What Is Menu Engineering?

Menu engineering has a bad reputation because too many versions turn into a laminated four-box chart that nobody trusts. The useful version is much simpler: what are guests buying, what does each dish really cost today, and which items deserve action?

That question matters more when input prices are moving. USDA ERS reported that food-away-from-home prices were 3.6% higher in April 2026 than April 2025, while food-at-home prices were 2.9% higher.¹ ERS also forecast 2026 food-away-from-home prices to rise 3.5%, with a 2.8% to 4.2% prediction interval.² BLS full-service meals and snacks were 3.8% higher in May 2026 than May 2025.³

Broad inflation does not tell a chef whether to change the steak frites, the Caesar, or the brunch sandwich. Menu engineering does that job dish by dish.

What is menu engineering?

Menu engineering is the process of sorting menu items by two plain questions:

1. Popularity: how many units did the dish sell?
2. Contribution margin: how many dollars does each sale keep after food cost?

The basic contribution margin formula is:

Menu price − current plate cost = contribution margin.

Food cost percentage still matters, but it can fool you. A $9 item at 24% food cost may keep less money than a $29 item at 35% food cost. Owners pay rent and payroll with margin dollars, not percentages.

Good menu engineering starts with recipe costing. If the plate cost is wrong, every menu decision built on it is wrong too.

The four menu engineering buckets, in kitchen English

The classic menu matrix uses four buckets. The labels matter less than the action.

Bucket	What it means	What to do first
High sales, high margin	Guests buy it and it pays.	Protect it. Keep prep, price, and quality steady.
High sales, low margin	Guests buy it, but it may be leaking money.	Check portion, supplier price, yield, and price.
Low sales, high margin	It pays when it sells, but not enough guests order it.	Rename, move, train servers, or feature it.
Low sales, low margin	It does not sell much and does not pay much.	Fix the recipe or remove it.

Do not treat the chart like law. It is a starting point. A low-margin dish might be a signature item, a traffic driver, or part of a prix fixe path. A low-sales item might be needed for dietary coverage. The review should make those tradeoffs visible, not pretend every dish is the same.

A worked example with a burger and a salad

This is why menu engineering should connect to invoice data. USDA said beef and veal prices rose 3.1% from March 2026 to April 2026 and were 14.8% higher than April 2025.⁴ If beef is a top seller, last month's cost sheet can get stale quickly.

What numbers do you need?

You do not need a giant system to start. You need clean enough numbers for the dishes that matter.

1. Menu price: what the guest pays before tax and tip.
2. POS units sold: item counts for the review period.
3. Recipe quantities: ounces, grams, eaches, and sauces by dish.
4. Usable yield: trim, cook loss, drained weight, peeled weight, and portion loss.
5. Current supplier price: the latest invoice or approved quote, not the old sheet.
6. Contribution margin dollars: margin per plate and total margin for the period.

Then rank the menu by total margin dollars. That keeps attention on the items that can move the business this month. A tiny issue on a huge seller can matter more than a dramatic issue on a dish that sells twice a week.

How often should you run it?

Run a full menu engineering review monthly or quarterly, depending on how large and stable the menu is. Do a faster weekly check on volatile, high-volume items. Beef, dairy, eggs, seafood, coffee, oil, and produce can all move enough to change the answer.

This does not mean rewriting the menu every week. It means keeping the facts current. Sometimes the answer is no price change. Sometimes it is a portion check, supplier quote, prep yield review, server note, special placement, or menu wording change.

If you only look at menu engineering once a year, the review becomes archaeology. It explains the margin you already lost.

What should change after the review?

A good menu engineering review ends with a small set of changes, not a 40-row spreadsheet nobody acts on. For each dish, write the next move in plain language.

For a high-sales, low-margin item, the move may be to reprice by $1, tighten the scoop size, switch the garnish, test a smaller portion, or ask the supplier for a better quote on the main ingredient. For a low-sales, high-margin item, the move may be to rename it, move it higher on the printed menu, add a server note, photograph it, or run it as a special for a week.

The best reviews also name what not to touch. If a dish is a house signature and the margin is acceptable, leave it alone. If a low-margin item brings people in the door and the rest of the check pays, say that out loud. Menu engineering should protect the story of the restaurant while making the money math honest.

The mistake that breaks menu engineering

The common mistake is using stale plate costs. A dish might look healthy because the recipe sheet still has the old mozzarella case, old fryer oil price, old beef yield, or old pack size. Then the menu matrix says “winner,” while the invoice says otherwise.

That is the same pattern behind supplier price tracking. Price changes only matter when they are tied to the dishes, sales volume, and choices they affect. Otherwise they sit in the invoice pile until month-end.

Mornay's angle is simple: keep the restaurant's cost memory current from invoices, supplier prices, recipes, POS data, and operator corrections. Then the menu review can show what changed, which dishes moved, how many dollars are at stake, and what decision needs approval.

Menu engineering should not be a poster. It should be a short owner conversation with numbers everyone trusts.