A dish can be one of your bestsellers without making as much money as you might expect.
Menu engineering helps restaurant operators across MENA see which dishes are making a strong contribution, which sell well but bring in less per sale, and which may need attention.
A customer favourite can still be a weak performer for the business. Looking at sales and contribution margin together gives managers a better idea of where each dish stands and what to do next.
Menu engineering looks at two things:
The method was developed by Michael L. Kasavana and Donald I. Smith in 1982. Plotting these two measures on a matrix places dishes into four groups:
Each group needs a different approach. A bestselling chicken shawarma with a thin contribution margin needs a different strategy from a premium dish with a strong contribution margin that barely sells.
Stars perform well on both measures. Customers order them often, and they generate a healthy contribution for the restaurant.
The priority is to protect that performance:
Plowhorses are popular but generate less contribution per sale than the restaurant would like.
Removing them can be risky. A popular dish may account for a large share of orders or play an important role in attracting repeat customers. Instead, look for ways to improve its economics without hurting demand.
Consider:
A small change to the price, recipe or add-ons may improve the economics without changing what customers already like.
Puzzles generate a strong contribution when they sell, but customers don't order them often enough.
The problem may be the dish itself, but it could also be the way it is presented. Customers may not understand the ingredients, the description may be too vague, or a more familiar dish may simply get their attention first.
Try:
Dogs perform poorly on both measures. They don't attract many orders and contribute little when they do sell.
That doesn't mean every Dog should be removed immediately.
Give the dish a defined test period. Change its description, presentation, price or menu position, then see whether sales improve.
If demand remains weak and the dish contributes little to the business, removing it can free up menu space, kitchen capacity and ingredients for stronger performers.
Imagine a restaurant in Dubai adds a premium lamb kofta and mixed-grill platter for AED 78.
The food cost is AED 30, leaving a contribution margin of AED 48 per portion.
That is a strong contribution margin compared with several other main dishes.
But the restaurant sells only 90 portions a month, while its chicken shawarma sells 450.
By sales volume alone, the lamb platter looks weak. By contribution margin, it looks attractive.
The lamb platter may therefore be a Puzzle: it makes good money when customers order it, but the restaurant needs to understand why demand is low.
Start with one menu category at a time.
Compare similar dishes rather than mixing completely different categories. Main courses should be compared with main courses, desserts with desserts, and beverages with beverages.
For example, imagine a restaurant analysing eight main dishes:
| Dish | Price | Food cost | Margin per portion | Food cost % | Sales/month | Total contribution |
|---|---|---|---|---|---|---|
| Chicken shawarma platter | AED 42 | AED 15 | AED 27 | 35.7% | 450 | AED 12,150 |
| Mixed grill | AED 68 | AED 27 | AED 41 | 39.7% | 320 | AED 13,120 |
| Chicken mandi | AED 55 | AED 21 | AED 34 | 38.2% | 280 | AED 9,520 |
| Beef burger | AED 52 | AED 22 | AED 30 | 42.3% | 250 | AED 7,500 |
| Lamb kofta platter | AED 78 | AED 30 | AED 48 | 38.5% | 90 | AED 4,320 |
| Seafood machboos | AED 75 | AED 34 | AED 41 | 45.3% | 80 | AED 3,280 |
| Chicken kabsa | AED 50 | AED 23 | AED 27 | 46.0% | 120 | AED 3,240 |
| Grilled hammour | AED 82 | AED 38 | AED 44 | 46.3% | 105 | AED 4,620 |
Add the total contribution from all dishes and divide it by the total number of portions sold.
Total contribution = AED 57,750
Total sales = 1,695 portions
AED 57,750 ÷ 1,695 = AED 34.07
A dish with a contribution margin above AED 34.07 is considered high-margin in this example.
First, calculate average sales per dish:
1,695 ÷ 8 = 211.9 portions
The Kasavana-Smith approach uses 70% of this average as the popularity threshold:
211.9 × 0.7 = 148.3 portions
A dish selling at least 148.3 portions per month is considered popular in this example.
That gives us the following matrix:
| Popularity / Contribution margin | Low contribution margin (< AED 34.07) | High contribution margin (≥ AED 34.07) |
|---|---|---|
| Popular (≥ 148.3 portions) | Plowhorses: Shawarma, burger | Stars: Mixed grill, chicken mandi |
| Unpopular (< 148.3 portions) | Dogs: Kabsa | Puzzles: Lamb kofta, seafood machboos, grilled hammour |
The lamb kofta platter has the highest contribution margin per portion, but it sells only 90 portions a month.
