Menu engineering with data: a 2027 guide for Indian restaurants

By Restrologic team Published Updated

The menu is the biggest profit lever a restaurant controls, yet many Indian restaurants change it based on chef instinct and complaints. This guide shows how to use sales analytics to build a menu that earns both volume and margin.

Why does menu engineering matter in India?

A typical casual dining or QSR menu in India has dozens of active SKUs across several categories. More options sounds like more choice and more revenue. The data usually says otherwise. A small share of dishes brings most of the revenue, and a long tail of slow items adds complexity, waste, tied-up stock and a slower kitchen.

Menu engineering uses sales, cost and preference data to decide which dishes to promote, reprice, remove or develop. With rising food costs, aggregator commissions and thin margins, profit per dish is a basic management practice and not a luxury.

What is the menu engineering matrix?

The framework was developed by Kasavana and Smith in 1982, and it is still the most useful tool for Indian operators. It sorts every dish on two scales: popularity (sales against the menu average) and profitability (contribution margin against the menu average).

What are stars?

Stars are popular and profitable. A well-placed biryani, a signature curry or a bestselling dessert made from low-cost ingredients at a premium price fits. Feature stars in physical and digital menus, and on Swiggy and Zomato place them in recommended sections and at the top of their category. Protect their quality and do not discount them without a reason.

What are plowhorses?

Plowhorses sell well but earn little per unit. Staples such as dal makhani, plain naan or jeera rice often land here, because they are priced low or cost more than expected. They are the easiest to act on. A small price rise, a better supplier rate or tighter portions can move them toward star status without losing volume.

What are puzzles?

Puzzles earn well when ordered but are not ordered often. They may be hard to see, poorly described or unknown. On delivery platforms, give them featured placement, better photos or a place in combos. A tandoori speciality with a strong margin that few guests order can become a real contributor if its order rate rises.

What are dogs?

Dogs take menu space, prep effort and stock for little return. Most should come off the menu. Use judgement in India. A regional speciality may stay for brand identity, and some apparent dogs are seasonal items judged on too little data. Investigate before you cut anything with strategic value.

How do you calculate contribution margin per dish?

The profitability axis needs contribution margin, not only gross margin. Contribution margin is the menu price minus the raw material cost per serving and, for delivery, the packaging cost per order. It isolates variable cost and leaves out fixed overhead.

For delivery orders, go one step further. The same dish carries costs on Swiggy and Zomato that it does not carry at the table: the platform commission, any tax on that commission you cannot claim back, the share of discounts you fund and its own packaging. A dish that is a star in the dining room can be a plowhorse or a dog on delivery once those are counted. Classify delivery on delivery numbers.

You need a recipe costing database that links each item to its ingredients and standard portions, plus current purchase prices. In many chains this sits in an inventory system or a chef’s cost card spreadsheet. Bring it into your analytics next to POS sales volume and the full classification becomes possible at scale. In Restrologic POS, recipes and stock deduction are built in, which keeps these costs current.

How does one dish compare at the table and on delivery?

Here is a worked example with assumed figures, for one dish sold at the same menu price in both channels. The commission rate and every other number are examples, not any platform’s terms. Replace them with your own.

Line Dine-in (₹) Delivery (₹)
Menu price 320.00 320.00
Less: discount you fund 0.00 32.00
Less: commission (example rate of 20 percent on the price after discount) 0.00 57.60
Less: GST on commission you cannot claim back (18 percent of the commission) 0.00 10.37
Less: ingredient cost 96.00 96.00
Less: packaging 0.00 24.00
Contribution per plate 224.00 100.03

On these figures the dish earns less than half as much per delivery order as it does at the table. If it also sells in high volume on delivery, it is a plowhorse there, and the fix is the price, the portion, the packaging or the offer, not the recipe. Run your own dishes through our cloud kitchen profit calculator to see the margin per order and per month on your real commission, packaging, ad spend and food cost.

Do dishes perform differently across regions?

Yes, and this matters for any chain in several cities. A North Indian chain that expands south may find that staples which are stars in Delhi are slow sellers in Chennai or Bengaluru. The dishes are fine. Local taste, familiarity and competing cuisines create different demand.

A single national menu strategy therefore underperforms one built from outlet-level data. For larger chains, build regional reports that group outlets by city or market and rank items by contribution within each group. Regional managers can then decide for their own markets.

How can you A/B test menu changes?

