Menu Engineering: Designing a Menu That Actually Makes Money

For every item on the menu, you need exactly two things:

• Popularity. How many sold in the last 90 days. Your billing system already knows this.
• Contribution margin. Selling price minus the food cost of that dish. Not margin percentage — rupees.

The second one is where most people stop, because working out true food cost per dish by hand is miserable. It needs a recipe attached to every menu item and current purchase rates behind every ingredient. Once you have recipe-linked costing, it is just a report.

Use rupees, not percentages. A dish with 35% food cost selling at ₹600 earns you more per plate than one with 20% food cost selling at ₹150. Percentage thinking pushes you toward cheap dishes that do not pay your rent.

Plot popularity against margin and every dish lands in one of four boxes:

• Stars. Sell well, earn well. Do not touch the recipe, do not touch the price, put them where the eye lands first.
• Plowhorses. Sell well, earn little. These are usually your signature dishes and you cannot remove them. Raise the price slightly, or trim the cost quietly — a cheaper garnish, a smaller side, a better supplier rate.
• Puzzles. Earn well, nobody orders them. The problem is usually visibility or the name, not the dish. Move them up the page, give them a box, let staff recommend them.
• Dogs. Neither popular nor profitable. Cut them. They tie up stock, add prep steps, and make the menu longer for no return.

Where an item sits matters nearly as much as what it costs:

• The top right of a page gets read first. Put stars and puzzles there, not your cheapest item.
• Boxes and borders draw the eye. Use them sparingly — box everything and you have boxed nothing.
• Drop the currency symbol and the trailing zeros. Write 320, not ₹320.00. It reads as a number rather than a cost.
• Never right-align prices in a column. That invites guests to scan down the prices and order the cheapest thing. Put the price right after the description instead.
• Keep each section to six or seven items. Long lists cause decision paralysis, and paralysed guests default to what they know.

One genuinely premium dish at the top of a section makes everything under it look reasonable. It does not need to sell much. It is doing its job by existing.

Beyond that, the reliable lifts to an average bill are boring and they work: a combo that bundles a main with a side and a drink, a clear add-on line under each dish, and staff who actually suggest one. A ₹60 add-on on a third of your covers is a real number by the end of the month.

Menus decay. Ingredient prices move, tastes shift, and dogs accumulate quietly because nobody wants to remove a dish. Once a quarter, pull the numbers and re-sort the four boxes.

If you run a QR menu, this gets easier still — you can change prices and reorder items without reprinting anything, and test whether a change actually moved the mix.

Billing Rasoi gives you sales count and profit per dish in the same report, which is the whole input for this exercise. Check the analytics features, or try it free for 14 days and sort your own menu into the four boxes.