← All insightsPILLAR GUIDE

Menu engineering for hotels and restaurants: margin, popularity and pricing

A menu-engineering guide to classifying items and deciding pricing, placement, promotion and range.

Menu engineering combines sales and profitability data to identify which dishes to protect, correct, promote or replace. Food-cost percentage alone can mislead: an item with a higher percentage may contribute more cash margin than an apparently efficient one.

Analyse a representative period by outlet with updated recipes. Star, Workhorse, Puzzle and Dog categories support decisions; they are not permanent labels for the menu.

To obtain a useful reading, data must be consistent with the period, service model and operating responsibilities. In a hotel, the same variance can come from occupancy, guest mix, board basis, outlet opening, purchase prices or recording quality. HPL therefore avoids automatic conclusions: it first reconstructs scope, then verifies source reliability and finally connects results to activities the team can genuinely control.

Indicators to monitor

The principal indicators are Contribution margin, Popularity, Food-cost percentage, Sales mix. They should not be read separately: a positive value may hide a weakness elsewhere, while a variance may be normal when explained by service conditions. The dashboard should move from the total to category, department and day without losing a shared formula definition.

  • Contribution marginNet selling price minus recipe cost for one portion.
  • PopularityItem sales compared with the average or category threshold.
  • Food-cost percentageRecipe cost divided by net selling price.
  • Sales mixEach item’s share of its category or total menu sales.

A five-step operating method

The recommended route includes Update recipes, Clean sales data, Build the matrix, Define action by item, Measure response. Sequence matters: acting before the baseline is reliable makes improvement difficult to demonstrate. Each stage should produce a simple output, an owner and a review date.

  1. Update recipesQuantities, yields, waste and prices must reflect real service.
  2. Clean sales dataCombine duplicate codes, variants and complimentary items consistently.
  3. Build the matrixCompare margin and popularity within homogeneous categories.
  4. Define action by itemPrice, portion, description, position, training or replacement.
  5. Measure responseReview sales, margin and satisfaction after the change.

Implementation does not necessarily require a new platform. A well-configured system can accelerate the work, but manual records and structured spreadsheets are enough for an initial diagnostic. Tool selection comes after defining the process, units of measure and people responsible for producing and using the data.

Common errors to avoid

  • Using theoretical recipes that the kitchen does not follow.Correcting this point makes the data more comparable and the action more credible.
  • Comparing starters, wine and main courses in one matrix.Correcting this point makes the data more comparable and the action more credible.
  • Automatically raising the price of every popular item.Correcting this point makes the data more comparable and the action more credible.
  • Removing a weak item without considering menu identity and function.Correcting this point makes the data more comparable and the action more credible.

The priority is not achieving a perfect number immediately, but building a measure stable enough to guide a decision. The first cycle can focus on products, departments or services carrying most value. After correction, repeat measurement on the same scope and verify that performance improved without harming quality, safety, availability or guest satisfaction.

Frequently asked questions

What data is required?

Net price, updated recipe cost and quantity sold over the same period.

What are Stars?

Items with popularity and margin above the threshold for their category.

How often should the analysis be repeated?

Whenever menus, prices or costs change and at a frequency appropriate to seasonality and volume.

From diagnosis to a control routine

The most delicate phase comes after the first analysis. Management must turn evidence into a routine understood by the general manager, F&B, kitchen, service, stores, purchasing and finance. Start with a short meeting that shares the formula, clarifies the source and assigns an owner to each action. Anomalies should not simply be corrected in the system: trace them back to the operating step that created them, otherwise the data will quickly drift again.

A sound cadence combines frequent operating checks with a more concise management review. The first corrects behaviours and records; the second evaluates financial impact, priorities and resources. Initially, show both absolute value and a normalised indicator because percentage alone can hide economic weight while totals may be driven by volume. Before-and-after comparison must use the same scope and disclose changes in occupancy, menus, prices or service conditions.

Once the process is stable, introduce traffic-light thresholds, benchmarks between comparable properties and progressive targets. Green confirms the routine is working, amber requires investigation and red requires an owner and deadline. This approach avoids generic alarms and makes improvement verifiable over time.

To keep control credible, every indicator needs a short definition sheet: formula, data source, frequency, owner and attention threshold. Exceptions must also be documented. Groups, events, partial closures, supplier changes or major menu variations may explain the result and should accompany the analysis. After two or three cycles, the team can remove redundant checks and focus on the levers that genuinely anticipate a problem. Discipline is not about producing more reports; it is about making the decisions that protect margin, quality and service continuity timely, consistent and verifiable.

Documenting decisions also makes the method transferable between departments, seasons and properties without losing accountability or operational context.