Delivery platforms let you test a change on some outlets before rolling it out chain-wide. Both Swiggy and Zomato allow menu edits at outlet level, so you can run different menus in different outlets and compare. With Restrologic POS, or any POS connected through our Swiggy, Zomato and Ownly integration, a menu change made in the POS reaches every delivery app at once.

Choose two similar outlets (similar market, customers and baseline revenue). Change the menu at one and leave the other alone. Run the test for several weeks so day-of-week effects even out, then compare revenue per order, item volumes and average order value. Good things to test:

  • a price rise on a plowhorse, to see whether volume falls by more than the margin gain
  • featuring a puzzle in the recommended section
  • moving desserts higher in the category order

What should you measure?

  • item-level order volume for the changed items
  • category-level volume, in case the change moved orders between categories
  • average order value
  • rating trends, if you changed portions or recipes
  • revenue per outlet per day, the final test of whether the change helped

How do you plan the menu by season?

Seasonal demand in India is often managed by instinct. Mango dishes peak in summer and then fade. Soups and warm drinks rise in winter in the north and centre. The wedding and festival season drives party and large-format orders. Monsoon moves demand toward comfort food in Mumbai and Pune.

Analytics makes this planned. Look at item-level demand by month over past years, link it to calendar events and decide in advance what to feature, remove or introduce. A seasonality calendar built from two or more years of sales is one of the most useful exercises a chain can do, and it lets you time Swiggy promotions to the dishes in demand.

What should change in your menu review for 2027?

Menu engineering itself has not changed since 1982. What has changed is how many costs sit between the menu price and the kitchen’s margin, and how many places the menu lives. A menu review for 2027 should cover these points:

  1. Re-cost every recipe at current prices. Cost cards drift. Use the last month’s purchase prices, not the prices from when the dish launched.
  2. Classify each channel separately. Run the matrix once for dine-in and once for each delivery app, using that channel’s own volumes and contribution.
  3. Check the commission base and the offers. Confirm whether commission is charged before or after the discounts you fund, and cost every running offer as a price cut. Our payout reconciliation guide shows how to check this on a real statement.
  4. Plan for more than two delivery apps. Ownly, the delivery app from Rapido, was live in Bengaluru, Hyderabad and Kolkata by November 2026. If you list there too, keep one menu master in your POS so prices, items and switches reach every app together. Our Ownly guide tracks where it runs.
  5. Cut dogs before the busy season. Fewer items means faster prep, less stock and fewer stockout cancellations at the dinner peak. Our guide to cutting cancellations and rejections explains why menus that are out of step with the kitchen cost ranking as well as revenue.
  6. Set a review rhythm. Look at delivery dishes every month and the whole menu every quarter. Write down what you changed and why, so the next review can tell whether it worked.

How long should you test a change?

Long enough to cover at least two or three full weeks, so each day of the week appears more than once, and longer if the change is small. A price change on a plowhorse needs enough orders on both sides of the change to tell a real fall in volume from an ordinary slow week. Avoid running a test across a festival or a long holiday, when demand changes for reasons that have nothing to do with the menu.

How does Restrologic help?

Our restaurant analytics includes a menu performance view that classifies your whole menu into stars, plowhorses, puzzles and dogs and refreshes it as sales flow in from your POS and aggregators. It uses your recipe costs for true margin-based classification, and for multi-city chains it adds regional breakdowns. See how it would look for your menu and get in touch.

Quick answers

What is menu engineering?

Menu engineering sorts every dish by popularity and contribution margin into four groups, stars, plowhorses, puzzles and dogs, so you know which dishes to promote, reprice, rework or remove.

How do you calculate contribution margin for a delivery dish?

Take the price the customer pays, then subtract the discount you fund, the platform commission and any tax on it you cannot claim back, the ingredient cost and the packaging for that dish. What is left is the dish's delivery contribution.

How often should a restaurant run menu engineering?

Check delivery menus every month, because commissions, offers and ingredient prices move quickly, and run a full review of the whole menu at least every quarter and before each big season.

Should the delivery menu be the same as the dine-in menu?

Not always. A dish can be a star at the table and a dog on delivery once commission and packaging are counted, and some dishes travel badly. Classify each channel on its own numbers before deciding.

Tell us what you want to build

Send a short note and Deepak will get back to you.

Deepak Kumar, Head of Operations +91 70275 16740 deepak@restrologic.com