Menu engineering is a way of looking at a menu not as a list of dishes, but as a portfolio of items that each earn money differently. The method is old, simple, and still works: every dish is plotted along two axes — popularity and profit margin — and lands in one of four quadrants: stars, plowhorses, puzzles, and dogs. Each quadrant calls for its own set of actions.
The real value of the method isn't the labels — it's the shift in perspective: menu decisions start getting made from sales data, not from the chef's or owner's gut feeling.
What You'll Need
The analysis needs just two numbers per item over a period (usually a month or a season):
- Units sold — pulled from the POS system.
- Margin in absolute terms — selling price minus the ingredient cost per the recipe card.
Important: margin is calculated in dollars per serving, not as a percentage. A dish with a 40% food cost that nets $400 per sale is more valuable to the register than one with a 20% food cost and $150 of margin. The percentage tells you about purchasing efficiency; the dollar amount tells you what the restaurant actually keeps.
It only makes sense to analyze within categories (mains with mains, desserts with desserts): soups and steaks have fundamentally different demand patterns, and lumping them together makes the comparison meaningless.
Building the Matrix
- Calculate the average margin for the category: total margin of every dish sold in the category divided by the number of servings sold.
- Calculate the popularity threshold. The classic approach: an item's share of sales is compared against its "fair" share (100% divided by the number of items in the category), multiplied by a discount factor — typically 70%. Put simply: if a category has 10 dishes, an item counts as popular once its share of sales tops roughly 7%.
- Every dish gets two flags: margin above or below average, popularity above or below the threshold.
That produces four quadrants:
| Margin above average | Margin below average | |
|---|---|---|
| Sales above threshold | Stars | Plowhorses |
| Sales below threshold | Puzzles | Dogs |
Stars: Protect Them
Popular and profitable. These are the backbone of the menu, and the main job is not to mess them up.
- Don't touch the recipe to save money. Trying to "optimize" a star is the most expensive way to cut costs — guests notice.
- Keep it consistent. A star has to taste the same on any day, on any shift — this is where strict adherence to the recipe card matters most.
- Give it visibility. The best real estate on the menu — the top of a category, a highlighted box — rightfully belongs to stars.
- Be careful with price. Stars usually have room to raise prices, but do it in small steps, and not for all of them at once.
Plowhorses: Raise the Margin
They sell great and earn little. Guests love them, so cutting them is out — the economics need careful improvement instead:
- Rebuild the cost structure. Often one expensive component that isn't critical to the flavor is eating the margin: a side, a garnish, a portion of sauce. A targeted swap or a smaller portion solves it without the guest noticing.
- Check the price against similar items. Plowhorses are often simply underpriced relative to comparable dishes. A small price increase on a popular item shows up clearly given the volume.
- Build profitable pairings. A server who suggests a high-margin drink or appetizer alongside a plowhorse lifts the check's overall margin without touching the dish itself.
Puzzles: Promote or Simplify
High-margin but poor sellers. There are usually three causes, and each has its own fix:
- The dish is invisible. Buried at the bottom of the category, with no description or photo. Fixed by repositioning it in the menu, writing a proper description, and adding it to server recommendations.
- The dish isn't understood. The name needs explaining, the ingredient list scares people off with unfamiliar words. Fixed by renaming it and writing the description in plain language: clear, appetizing wording sells better than culinary jargon.
- Nobody actually wants the dish. Sometimes it's simply an item the kitchen finds interesting and guests don't. If a month of promotion doesn't move sales, it's a candidate for the dogs quadrant.
Puzzles are the only quadrant worth experimenting on before deciding to cut anything — the margin justifies the effort.
Dogs: Cut Them Without Sentiment
Unpopular and low-margin. Every dog ties up inventory in storage, adds an extra line to prep, and spreads the kitchen's attention thinner for nothing.
The standard move is to cut it. Exceptions are rare, but they exist:
- the item serves an important segment's needs (kids' options, gluten-free) and brings in whole groups;
- its ingredients fully overlap with other dishes and don't create any dedicated leftover stock;
- it's a signature, image-defining dish for the concept — in which case it's a deliberate cost of positioning, not just something left hanging around.
A Menu Is a System, Not a List
The matrix is the core of the method, but a few rules around it make it work harder:
- Menu size should match the kitchen's capacity. A compact menu with overlapping ingredients holds a more consistent quality and cheaper inventory than a "phone book" of a menu. Every new item has to earn its spot.
- Regularity. Engineering isn't a one-time exercise: purchase prices shift, tastes change. A sensible rhythm is to revisit the matrix once a season, and recheck item margins whenever purchase prices move noticeably. If recipe cards and sales data live in the same system (in Cenaly, for example), the recalculation comes down to a single report.
- Seasonal and event menus. Short-run specials are a testing ground for future stars: a new dish gets validated at low volume, without risking the core menu.
- Servers know the map. The floor team needs to understand which items to push first — not "upselling the expensive stuff," but confidently recommending what guests genuinely like and what's profitable for the restaurant.
Key takeaways
- Plot dishes along two axes — margin in absolute terms and popularity — within each menu category.
- Protect stars from "optimization," raise plowhorses' margin through cost structure and pairings, promote puzzles with better descriptions and visibility, and cut dogs.
- Calculate margin in absolute terms per serving, not as a food-cost percentage.
- Keep the menu compact: every item has to earn its place through sales or its role on the menu.
- Repeat the analysis regularly — the matrix goes stale along with purchase prices and the season.