momoscreen

Guide with a worked example

Menu engineering:which dishes to push, change or drop

Menu engineering means checking every dish with two numbers you already have: how often it's ordered and what it earns you per plate. We work it through with Café Lina in Austin, 8 dishes, step by step. At the end you know what to change on your menu.

Pen and paper are enough for the math.

Step 1

Two numbers you already have

Menu engineering needs no software and no consultant. You need two numbers per dish.

  1. That's in the item report of your POS. Take four weeks, so a quiet Monday or a rainy day doesn't skew everything. In September, Café Lina sold 1,000 portions of 8 dishes.

  2. That's the gross profit: price before tax minus food cost. In the US, sales tax goes on the check, so the menu price is already the price before tax. Work out the food cost per portion with our menu price calculator, free and without signing up.

Important: take the amount, not the percentage. On the grilled cheese, food cost is only 19% of the price, on the brisket 38%. Still, the brisket earns 12.08 per plate, the grilled cheese only 9.37.

Step 2

The math with 8 dishes

This is Café Lina's September. All 8 dishes sit in the category “Lunch”; coffee and cake Lina works out separately. Now you draw two lines: one for “popular”, one for “earns well”.

Café Lina, Austin: lunch dishes, September 1 to 28. Menu prices before sales tax.
DishGross­profitPortions
Mac and cheeseStars12.57160
Grilled cheeseStars9.37210
Turkey clubPlowhorses7.55220
Avocado toastPlowhorses6.75150
Smoked brisket platePuzzles12.0880
ShakshukaPuzzles10.4745
BowlDogs7.4260
Goat cheese saladDogs6.5175
Total9,023.401,000

Line 1: when is a dish popular?

  1. 1,000 portions ÷ 8 dishes = 125
  2. 125 × 70% = 87.5

= 88 portions

If all dishes had sold equally well, each would have 125 portions. A dish is popular from 70% of that. Why not 100%? Then almost half the menu would be “unpopular”, including dishes that do perfectly fine. The 70% is the usual rule of thumb.

Line 2: when does a dish earn well?

  1. Gross profit × portions, all 8 together = 9,023.40
  2. 9,023.40 ÷ 1,000 portions

= 9.02 per portion

A dish earns well when it brings in more than the average of your menu. Work out the average with the portions: a dish that goes out 220 times counts more than one that goes out 45 times.

Compare only what belongs together

Work out each category on its own: mains with mains, cake with cake, drinks separately. An espresso always earns less than a brisket plate. That says nothing about whether it belongs on the menu.

Source: The method with the 70% line and the four boxes comes from Michael L. Kasavana and Donald I. Smith: Menu Engineering. A Practical Guide to Menu Analysis. Hospitality Publications, Lansing (Michigan) 1982.

Step 3

The menu engineering matrix: four boxes

Now every dish gets its place. To the right it's about portions, upward it's about gross profit. The two lines split the area into four boxes. Their names are the ones you'll find in books and courses too.

  • Stars

    popular and earn well

    Mac and cheese and grilled cheese. They carry your menu. Leave recipe and price as they are.

  • Plowhorses

    popular, earn little

    Turkey club and avocado toast. Guests love them, but little is left per plate.

  • Puzzles

    earn well, rarely ordered

    Brisket and shakshuka. Good money per plate, but too few guests order them.

  • Dogs

    rarely ordered, earn little

    Bowl and goat cheese salad. They take up space on the menu and work in the kitchen.

Close call: the brisket, with 80 portions, is only 8 below the line. A small nudge may be enough.

Step 4

One move per box

Every box has a clear next step. Make only one move per dish each round, or you won't know later what worked. This is how Café Lina does it with momoscreen.

StarsGrilled cheese

Keep and show

You change nothing about the grilled cheese, neither recipe nor price. Just make sure everyone sees it, for example with your digital menu board at the counter.

In momoscreen, you tick it in the Menu Board under “Products on the board”.

PlowhorsesTurkey club

A small price step

The turkey club goes from 13.75 to 14.50. Gross profit rises from 7.55 to 8.30. If it stays at 220 portions, that's about 165 more gross profit in four weeks. A small step is noticed less than a big one. If you don't want to touch the price, save on purchasing.

In momoscreen, you change the price once under “Change name or price”. Your QR menu shows it automatically, and if you want a printable menu, it's already in there.

PuzzlesSmoked brisket plate

Describe it better

The brisket earns well, but only 80 guests ordered it. Give it a description that makes people hungry: how long it smokes, what comes with it. And put it where guests look, say on the TV.

In momoscreen, you add the description under “Change name or price”. Guests read it in your QR menu.

DogsGoat cheese salad

Drop it or rethink it

Rarely ordered, earns little: the goat cheese salad goes. Its space on the menu goes to dishes that sell. If you'd rather try something new instead, it gets the tag “new”, but only as long as it really is new.

