Make a partner menu easier to choose—and worth selling

Menu engineering for a US or Canadian restaurant

Menu engineering compares item contribution with popularity, but the definitions need to be explicit. For this international teaching exercise, contribution means menu revenue minus the item-variable costs included in the worksheet. It is not net profit after rent and the whole payroll. Record the currency, portion, channel, price basis and included costs before comparing dishes. The existing calculator above is labelled INR; do not mistake it for a USD/CAD conversion tool.

In a fictional USD menu, pasta sells for 18 and has 7 of included variable cost, producing 11 per sale; 180 sales generate 1,980 contribution. A burger sells for 16 with 8 cost, producing 8; 220 sales generate 1,760. A salad sells for 20 with 8 cost, producing 12; 60 sales generate 720. Soup sells for 10 with 5 cost, producing 5; 40 sales generate 200. The total is 500 items and USD 4,660 contribution on this deliberately limited scope.

For demonstration only, choose popularity of at least 100 sales and contribution of at least USD 10 as the high thresholds. These are invented teaching cutoffs, not an industry benchmark. Pasta is high on both dimensions; the burger is popular with lower contribution; the salad has higher contribution but fewer sales; soup is lower on both. In a CAD case, recalculate using actual CAD prices and recipe costs rather than simply relabelling a US menu.

These groups suggest questions, not automatic removals. For pasta, check recipe consistency and whether production can handle demand. For the burger, investigate whether an evidence-backed portion or purchasing improvement preserves the customer proposition. For salad, test a clearer description or suitable placement and verify that it fits what guests want. For soup, ask whether it provides an important dietary option, supports another sale or uses ingredients efficiently before judging it solely by two thresholds.

Delivery can change the classification. Packaging, channel fees and restaurant-funded incentives can reduce the contribution of a dish that looks attractive on a dine-in worksheet. Calculate the correct channel scope rather than applying one delivery percentage indiscriminately. Capacity matters too: an item with a high dollar contribution can consume scarce preparation time during the peak. A measured bottleneck review can be more useful than putting the most expensive dish at the top of the menu.

A controlled menu experiment should change one useful factor at a time and record item sales, contribution, order mix and service outcomes. If a photograph, description and price all change at once, it becomes harder to explain the result. Keep the observation window and unusual conditions visible. Do not copy another restaurant's descriptions or images, invent customer reviews or claim that a platform's proprietary ranking formula is known.

Swiggy Menu Score guidance in the earlier section is platform-specific. It should not be represented as a DoorDash, Uber Eats, Square or Toast scoring feature. The general contribution-and-popularity method here is independent of those platforms. For supporting research, see Toast's original menu-engineering research; the four-dish case and its thresholds are independently invented.

For the next review, keep a small table with item quantities, consistent prices, recipe versions, included variable costs and the reason for each proposed experiment. A strong candidate is one whose evidence can be checked and whose operational tradeoff is understood. If your business needs export data from a DoorDash or Uber Eats dashboard to perform this analysis, use the international extractor enquiry link below. Those integrations are not currently available in our extension.

PRACTISE WITH EVIDENCE

Use the comparison without confusing a menu score with profit

Start with one accurately costed basket

The comparison above is an independent calculator, not Swiggy’s actual Menu Score tool and not a Zomato ranking score. Type the two item scenarios into their separate columns and click “Compare items.” Its contribution result subtracts the food, packaging, platform and promotion amounts entered. It does not retrieve prices or costs from a partner account.

Start with the supplied fictional pair. Item A contributes ₹100 per order across 100 orders; Item B contributes ₹160 across 60. Their total contributions are ₹10,000 and ₹9,600. Now reduce Item B’s packaging cost by ₹5 after a real packaging test: contribution becomes ₹165 per order and ₹9,900 in total, if sales and other costs stay unchanged. Verify leak resistance and delivered quality before accepting the cheaper pack.

Make one menu test measurable

For an item with incomplete description, record baseline completed orders, availability, complaints and contribution before changing the description. State the actual quantity, ingredients and included accompaniments. Review the next equivalent service periods, while noting any ad campaign or price change. Improved orders during a festival are not evidence that the wording alone caused growth.

For an add-on, cost the complete extra portion and packaging, then include the service charges caused by the extra revenue. A ₹40 add-on with ₹12 ingredients, ₹3 extra packing and an assumed ₹8 service cost contributes ₹17 before other costs. Those numbers are a teaching exercise, not a universal fee schedule. If the add-on delays a constrained station or creates frequent packing errors, the rupee calculation alone is insufficient.

