these. Do not tinker with them.
- Ploughhorses — high popularity, lower profit. Can you reduce cost without changing the dish guests love?
- Puzzles — high profit, lower popularity. Can you make them more visible on the menu or train servers to recommend them?
- Dogs — low popularity, low profit. These are candidates for removal or a complete rethink.
The value of this analysis is not in the categories themselves — it is in the decisions they make possible. Without it, a restaurant’s best-performing dish might be sitting in small text at the bottom of a section, while a low-margin crowd-pleaser leads every page.
Price anchoring and the role of the most expensive item
One of the most counterintuitive findings in menu psychology is that the most expensive item on a menu affects how much everyone spends — including people who do not order it. A high-priced anchor makes the items around it appear more reasonable. Remove it, and average spend often drops.
The cheapest item on your menu influences your average spend more than the most expensive one. Customers use it as a reference point for everything else.
What this means operationally
Menu engineering is not a one-time exercise. It should be reviewed quarterly — or whenever you are considering a price change, a seasonal menu update, or a decline in average spend. The data you need already exists in your POS system. The framework is a way of reading it.
The goal is not to optimise every dish into profitability. It is to understand the whole menu as a financial instrument — and make deliberate decisions about what it does and does not do for your business.