In momoscreen, you delete it under “Edit category”. It then disappears from the QR menu and from your Menu Boards too.

Space on the menu

Where guests look: paper, phone, TV

Where a dish sits helps decide whether it's ordered. But not every rule you read about it is true.

Paper: guests read like a book

You often read about the “golden triangle”: the eye goes to the middle first, then to the top right and top left. The top right is where your best dish belongs. An eye-tracking study by Sybil S. Yang didn't confirm that. The test subjects read the menu more like a book, from left to right and top to bottom, slowly and carefully. She found no “sweet spot”. So a clear order and good descriptions do more for you than a placement trick.

Phone: what's at the top, everyone sees

On a QR code menu, guests scroll from top to bottom. Everyone sees the first category, not everyone the last. Move the category with your stars to the front: in momoscreen, you tap “Sort” on the QR menu page, move it up with the arrow and tap “Save QR menu”. A photo helps too: give your stars one and pick a “With photos” style. When scrolling, the eye stops at the picture first.

TV: only what's on it

A Menu Board shows only the dishes you tick there. Whoever waits at the counter reads what's big on the screen. That's where your stars and puzzles belong, not all 8 dishes.

How to lay out categories, descriptions and prices clearly is in our menu design tips.

Source: Sybil S. Yang: Eye movements on restaurant menus: A revisitation on gaze motion and consumer scanpaths. International Journal of Hospitality Management 31 (2012), pp. 1021–1029. Read the study

Exceptions

When a dog may stay

The matrix knows only two numbers. You know your guests. A dish may stay despite weak numbers if …

DogsBowl

That's why Café Lina's bowl stays even though it's a dog: it's the only vegan dish on the menu.

  • … kids eat it. A kids' dish brings the whole family to the table.

  • … it's the only vegan or gluten-free dish. Otherwise a whole group goes to eat somewhere else.

  • … regulars order it every week. Few portions, but loyal guests who also take coffee and cake.

  • … it sells drinks. A small snack with a lemonade makes its money through the glass.

  • … it's younger than your numbers. If a dish has been on the menu for only two weeks, it doesn't have four weeks of numbers. Give it a full round.

Routine

Every three months, 30 minutes

Menu engineering isn't a project, it's a short appointment. Best at the change of season, when you're touching the menu anyway.

  1. MarchSpring30 min
  2. JuneSummer30 min
  3. SeptemberFall30 min
  4. DecemberWinter30 min

How the half hour goes

  1. 10 min

    Pull the item report of the last four weeks from the POS.

  2. 10 min

    Check the gross profit per dish, enter new purchase prices.

  3. 5 min

    Work out the two lines and put every dish in a box.

  4. 5 min

    Pick one move per box and change it in the app.

Common mistakes

  • Changing everything at once. Then you don't know what worked.

  • Judging after one week. Weather, vacations and holidays skew too much. Take four weeks.

  • Raising all prices at the same time. Guests notice that right away. Only touch the plowhorses.

  • Mixing categories. A cake next to a brisket plate tells you nothing.

Frequently asked questions

What are the four categories of menu engineering?

Stars are popular and earn well. Plowhorses are popular but earn little. Puzzles earn well but are rarely ordered. Dogs are rarely ordered and earn little. Two lines draw the borders: for portions, 70% of the average; for gross profit, the average.

Do I need software or an Excel template for it?

No. Pen, paper and a calculator are enough; a spreadsheet makes it more comfortable. The portions come from your POS: momoscreen counts no sales and isn't connected to any POS. It helps you act on the result, because you change prices, descriptions and dishes in one place.

How often should I do menu engineering?

Every three months, best at the change of season. More often isn't worth it: you need four weeks of numbers, and a change takes time to show.

Does this work for drinks too?

Yes, but separately. Work out coffee, soft drinks and wine each on their own, with their own lines. Sales tax goes on the check, so the menu price is yours. One exception: if a tax is built into a drink price, as some Texas bars do with mixed drinks, take it out first.

What's the difference from pricing a dish?

A menu price calculator works out the price for one dish: what does it have to cost to be worth it? Menu engineering looks at the whole category: which dishes earn you the most, and what do you do with the rest? First the pricing, then the menu engineering.

Is it worth it with a small menu?

Yes, especially there. With 8 to 10 dishes, every spot counts, and the math takes only a few minutes. If a category has only two or three dishes, simply compare them directly.

What does menu psychology have to do with it?

Menu engineering tells you which dishes you want to push. Menu psychology asks how: with placement, description and clear prices. Be careful with tricks like the “golden triangle”: an eye-tracking study showed that guests read a menu more like a book. Good words and a clear order work more reliably.

By momoscreen (JetCoders GmbH, Vienna)Updated:

One move per box, in one place

Prices, descriptions and dishes you keep in momoscreen in one place. Your QR menu is automatically up to date, and Menu Boards show the dishes you tick there.

Start for free

No subscription, no commission.