Before a Chennai restaurant reprints its lunch menu, the POS export and recipe cards should meet on the same desk. The best-selling dish may leave fewer rupees per plate than a quieter one. The dish with the neatest food-cost percentage may contribute least of all.
That review has a current reason to happen. September reporting on Indian restaurants using smaller plates describes chefs reconsidering portions, variety and seasonal dishes. Those accounts do not establish a national trend or prove that a smaller plate earns more. They make the purchasing and menu question concrete: what changes when a guest chooses one dish instead of another?
Start with a comparable set of dishes
Compare lunch mains with lunch mains at one outlet and in one sales channel. A side dish ordered with a main serves a different purpose. Combining both in a single ranking can turn a useful accompaniment into an apparent failure.
The basic inputs are selling price, portion food cost and units sold. Northern Arizona University’s menu-engineering exercise sets out contribution as menu price less food cost, alongside each item’s share of sales. Its four labels describe high or low contribution and popularity: star, plowhorse, puzzle and dog.
For the working sheet here, use realised selling prices after restaurant-funded discounts and refunds, excluding GST. Keep recipe costs on a consistent accounting basis. If discounted sales are mixed with full-price sales, calculate their average realised price rather than copying the printed menu.
Recipe contribution per plate = realised selling price − portion food cost
This contribution still has other bills to pay. It is not the restaurant’s net profit. Portion food cost should include the actual accompaniments and usable yield, while rent, salaries and other costs remain outside this first screen.
A recipe card also cannot explain every stock loss. The separate guide to stock-based food-cost percentage explains the opening-stock, purchases and closing-stock calculation. Keep that outlet-level check beside the dish sheet, without treating the two as interchangeable.
Work the rupees before assigning labels
The table below is an illustrative lunch menu, not a customer case or a set of recommended Chennai prices. All prices, costs and counts are chosen assumptions. Each dish is sold as a complete lunch option during the same review period.
| Lunch option | Net price | Portion food cost | Units sold | Contribution per plate | Total contribution |
|---|---|---|---|---|---|
| Vegetable meals | ₹180 | ₹60 | 160 | ₹120 | ₹19,200 |
| Vegetable pulao | ₹220 | ₹70 | 100 | ₹150 | ₹15,000 |
| Mushroom rice | ₹260 | ₹100 | 40 | ₹160 | ₹6,400 |
| Curd rice | ₹140 | ₹40 | 20 | ₹100 | ₹2,000 |
| Total | 320 | ₹42,600 |
Mushroom rice has the highest food-cost percentage: ₹100 ÷ ₹260, or about 38.5%. It also leaves the most recipe contribution per plate, ₹160. Curd rice has the lowest percentage, about 28.6%, but leaves ₹100. Ranking by percentage alone reverses the per-plate contribution ranking of those two dishes.
Volume changes the picture again. Vegetable meals contributes ₹19,200 across the period, more than any other row. Removing it because its ₹120 per-plate contribution is below average would put a large part of this menu’s contribution at risk.
For this example, choose a sales-weighted contribution benchmark:
₹42,600 total contribution ÷ 320 plates = ₹133.125 per plate, rounded to ₹133.13 for display.
This weighting is a stated method choice. The NAU exercise instead demonstrates a simple average across menu items. Neither benchmark should be silently swapped midway through a comparison; record which your sheet uses.
For popularity, use the exercise’s conventional screen of 70% of an equal share. With four competing options, that is 0.70 ÷ 4 = 17.5% of units sold. This is a classification convention, not a minimum viable demand level for an Indian restaurant.
| Lunch option | Share of units | Position against these thresholds |
|---|---|---|
| Vegetable meals | 50% | Plowhorse: popular, lower contribution per plate |
| Vegetable pulao | 31.25% | Star: popular, higher contribution per plate |
| Mushroom rice | 12.5% | Puzzle: quieter, higher contribution per plate |
| Curd rice | 6.25% | Dog: quieter, lower contribution per plate |
Find out what the promoted sale replaces
Suppose a menu change sells 20 more mushroom-rice lunches. At ₹160 contribution each, those plates account for ₹3,200. That does not automatically mean the restaurant gained ₹3,200.
If those guests would otherwise have ordered vegetable pulao, the category gives up 20 × ₹150 = ₹3,000. The increase is only:
20 × (₹160 − ₹150) = ₹200, before any other costs change.
In this explicitly fixed-volume scenario, category contribution rises from ₹42,600 to ₹42,800, while total plates stay at 320. If all 20 orders were additional business instead, the recipe-contribution increase would be ₹3,200. These are different assumptions, so the report needs to say which is supported by the sales record.
This concern has original research behind it. Breffni Noone and Guillaume Cachia’s 2020 study examines within-category substitution using two years of data from 48 outlets of a US steakhouse chain. It supports considering relationships between dishes when changing pricing and placement. It does not supply an Indian uplift forecast or validate the illustrative rupee results above.
Twenty extra plates of one dish can mean twenty fewer plates of another.
A POS comparison can flag that possibility without proving why each guest switched. Festival dates, weather, opening hours and stockouts may also change the mix. Keep those explanations visible rather than assigning every movement to the new menu design.
The kitchen adds another constraint. If the promoted dish needs the same crowded pan station at lunch, extra demand may bring delays or overtime. Record the actual service consequence before assuming the higher recipe contribution reaches the bottom of the operating statement.
Turn one menu decision into a small test
Start with a change the team can identify afterwards: one description, one placement or one price. Changing the whole page, portion and promotion together makes a before-and-after result difficult to interpret.
The following is a proposed operating routine, not a research-tested promise of savings.
A menu review the next shift can repeat
Record the outlet, channel, meal period, dates and competing dishes. Note closures, unavailable items and any promotion. Keep cancelled orders out of fulfilled unit counts and reconcile refunds separately.
Match POS item codes to current recipe cards. Check portion yield, accompaniments and actual selling prices. A renamed dish should not disappear into a new code without its history being noted.
State which item changes, why, and what would count as an acceptable outcome. Keep the benchmark method and cost boundary with the sheet. Include a service or waste limit that matters to this kitchen.
Compare contribution across all competing dishes over matching trading periods. Check unit counts, stockouts, ticket delays and waste alongside the promoted item. Record what to retain, reverse or test again.
Where a dish needs its own short-lived preparation, include unsold prep in the review. A recipe card based only on plates sold will miss it. The hotel food-waste log provides a way to record where food leaves the process and whether the log itself is complete.
A useful decision record can be brief: the dish, the change, the comparison dates, category contribution and one explanation for the result. Keep the old sheet. When the next purchase rate or menu price changes, the team needs to see what moved rather than rebuilding the argument from memory.