Menu Engineering for Restaurants Explained

In short: menu engineering is a simple way to see which dishes make your money. For each dish you work out how many guests order it (popularity) and how much cash it leaves after ingredient cost (contribution margin). That gives four groups: stars, plowhorses, puzzles and dogs. Each group gets a different action.

Menu engineering was described in 1982 by Michael Kasavana and Donald Smith in their book on menu analysis. Many articles and tools repeat their four-box idea today. It is a decision aid, not a law: it will not run your restaurant, but it is a very good way to spot what to look at first.

The two measures

Popularity

How often a dish is ordered compared with the other dishes on the same menu section. Use sales from a fixed period of at least four weeks. A restaurant's reports show top products by quantity, so you do not need to count by hand.

Contribution margin

The money one sale leaves after the cost of the ingredients: selling price (excluding sales tax or VAT) minus the cost of the ingredients on the plate. Work this out with recipe costing.

Use money, not percentages. A dish with a 40% food cost that sells at 20 leaves 12. A dish with a 25% food cost that sells at 8 leaves 6. The first has the worse percentage and the better contribution. Percentages tell you about efficiency, but money pays the rent. See food cost percentage for how the two relate.

The four groups

How to set the dividing lines

You need two cut-offs. Most guides use your own menu's averages, not industry numbers:

• Margin line: the average contribution margin across all dishes, weighted by how many were sold.
• Popularity line: a common version of the classic method says a dish is popular if it sells at least 70% of an even share. With 10 dishes, an even share is 10%, so the line is 7% of all sales.

Both lines are conventions. Use them to start, then adjust if they put an obviously popular dish in the wrong group.

A worked example

A made up example with round numbers, for one menu section of eight main courses over four weeks.

Total sold is 1,600. The popularity line for eight dishes is 70% of 12.5%, or 8.75%. The weighted average margin is about 8.48 (the total margin of 13,574 divided by 1,600).

Now sort:

• Stars (popular, margin above 8.48): steak frites, and the chicken burger (its 8.40 is just under the line, so check it).
• Plowhorses (popular, lower margin): margherita pizza, fish and chips, pasta of the day.
• Puzzles (not popular, margin above average): mushroom risotto, lamb shank.
• Dogs (not popular, low margin): caesar salad.

The burger is on the line. In practice you would treat it as a star, since it sells the most and its margin is close. This is why the lines are a start and not the final word.

What to do with each group

Stars: protect and promote

Keep quality and consistency high, make sure they are never unavailable, and position them well. Put them where eyes go first on the menu or the ordering page. Do not discount them.

Plowhorses: improve the margin without losing the guests

Options, from gentlest to boldest:

1. Check the recipe cost. Are portions bigger than the spec? Has a supplier raised a price? See recipe costing.
2. Adjust the portion or an expensive garnish slightly.
3. Raise the price a little. Popular dishes with loyal guests often bear a small rise. See how to price a menu.
4. Add a paid extra or a pairing, such as a side or a drink, to raise the average spend.
5. Make it part of a combo where a better margin item balances it.

Puzzles: help guests discover them

A good margin and low sales usually means guests do not see the dish or do not trust it. Try a better name and description, a photo online, a spot as a server recommendation, or a tasting. Online, move it higher in its category. If it still does not sell after a few weeks, it may not suit your guests.

Dogs: decide

Ask whether the dish has a job. Does it serve vegetarians, children or a group order? Does it share ingredients with other dishes, so removing it saves little? If it has no job, remove it and make the menu shorter. A shorter menu cuts waste and speeds the kitchen. See reduce food waste.

Include the whole picture

Menu engineering looks only at sales and ingredient cost. Before acting, also think about:

• Labour. A dish with a high margin that takes three cooks ten minutes is not a star in a rush.
• Kitchen capacity. A dish that blocks the fryer or the grill at peak slows everything else. See ticket times.
• Ingredient overlap. Dishes that share ingredients keep stock turning and reduce waste.
• Allergens and dietary needs. Never remove the only dish for a group without a plan. Keep allergen data up to date. See allergens and calories.
• Guests who come for one dish. Some low margin dishes are the reason people visit.

Treat channels separately

A dish may be a star in the dining room and a dog for delivery, because it travels badly or because delivery fees change the price. Analyse dine in, collection and delivery separately where you can. See takeaway and delivery profit.

Drinks and desserts should also be analysed on their own, since their margins are very different from mains.

A simple routine

1. Export four weeks of sales by dish.
2. Update the recipe cost for each dish.
3. Work out margin per dish and the two lines.
4. Sort into four groups.
5. Choose three actions, with one owner for each.
6. Make the changes, note the date.
7. Check again after four weeks. Keep what works.

If you use DineHarbor, reports give you the sales side, and the weekly numbers guide shows how to read them. If you keep ingredient costs in your stock records, your costs are easier to keep current. See stock control.

What menu engineering cannot tell you

It cannot tell you why a dish sells. It cannot replace a conversation with your chef and your servers. And it does not know about next month's supplier price rise. Use it to ask better questions, not to avoid them.

Acting on the example

Look back at the eight dishes. The steak and the burger carry the menu, so the first rule is to keep them available and consistent. The pizza, fish and chips and the pasta sell well but leave less per plate, so check their recipe cost first. Perhaps the fish has crept up in price, or the pasta portion has grown. A small, honest price rise on the pizza of 0.50 would add about 150 to the month on 300 sold, with no new work.

The risotto and the lamb are puzzles. The risotto leaves 10.00 a plate and sold only 90. Moving it to the top of the vegetarian group, adding a clear photo online and asking servers to recommend it might lift it. The lamb is a long cooked dish that suits a smaller number of guests; a weekly special might suit it better than a permanent place. And the salad is the one dog. If it is the only light dish, keep it; if guests choose other lighter dishes, replace it with something that shares ingredients with the kitchen's other prep.

Ready to bring your menu online and watch what sells? Start the 30 day free trial with no card needed.