Choose the right companion tool

Use the on-page comparison for two already-costed items. Use the full margin calculator for percentage-based fees, refunds and period overhead. Use the settlement analyser to review cash adjustments—not to rank dishes. The beginner tool manual explains those boundaries with arrows and examples. Preserve the recipe source and test notes so the next manager can repeat the comparison.

The complete beginner tool manual explains controls, units, privacy and access limits. Examples are invented and are not platform rates or restaurant results.

A menu workshop: improve the choice without losing the money or the promise

Your menu is both a customer conversation and a production plan. It tells a guest what they can buy, then tells the kitchen what it must deliver. A clearer menu can reduce confusion, but a beautiful catalogue cannot rescue an unavailable dish or a recipe that costs more than expected. The following workshop connects descriptions, portions, costs, popularity and preparation without pretending that one score measures everything.

We will use a fictional outlet, Orchard Table. It has twelve main dishes, four sides and several drinks. Its owner wants a smaller, easier menu and worries that removing an item will upset regulars. The team starts with evidence rather than a universal rule about the ideal number of dishes. All figures, names and events below are original teaching examples, not a real restaurant case or platform ranking study.

Start with a menu copy you can trace

Keep a dated private copy of the current item list, descriptions, prices, options and availability. Record where changes are made: a partner account, the restaurant's POS or an authorised menu provider. POS means the point-of-sale system used to record orders and payments. If an integration controls the menu, changing another screen can be overwritten later. Check the actual source of the listing before making competing edits.

Put the customer-facing item name beside an internal recipe reference. Two listings may use the same base recipe but different packaging or accompaniments. A large bowl and a small bowl should not share one cost merely because their names look similar. Likewise, a delivery listing with a side and drink is a basket, not a standalone dish. Your cost sheet needs to match what the customer actually receives.

Keep old versions long enough to explain a test under the business's record policy. If the price or portion changed during the measurement period, note when. A monthly total that combines two different versions may hide the effect you wanted to study. Do not alter the old recipe retrospectively to make an earlier week appear comparable; show the version change and choose a sensible comparison.

Read descriptions through a beginner's eyes

Ask someone unfamiliar with the menu to describe what they think each item includes. Can they identify the main ingredients, portion or serving size, included sides and required choices? Replace vague words with useful information. “Chef's special bowl” says little without an explanation. Do not invent appealing ingredients or quantities that are not reliably supplied. Plain, accurate words are better than elaborate promises the packer cannot fulfil.

Distinguish a choice from an addition. If rice is included and the guest chooses one type, say that. If extra rice costs more, make the addition clear through the permitted menu controls. An ambiguous description can create a complaint even when the kitchen packed exactly what its staff expected. Review descriptions alongside packing instructions so the customer promise and internal process use the same meaning.

Dietary and allergen statements need particular care. A dish described as vegan is not automatically free from every allergen, and removing one ingredient does not by itself prevent cross-contact. Follow the applicable requirements and the restaurant's trained safety process. Do not create a “free from” claim to improve menu appeal unless the operation can properly support it. This commercial workshop does not certify allergy safety.

Use photographs to explain, not disguise

A useful item image shows the actual product and portion honestly. Use original or properly licensed images that match what you sell. Keep lighting and framing consistent enough that customers can compare dishes. Avoid a photo that shows several sides when the listing includes none. A more attractive image can increase expectations as well as orders, so the kitchen and packing process need to match it.

If you test a new photo, keep the item price, description and availability unchanged where practical. Record when it went live and compare an appropriate period, including demand and complaints. A rise in sales alone does not prove the image caused it; other conditions may have changed. Do not claim a universal conversion uplift from one before-and-after test. Treat the result as evidence for your own next decision.

Do not copy a platform screenshot with customer details into a public article or support message. Our illustrated examples are teaching diagrams, not current authenticated interfaces. If you need to show a menu problem to official support, use the permitted private route and remove unrelated sensitive details. The aim is to make the issue understandable without disclosing information that does not help resolve it.

Cost the usable recipe, not just the supplier pack

List ingredient quantities in consistent kitchen units and connect them to the purchase unit. Yield means the useful amount after normal preparation. A kilogram bought does not always produce a kilogram served. Count the usable part and apply the cost consistently. If the recipe changes, update the quantity and version rather than leaving a historical cost in place because it makes the margin look better.

