Cocktail Menu Engineering: Read the Numbers Carefully

Define the decision before opening the spreadsheet

Menu engineering is most useful when it answers a specific question: which drinks should stay, change, receive clearer descriptions or leave the list? A large report without a decision can create work without improving service. Choose a review period that reflects ordinary trading and note unusual events, promotions or closures before comparing individual items.

Begin with a stable recipe and a consistent costing method. If every drink uses a different definition of cost, sorting by margin produces a tidy but unreliable ranking. The cited Diageo guidance combines popularity and profitability as a starting point. The framework below adds practical questions that keep those numbers connected to the actual guest experience.

Build a small, reliable set of measures

For each drink, collect units sold, net revenue on a consistent basis, ingredient cost per serving and days genuinely available. Add a short operational note covering preparation complexity, stockouts, remakes and guest questions. You do not need a sprawling dashboard to discover that a frequently unavailable item has a misleadingly low sales count.

Decide how discounts, complimentary drinks and returned orders enter the report. A point-of-sale count may not equal paid portions, and a promotional price may change the contribution. Ask whoever manages the accounts to confirm the definitions. Keep the raw data separate from editorial judgments so a disagreement about a drink does not silently change the numbers.

When comparing two periods, check whether opening hours, covers or menu visibility changed. A lower count can reflect fewer opportunities to order rather than a decline in appeal. Simple context often explains more than an elaborate chart.

An illustrative two-drink comparison

Suppose Drink A sells 120 paid servings at a net price of 12, with ingredient cost of 3 per serve. It produces 1,080 currency units after ingredient cost. Drink B sells 50 at a net price of 16, with ingredient cost of 5, producing 550 after ingredient cost. These are invented figures for explanation, not actual venue results.

Drink B contributes more per serving, while Drink A contributes more in total during the example period. Neither observation alone tells you to remove the other. They may serve different preferences, occupy different occasions or require different amounts of work. The figures are also not net profit because operating costs outside the ingredient calculation remain.

Now imagine Drink B was unavailable for half the period. Its low count needs a supply investigation before a popularity judgment. Or imagine Drink A repeatedly blocks the only blender during peak service. Its impressive total should prompt a workflow review, not a declaration that all other drinks should imitate it.

Use four questions rather than four permanent labels

A popular drink with a useful contribution may deserve protection: maintain its quality and keep its ingredients available. A popular drink with weak contribution may need a purchasing, yield or portion review before a price change. Do not quietly reduce the promised serve to improve a spreadsheet.

A less ordered drink with useful contribution may need a clearer description, better staff familiarity or simply more relevant exposure. A less ordered drink with weak contribution may be a candidate for redesign or removal. In both cases, check whether it fills an important gap, such as a particular flavor preference or a considered alcohol-free choice.

These are questions to investigate, not permanent categories attached to a drink's identity. An item can move as prices, demand and service change. Avoid letting a nickname in a spreadsheet become a substitute for understanding why people do or do not order it.

Change one thing and state what success means

If you rewrite a description, keep the recipe and price stable during the comparison where practical. If you change the recipe, record the version and retrain the team. Several simultaneous changes make it difficult to know what caused the result. In a live venue you cannot control every variable, but you can keep a clear record of the ones you chose.

Before the change, state the desired outcome in observable terms. Perhaps guests should understand the flavor more readily, the preparation should require fewer handoffs or a perishable component should generate less waste. Sales may be part of the evaluation, but they should not erase quality, accessibility or staff workload.

Allow enough relevant exposure to draw a useful conclusion. One unusual evening is weak evidence for a permanent decision. Equally, do not keep an unresolved operational problem indefinitely because a favorite concept has emotional appeal to its creator.

Protect the menu's overall usefulness

A list is a set of choices, not merely a collection of individually profitable items. Removing every less ordered option can leave several near-identical drinks and no answer for a guest with a different preference. Review the range of flavors, methods, price points and alcohol-free options after any proposed changes.

Keep descriptions truthful and readable. The menu-design reference supports consistency with the venue's character; that does not require exaggerated claims or visual tricks. A clear explanation of the drink can be more helpful than repeatedly highlighting the item the spreadsheet prefers. Staff should be able to recommend honestly based on what the guest wants.

Finish with an action log

A review is complete when each decision has an owner and a way to check its effect. Limit the number of changes to what the team can implement carefully.

Menu engineering should make the bar easier to run and the list easier to choose from. Treat numbers as evidence, retain their context and use the review to make a small number of well-supported decisions.

• Keep: protect specification, availability and training.
• Investigate: name the missing evidence and who will obtain it.
• Trial: record the precise change and comparison period.
• Replace: confirm the menu gap the new drink must fill.
• Review: set the next decision point using a relevant amount of service data.

Editorial method and review

Prepared with AI assistance by CocktailsPro Editorial using the primary references below. The examples and checklists are original editorial proposals, not reported hands-on tests. This draft needs human editorial review and, where applicable, practical tasting before publication. Research checked on 12 October 2026.