Menu Engineering: Turning Your Menu Into a Profit Tool

Menu engineering is the practice of analysing each dish on two axes — how profitable it is and how often it sells — and then changing the menu based on where each dish lands.

It exists because food costing answers only half the question. Costing tells you what a dish returns when someone orders it. Menu engineering tells you what a dish returns to the business, which depends entirely on whether anyone does.

The two numbers you need

For every dish on the menu, over a defined period — a month is usually enough — you need:

1. Cash gross profit per serve. Ex-GST selling price minus true plated cost. Cash, not percentage. This matters and is covered below.
2. Units sold. Straight off the POS.

Multiply them and you have each dish's total contribution. Sort that list and you usually have a surprise: the dish the owner is proudest of is rarely in the top three, and something unglamorous is quietly paying the rent.

The matrix

Plot each dish against the menu's average gross profit and average unit sales. Four quadrants fall out.

Stars — high profit, high popularity

These work. The instruction is to protect them, and the most common mistake is meddling. Do not reduce the portion, do not substitute a cheaper ingredient, do not move them off the prime position on the page. If you raise the price, raise it in small increments and watch the volume.

Plowhorses — low profit, high popularity

Everyone orders it and it barely pays. This quadrant holds the most recoverable money on most menus, because a small change multiplied by high volume is a large number.

The levers, roughly in order of how much damage they do: raise the price modestly; reduce plate cost without touching what the customer values — usually the garnish, the accompaniment or the oil rather than the protein; or add a paid accompaniment. Cutting the portion of the thing people actually come for is the one move that reliably backfires.

Puzzles — high profit, low popularity

It makes good money when it sells, and it does not sell. This is normally a description, position or price-perception problem rather than a dish problem.

Move it to a position the eye actually reaches. Rewrite the description so it says what the dish is rather than what it contains. Have the floor recommend it for a fortnight. If none of that moves it after a genuine attempt, it becomes a dog — and it is the section of the matrix where owners waste the most time before accepting that.

Dogs — low profit, low popularity

Take it off. The exception worth respecting is a dish carrying strategic weight — the one vegetarian main, a dish for a regular who spends heavily, a legacy dish the venue is known for. Everything else goes, and it takes its orphan ingredients with it, which is where the real saving is.

Why cash gross profit, not percentage

This is the point that changes decisions.

Take two dishes. A pasta costing $5.20 at $26 on the menu runs 22% food cost and returns $18.44 ex-GST. A snapper costing $12.80 at $42 runs 33.5% and returns $25.38.

The pasta wins on percentage. The snapper returns nearly seven dollars more every time it leaves the pass. If you engineered this menu by percentage alone you would promote the pasta and quietly discourage the fish, and you would be making the business less money while congratulating yourself on the food cost.

Percentage is useful for spotting dishes that are drifting. Cash is what pays the rent.

Where the menu itself does the work

Once you know which dishes you want to sell, the physical menu is the cheapest lever you have. No kitchen change, no cost change.

• Position. On a single-page menu, attention concentrates near the top of each section and on the first two items. Put stars and puzzles there; do not waste those positions on dishes that sell anyway.
• Section length. Around five to seven items per section. Beyond that people stop reading and default to whatever is familiar — which is usually your plowhorse.
• Prices without dollar signs, not right-aligned in a column. A column of right-aligned prices invites reading down the prices and choosing the cheapest. Set the price after the description in the same type.
• Boxes and highlights, sparingly. One highlighted dish gets attention. Four highlighted dishes get none.
• Descriptions that name provenance. "Bronte pistachio" outsells "pistachio" at a higher price, because it tells the customer why it costs what it costs.

Doing it in practice

A workable cycle for most venues:

1. Pull a month of POS data by item.
2. Cost each dish properly — yielded weights, everything on the plate.
3. Calculate cash gross profit per dish and total contribution.
4. Plot the four quadrants against your menu averages.
5. Change one thing per quadrant, not five.
6. Wait a month and measure before changing anything else.

The discipline in step five and six is what separates menu engineering from menu fiddling. Change six things at once and you will have no idea which one worked, which means you cannot repeat it next quarter.