Menus
Menu engineering made simple: which dishes to push and which to drop
By Sulakshi (Sue) Pallethenna
Every café has that one dish. The one the chef loves, that takes up three containers in the fridge, and that maybe four people order a week. Nobody wants to be the one to say it, so it just… stays.
Menu engineering is how you say it with numbers instead of feelings.
It sounds fancy, but it's really just looking at two things for every dish: how popular it is and how much money it makes each time you sell it. Put those together and every dish falls into one of four groups. Once you see the groups, you know what to push, what to fix and what to quietly retire.
It's common for a dish to stay on the menu out of habit, long after it stopped selling or making money. Menu engineering gives you the numbers to make that call with confidence.
What you'll need
- Your sales by item for a set period (a month is a good start). Your POS should give you this.
- The selling price of each dish (excluding GST).
- The food cost per plate of each dish. If you haven't worked these out yet, start with my guide on how to work out your real food cost per plate.
Do one menu section at a time, like breakfast mains or lunch, because comparing a muffin to a big brunch plate doesn't tell you much.
The two numbers that matter
1. Contribution margin (profit per plate)
This is what each sale leaves you after the ingredients:
Contribution margin = selling price (excl. GST) − food cost per plate
Notice this is a dollar figure, not a percentage. A dish with a higher food cost percentage can still put more dollars in your pocket per sale. That's one of the most useful lessons menu engineering teaches.
2. Popularity (menu mix)
This is each dish's share of the total number of dishes sold in that section:
Menu mix % = number sold of this dish ÷ total sold in the section × 100
Setting the benchmarks
To sort your dishes, you need a line for "high" and "low" on each number.
Profit benchmark: work out the average contribution margin for the section. Total contribution (each dish's margin × number sold, added together) divided by the total number of dishes sold. Dishes above it are high profit; below it are low.
Popularity benchmark: a common rule of thumb is 70% of an equal share. If you have 8 dishes, an equal share would be 12.5%, so the line is 12.5% × 0.7 = 8.75%. Dishes above it are popular; below it are not. (Why 70%? Because in practice a few dishes always lead, so a straight equal share would make too many dishes look "unpopular".)
The four groups
⭐ Stars: popular and profitable Your heroes. Keep them consistent, keep them visible and don't mess with them.
🐴 Plowhorses: popular, but lower profit People love them, but they don't earn much per sale. Don't drop them, because your regulars will notice. Instead, gently improve the margin.
🧩 Puzzles: profitable, but not popular They make good money when they sell, but they don't sell often. Figure out why. Is it the name, the description, where it sits on the menu, or the price?
🐶 Dogs: not popular and not profitable These are the candidates to drop or rework. Be honest with yourself here.
A worked example
All of these numbers are made up to show the method. They're not real café data or recommended prices.
Here's an example breakfast section with six dishes, over one month:
| Dish | Sold | Price excl. GST | Food cost | Margin per plate | Total margin |
|---|---|---|---|---|---|
| Smashed avo | 400 | $20.00 | $6.00 | $14.00 | $5,600 |
| Big breakfast | 300 | $24.00 | $9.00 | $15.00 | $4,500 |
| Eggs on toast | 350 | $12.00 | $3.00 | $9.00 | $3,150 |
| Chilli scrambled eggs | 60 | $21.00 | $5.00 | $16.00 | $960 |
| Acai bowl | 140 | $16.00 | $6.50 | $9.50 | $1,330 |
| Corn fritters | 50 | $19.00 | $7.00 | $12.00 | $600 |
| Total | 1,300 | $16,140 |
Profit benchmark: $16,140 ÷ 1,300 = about $12.42 per plate.
Popularity benchmark: 6 dishes, so an equal share is 16.7%. 70% of that is about 11.7%.
Now the menu mix for each dish:
| Dish | Menu mix | Popular? | Margin vs $12.42 | Group |
|---|---|---|---|---|
| Smashed avo | 30.8% | Yes | Higher | ⭐ Star |
| Big breakfast | 23.1% | Yes | Higher | ⭐ Star |
| Eggs on toast | 26.9% | Yes | Lower | 🐴 Plowhorse |
| Chilli scrambled eggs | 4.6% | No | Higher | 🧩 Puzzle |
| Acai bowl | 10.8% | No | Lower | 🐶 Dog |
| Corn fritters | 3.8% | No | Lower | 🐶 Dog |
Interesting, right? The eggs on toast is the cheapest dish to make, with the lowest food cost of anything here. But it only leaves $9 a plate, so it's a plowhorse. Meanwhile the chilli scrambled eggs makes the most per plate but hardly anyone orders it.
What to do with each group
Stars: protect them
- Keep the recipe and portion consistent (an SOP helps a lot).
- Put them where eyes land first on your menu and specials board.
- Train staff to recommend them.
- Be careful with price rises. Small and occasional is fine; a big jump can turn a star into a plowhorse.
Plowhorses: nudge the margin up
- Look for a small portion tweak on the most expensive ingredient.
- Offer paid add-ons (bacon, avocado, halloumi) that lift the average spend.
- Consider a modest price increase. Popular dishes can often handle it.
- Pair them with a high-margin drink or side.
Often the fix is small: a slightly smaller portion of the most expensive ingredient, a cheaper garnish, or a few dollars added to the price.
Puzzles: help them sell
- Rename it or rewrite the description. "Chilli scrambled eggs" might do better with a more tempting description of what makes it special.
- Move it to a better spot on the menu, or feature it as a special.
- Get staff to taste it so they can recommend it with confidence.
- Check the price. If it's a lot higher than nearby dishes, it might be scaring people off.
- If you've genuinely tried and it still won't move, it might be time to let it go.
Dogs: rework or remove
- Ask whether the dish has a reason to stay. Is it the only vegan option, or a regular's favourite you're happy to keep?
- If it stays, rework it: change the ingredients, cut costs or turn it into a special.
- If it has no reason to stay, take it off. Fewer dishes usually means less stock, less waste and a faster kitchen.
How often to do this
Once a quarter is a good rhythm for most cafés, plus whenever you change your menu or a key ingredient price moves a lot. Seasons matter too, so don't judge your acai bowl on its July numbers alone.
Common traps
- Judging by food cost % alone. Look at dollars per plate too.
- Mixing sections. Compare breakfast with breakfast, not with cakes.
- Using too short a period. One week can be thrown off by a rainy Saturday or a public holiday.
- Forgetting the human side. Numbers tell you what's happening, but your team and regulars can tell you why.
FAQ
What's the difference between a plowhorse and a dog? Both earn less than average per plate, but plowhorses are popular and dogs aren't. That's why you improve a plowhorse's margin but question whether a dog should stay at all.
Should I always remove dogs from the menu? Not always. Some dishes earn their place in other ways, like being your only gluten-free option. The point is to keep them on purpose, not by habit.
Can I do menu engineering without a POS system? Yes, but it's harder. You can tally orders from dockets for a few weeks. A POS that reports sales by item makes it much quicker and more accurate.