For a fictional ingredient, a 1,000-gram purchase costs ₹200 and yields 800 usable grams. Each usable gram costs ₹0.25. A 160-gram serving costs ₹40 for that component. If the dish also uses ₹18 of other ingredients and ₹12 of garnish and sauce, recipe cost is ₹70. At ₹280 revenue on the stated consistent basis, food cost is 25%. Packaging and other order costs remain outside that food-cost calculation.

Now imagine the serving quietly increases to 180 grams. That component costs ₹45 instead of ₹40 and the recipe becomes ₹75 if other amounts stay unchanged. Food cost at ₹280 is about 26.8%. A portion drift can therefore change the result without any supplier price rise. Measure normal preparation with the team rather than assuming the written recipe is exactly what happens at service.

Include the items that look too small to matter

Sauces, garnish, disposable cutlery and packaging can be missed because each individual amount seems small. A small unit cost becomes material across repeated sales. Use an evidence-based cost rather than rounding every minor item to zero. Keep the scope explicit: recipe ingredients belong in food cost, while packaging can be a separate variable line. This separation helps explain channel differences without losing the total cost.

Suppose a bag, container and sauce cup together cost ₹14 per delivery basket. Across 400 baskets that is ₹5,600. If the owner assumed ₹8, the missing ₹6 per basket understates the included cost by ₹2,400. This is not a platform deduction; it is a restaurant purchase and usage issue. Check the actual supplier pack, unit count and consumption instead of looking only at the settlement statement.

Some packaging is shared across a basket. Do not charge the full bag to every dish in the same order and then call the resulting item margins precise. Choose an allocation method for the intended decision and explain it. For a complete basket comparison, use the actual basket packaging once. For an item worksheet, a reasonable documented allocation can help, but remains an allocation rather than a directly observed per-item fact.

Understand markup, margin and contribution in ordinary words

Markup compares the extra selling amount with cost. Margin compares the amount remaining with sales. They have different denominators, so the same percentage produces different prices. Contribution subtracts the variable costs included in your stated scope and leaves an amount toward fixed expenses and profit. A food-cost-only margin should not be presented as full order contribution when delivery fees and packaging have been left out.

Take a ₹100 cost and ₹150 price. The extra ₹50 is 50% of cost, so markup is 50%. The same ₹50 is one third of ₹150 sales, so margin is about 33.3%. To achieve a 50% margin on that ₹100 cost alone, price would be ₹200 because half of the selling price remains after the cost. This is arithmetic, not a recommendation that customers will accept that price.

If the ₹200 basket also incurs ₹50 of other included variable costs, its contribution is ₹50, not ₹100. Rent and core payroll still need paying. Choose the price only after considering the full decision scope, customer demand, competitive alternatives, quality and legal price-display requirements. No formula can guarantee that a calculated price creates a viable market.

Compare dishes with three columns rather than one winner

Use units sold, contribution per unit and total contribution for the same period. Popularity tells you how often the dish was selected; contribution per unit tells you what one sale adds under your scope; total contribution combines both. A dish can rank differently in each column. The purpose is to find an operational or customer question, not automatically remove everything below an invented threshold.

Orchard Table sells 150 pasta dishes contributing ₹120 each, 100 rice bowls contributing ₹140 each and 30 salads contributing ₹180 each. Total contribution is ₹18,000, ₹14,000 and ₹5,400 respectively. Salad has the highest contribution per item but the smallest total. Pasta contributes most overall at the observed demand. These figures do not include fixed expenses, and they do not prove one item deserves the most prominent position.

A salad may serve an important customer need, use spare preparation capacity or support a group order. Pasta may consume the busiest station. Add those facts to the decision. If a low-selling item requires unique ingredients that repeatedly spoil, investigate its availability, demand and role. Avoid an automatic rule that every low seller is bad or that every popular item is profitable.

Check whether availability explains low popularity

An item cannot be fairly judged against another if it was unavailable for much of the period. Record listing hours or services where you can verify them. If a dish was sold out every Friday evening, its low weekly units may reflect supply rather than low demand. Likewise, a recently added item has had less opportunity to be selected. Put that limitation next to the comparison instead of treating it as an afterthought.

In a fictional week, a salad sells 25 units while available for two lunches. Pasta sells 80 across ten services. The total ranking alone does not establish relative customer preference. You might compare suitable available periods and observe whether the salad can be produced reliably. Do not simply multiply the two-lunch sales to forecast a full week; demand varies by day and service.

Review the reason for unavailability. Ingredient supply, preparation capacity and poor forecasting require different actions. A description edit will not fix a recurring supplier shortage. A larger batch may create waste if demand is uncertain. Test a manageable replenishment or availability process, keep safe production constraints and record whether the item actually becomes available when customers want it.

