Which dishes on your menu actually make you money?

Ask most owners which dish makes them the most money and they will name their best seller. Those are different questions, and the answers are usually different dishes.

A dish that sells 300 times a month at ₹40 of margin contributes more than one that sells 40 times at ₹120. And a popular dish with a thin margin can quietly be the reason a busy restaurant is not a profitable one.

The method, in four steps

This is menu engineering — a standard framework in the industry, and you can run it on paper in an afternoon.

Step 1: Pull one month of item-wise sales. How many times each dish sold. One month minimum, ideally three to smooth out anomalies.

Step 2: Work out the contribution margin per dish. Selling price minus the cost of the ingredients. Not your full cost with rent and salaries — just food cost. Contribution margin is what the dish contributes toward covering everything else.

Step 3: Find your two midpoints. Average sales volume per dish, and average contribution margin per dish. These two lines split your menu into four groups.

Step 4: Sort every dish into one of the four.

The four groups and what to do with each

High popularity, high margin — your stars. These are the dishes your restaurant should be known for. Put them where the eye lands first on the menu, make sure staff recommend them by name, and protect their quality obsessively. Do not discount them; they do not need it.

High popularity, low margin — your workhorses. Loved, ordered constantly, and barely paying for themselves. Do not remove them, because they are often why people come. Instead: re-cost the recipe, check the portion, negotiate the main ingredient's price, or raise the price modestly. On a dish selling 300 times a month, ₹10 is meaningful money and is rarely the reason anyone stops ordering.

Low popularity, high margin — your puzzles. Good money when they sell, and they do not sell. Usually a visibility or description problem rather than a food problem. Try renaming, repositioning on the menu, a photo, or making it the dish staff suggest when asked. Give it a fair trial before judging it.

Low popularity, low margin — your dogs. They do not sell and they do not pay. Every one of these costs you inventory, prep time, menu space, and kitchen complexity. Remove them. Owners resist this more than any other step, usually for one regular customer's sake.

The менu-size problem nobody wants to hear

If this exercise puts a third of your menu in the dogs quadrant, the finding is not about those dishes individually — it is that your menu is too big.

A large menu costs you in ways that never appear as a line item: more SKUs in inventory, more spoilage, longer prep, slower service, more training, and more scope for the portion drift that inflates food cost. Restaurants that cut their menu meaningfully usually find food cost improves and service speeds up at the same time.

Doing it without software

Entirely possible, and worth doing once by hand because the manual version teaches you the logic.

You need a month of item-wise sales — countable from bill books if you have the patience — and a costed recipe for each dish. Our recipe cost calculator is free and handles the second part without a signup.

The catch is repeatability. Done by hand, this is a full day, so it gets done once and never again. Its value comes from running it every quarter, because prices, suppliers, and customer preferences all move.

Doing it in ten minutes instead

If your billing already captures every order, item-wise sales is a report rather than a counting exercise. In reports and analytics you can pull item-wise sales for any period and export to Excel, then drop in your per-dish costs and sort.

The point is not the report. It is that a quarterly ten-minute exercise actually gets done, while a quarterly full-day exercise does not.

Start here

Do not analyse the whole menu on your first attempt. Take your top twenty dishes by volume — they are most of your revenue — and sort just those four ways. The decisions that come out of twenty dishes will be worth more than a perfect analysis of 120 that you never finish.