It's therefore a Puzzle, not necessarily a bad dish.
Before removing it, investigate why demand is low.
Price can be one factor.
If AED 78 is the most expensive item in a section, customers may immediately see it as a premium purchase.
A restaurant could test its overall price structure by offering a higher-priced option alongside it. For example, a premium mixed seafood platter at AED 95 could make the AED 78 lamb platter appear more accessible by comparison.
This is known as price anchoring.
But price isn't the only possible explanation. A low-selling dish may also have:
Menu engineering can point you towards these questions; it doesn't answer them on its own.
Contribution margin and popularity tell you a lot about how individual dishes are performing, but they do not show the full picture. A dish can generate strong revenue but have unpredictable demand, while another may contribute less but sell steadily every week. ABC/XYZ analysis helps restaurants look at both sides.
ABC analysis groups items according to their importance to the business.
Depending on the metric being analysed, this can include sales volume, revenue or profit.
The basic categories are:
The exact distribution depends on the metric and methodology being used.
A restaurant may find that a relatively small number of dishes account for a large share of revenue. Changes to the pricing, recipes or availability of those dishes can therefore have a significant effect on the business.
XYZ analysis looks at demand stability.
It asks a simple question: How predictable are this item's sales?
Generally:
The coefficient of variation can be used to measure demand variability:
Coefficient of variation = standard deviation ÷ average demand
A lower coefficient indicates more consistent demand.
ABC tells you how important an item is to the business.
XYZ tells you how predictable its demand is.
Together, they give managers more context when deciding what to do with each dish.
| ABC/XYZ classification | Demand pattern | What to do |
|---|---|---|
| A + X | High importance, stable demand | Protect availability and consistency |
| A + Y | High importance, changing demand | Monitor closely |
| A + Z | High importance, unpredictable demand | Control stock and investigate demand |
| B + X | Medium importance, stable demand | Maintain and develop |
| B + Y | Medium importance, variable demand | Review regularly |
| B + Z | Medium importance, unpredictable demand | Test and monitor |
| C + X | Lower importance, stable demand | Optimise where practical |
| C + Y | Lower importance, variable demand | Review performance |
| C + Z | Low importance, unpredictable demand | Consider reducing or removing |
A popular, profitable dish with stable demand may need protection from stock shortages and recipe changes.
A high-margin dish with unpredictable demand may need closer attention to purchasing and stock.
A low-performing item with unstable demand may deserve a more serious review.
Menu engineering depends on having reliable sales and cost data.
Syrve provides restaurant operators across MENA with reporting and analytics tools, including ABC/XYZ menu analysis. Its platform also provides customisable sales and Online Analytical Processing (OLAP) reports, allowing users to create reports using different parameters.
Managers can use this data to identify:
Syrve also supports menu updates across multiple locations, allowing restaurant groups to customise menus for individual branches and respond to local preferences, seasonal changes and nearby events.
For groups operating across different MENA markets, that can be useful when customer preferences and demand vary between locations.
The 2×2 matrix answers two questions:
How strong is this dish's contribution margin?
How popular is it?
ABC/XYZ adds another:
How predictable is its performance?
A mixed grill may be a Star because it has strong demand and a healthy contribution margin. If demand is also stable, keep it consistently available, maintain the recipe and avoid unnecessary changes.
A lamb kofta platter may be a Puzzle because it has a strong contribution margin but weaker demand. Test its description, placement, presentation or pricing before deciding whether to remove it.
A seasonal dish may sell well during certain periods but become less predictable outside them. Adjust purchasing and stock levels to match the seasonal demand rather than treating its lower off-season sales as a permanent problem.
The goal is to understand why each dish performs the way it does and take the appropriate action.
A practical menu engineering process:
Which matters more, low food cost or high contribution margin?
For profit, contribution margin in currency matters more than the percentage. A dish with a higher food cost percentage can still generate more contribution per portion.
Should dine-in and delivery menus be analysed separately?
Yes. Commissions, packaging and logistics can change a dish's economics. Analyse each channel separately to see which items remain profitable after channel-specific costs.
How often do dishes move between categories?
Regularly. Seasonality, promotions, ingredient costs and customer demand can change both popularity and contribution margin. Menu engineering should be an ongoing process, not a one-time audit.
Can poorly performing dishes be saved instead of removed?
Often. Try changing the name, presentation, menu placement, recipe or price first. Removal should come after testing reasonable improvements.
Where should you start with a menu of over 100 items?
Start with the categories generating the most revenue, such as mains, grills, and beverages. Analyse those first, then move to the remaining categories.