Build choices the kitchen can follow

Every required modifier adds a decision for the guest and a task for the team. Modifier means a selectable change, such as size, sauce or side. Use choices that serve a real customer need, with clear defaults where permitted. Too many options can make a simple dish difficult to understand and increase packing errors. Before adding a choice, check the recipe, price, availability and way the order will be displayed internally.

Orchard Table discovers that three sauces share confusing names on its order tickets. Staff sometimes read the default sauce instead of the chosen one. The fix may be a clearer label and packing verification, not removing every option. Try a small internal rehearsal for different baskets and have the packer explain each order. Do not rely on memory or an unofficial workaround that loses the original request.

Cost the choice honestly. An extra protein portion with a higher purchase cost should not inherit the base recipe's contribution. Keep included choices and paid extras distinct. If a dietary request affects food safety, follow the trained process and applicable rules rather than letting a commercial worksheet decide what is safe. Where the platform's controls cannot represent a promise clearly, reconsider the offer before publishing it.

Test a combo as a basket, not an assumed bargain

A combo groups items into one offer. Calculate its revenue, ingredients, packaging and relevant order costs together. The customer may buy the combo instead of buying the components separately, so extra combo sales are not automatically extra demand. Compare average basket composition and total contribution alongside the apparent popularity. Keep the advertised contents precise so the packer does not need to guess which drink or side is included.

Suppose a main dish has ₹300 revenue and ₹170 included variable cost, contributing ₹130. A drink alone adds ₹60 revenue and ₹20 cost. A combo priced at ₹340 with total included cost ₹190 contributes ₹150. It contributes ₹20 more than the main alone but ₹20 less than the ₹360 separately priced pair, which would contribute ₹170. Whether the combo helps depends on what customers would otherwise have bought.

If the combo makes packing slower or requires another container, add the relevant cost to the comparison. If it increases the use of a bottleneck ingredient, note that constraint. A lower apparent food-cost percentage can hide a weak contribution after funded discounts. Test a defined offer for a bounded period and review substitution, service quality and cost before keeping it permanently.

Consider scarce kitchen time without inventing an exact answer

When one station is full, time becomes part of the choice. Observe the actual work needed at that station, not just the total guest wait. Some dishes cook together or share a preparation batch, so time per item is not always additive. Use a measured, clearly simplified model to identify questions. Do not pretend a contribution-per-minute ranking is a complete scheduling rule for a working kitchen.

In a fictional peak, Dish A contributes ₹150 and needs five measured bottleneck minutes, or ₹30 per minute. Dish B contributes ₹180 and needs nine, or ₹20 per minute. Dish B contributes more per item, but A provides more contribution per bottleneck minute under these assumptions. Customer demand, batching, commitments and quality still matter. If the station has spare capacity, the same ranking may be less important.

Before changing availability, investigate avoidable delay: missing mise en place, confusing tickets or a handover block. Mise en place simply means ingredients and equipment prepared for service. Preparation must follow safe procedures. A cost sheet never justifies unsafe advance preparation or holding. Coordinate operational changes with trained staff and use the current authorised platform controls when changing availability or preparation settings.

Plan a price test without misleading guests

A price change is not just a spreadsheet edit. Review applicable display rules, platform terms and the guest's understanding of the portion and contents. Record the old and new price, the effective date and the reason for the test. Do not disguise a smaller portion as the same offer. If the product changes, update the description and cost version so the comparison reflects the actual promise.

Suppose an item rises from ₹300 to ₹320 while its included variable cost remains ₹200. Contribution rises from ₹100 to ₹120 per unit. At 100 sales the old contribution is ₹10,000. To match that with the new contribution, the item needs about 83.34 units, so at least 84 whole sales in this simple model. The model assumes unchanged costs and no effect on other basket items; real demand must still be observed.

A drop from 100 to 90 sales would produce ₹10,800 contribution, but could also affect drinks, side sales or repeat demand outside the narrow worksheet. Do not call it a complete success based on the single-item total. Check complaints, basket mix and comparable conditions. A one-week result does not establish that the same price will work through every season or for every customer group.

Write a menu action log that survives the busy shift

For each proposed change, state the problem, evidence, action, owner, date and measure. “Improve salad” is too vague. “Clarify the included side in the description; compare missing-side complaints and item sales over comparable available services” is usable. Keep the actual menu version and operational notes together. If several people can edit, agree who approves and verifies publication through the authorised source.

Put a check after publication into the workflow. Does the live listing show the intended text, price and options? Does the internal order view represent the same choices? If an integrated menu overwrote the edit, resolve the source-of-truth issue rather than repeatedly changing competing screens. Never bypass a platform's access controls or share an account token to make a third-party tool perform menu changes.

Our extension exports selected supported Zomato and Swiggy records; it does not automatically publish menu edits or change prices. The calculators help explore entered values but do not know your complete agreement or guarantee the outcome of a test. Owners using DoorDash or Uber Eats can ask about a future export requirement through support, but those integrations are not currently available in our extension.

Handle shared ingredients without pretending allocation is exact

A common sauce batch may serve several dishes. First calculate the genuine batch cost and usable output, then use the measured recipe amount for each dish. Do not assign the whole batch to whichever item was sold most recently. If part of the batch was discarded, keep actual usage and the reason for waste in the weekly review. The standard recipe cost and actual weekly consumption answer related but different questions.

For an original example, a sauce batch costs ₹600 and yields 3,000 usable grams. Its cost is ₹0.20 per gram. An item using 50 grams carries ₹10 of sauce in its standard recipe; another using 80 grams carries ₹16. If 500 grams are discarded, the batch's recorded waste is ₹100 at this stated cost. It does not mean every remaining serving's standard quantity changed. Explain the discarded amount once in actual usage.

If you cannot measure the usable output reliably, label the recipe estimate and improve the next batch record. A decimal-filled sheet can create false confidence when the underlying measurement is only a guess. Choose useful precision and record the assumptions. The same care applies to shared packaging, bulk condiments and discounts that belong to a whole basket rather than a single item.

Keep tax and pass-through money out of a misleading comparison

Two dishes with identical displayed prices can have different accounting treatment, depending on the jurisdiction, item and arrangement. Keep the price basis consistent with the costs. Recoverable purchase tax may not be treated like a final ingredient expense; non-recoverable tax may affect cost. A customer's total can also include amounts that are not restaurant revenue. Ask the accountant to establish the correct basis rather than subtracting a generic rate because a calculator accepts it.

For a simple fictional worksheet, label revenue “excluding pass-through tax” and use costs on the corresponding agreed basis. This is a teaching definition, not tax advice for your business. If you later use a tax-inclusive customer price, explain the conversion supported by the actual rules before comparing it with that worksheet. Do not mix one gross price and one net price and declare the difference a menu improvement.

The same principle applies to staff tips and other collected amounts that the restaurant does not retain as its own sales. Their treatment depends on the arrangement and applicable rules. This guide's contribution examples exclude such pass-through amounts. If the source report combines them with sales, identify the separate lines before using the total. A larger cash transfer can otherwise make an item appear more profitable without any improvement in the dish itself.

Know when not to change the menu

If the records are incomplete, a supplier delivery changed, or the platform listing was offline for half the test, gather better evidence before making a sweeping decision. The cost of waiting briefly for a clear answer can be lower than printing a new menu, retraining staff and upsetting guests based on an unreliable week. Keep urgent safety and accuracy corrections separate; those should not wait for a commercial experiment.

A menu review should leave the customer with clearer choices and the team with a manageable promise. It does not need constant redesign. Keep a well-supported item stable while investigating a specific weak point elsewhere. Record why no change was made when that is the right decision. Stability can be an informed choice, not a failure to optimise.

Reader exercise: choose the next useful question

Imagine a dish with high sales, low contribution and frequent preparation delays. Which action comes first: more advertising, removal, or an investigation of its recipe, price and station workload? The useful first move is to identify the repeated cause with the strongest evidence. Advertising could increase the problem. Removal could lose valued demand. A targeted preparation or cost test may help, but needs a real observation and a clear stop condition.

Now imagine a profitable-looking dish with few sales and a unique ingredient that repeatedly spoils. Recheck the contribution with actual usage and waste rather than only ideal recipe cost. Ask whether the description, availability or customer need explains demand. A planned smaller batch may be appropriate if safe, but it is not automatically the answer. Retain the facts that would support keeping, changing or removing it.

Finally, explain your conclusion to someone new to the business without using a platform score as the entire reason. Can you state the customer promise, included cost, observed demand, operational tradeoff and next test? If so, you have a practical menu decision rather than a ranking with no clear action. Use the relevant examples above when one of those parts is missing.

All added menus, costs, thresholds and experiments are fictional exercises. Menu-score features remain platform-specific; accounting comparisons are independent of proprietary rankings. Recipe and food-cost detail · Contribution versus final profit · Weekly operational review.