What Is Menu ABC Analysis and How Does Menu Engineering Work
Contents
- What menu ABC analysis is and which metric to choose
- What sales data and recipe cards you need for menu engineering
- How to build a popularity and contribution matrix
- What to do with stars, plowhorses, puzzles, and dogs
- What the calculation looks like with a fictional example
- What mistakes distort ABC analysis and the menu matrix
- Where to run menu analysis and how to track changes
- Frequently asked questions
What is menu ABC analysis? It is ranking dishes by a selected metric and dividing them into groups A, B, and C. Menu engineering helps you make assortment decisions by comparing demand with item unit economics. To do this, after ranking, you build a separate popularity and contribution matrix based on contribution after food cost.
Suppose your manager suggests removing a rarely ordered dish and boosting sales of the most popular one. Before deciding, check how much each item leaves after ingredient costs, what it is ordered with, and when it is available. Below, we break down the calculation, the four matrix groups, and the process for adjusting your menu without hasty conclusions.
In brief:
- For ABC, choose a metric: revenue, portion count, or total contribution. The results will differ.
- For the matrix, separately compare popularity and contribution per portion within a comparable group.
- Link sales data with current recipe cards and maintain consistent accounting rules for discounts and returns.
- Stars, plowhorses, puzzles, and dogs require different checks and actions.
- First verify the explanation, then adjust the menu, and review the results over a comparable period.
What menu ABC analysis is and which metric to choose
Start with a management question. To evaluate revenue structure, rank dishes by revenue. To see kitchen workload, look at portion counts alongside preparation complexity. For product economics, take total contribution after ingredient costs. Label the chosen metric directly above the table.
Sort items in descending order of value. Calculate each item's share of the total, then the cumulative share from top to bottom. Agree on the boundaries for groups A, B, and C. Note where an item belongs if it crosses a boundary. Choose the classification threshold as a standard rule for your menu analysis and keep it to compare periods.
Group A highlights the items that form the chosen core share of your results. Groups B and C continue the ranking. The letter category depends on the metric: a dish can rank high in revenue and much lower in contribution. Check both columns before discussing the menu mix.
ABC helps prioritize your review. The menu matrix answers the next question: how order frequency combines with the contribution of an individual portion. Therefore, keep your original ranking first, then add the two matrix axes. The overall framework for working with these numbers is covered in our restaurant analytics guide.
Isolate zero and negative contributions separately. If such items are included in the calculation, verify the meaning of cumulative shares: negative amounts alter their progression and disrupt standard group interpretation. Find the root cause first, including returns and costing errors. Retain the original values and clearly describe the rule for handling these rows in your report.
What sales data and recipe cards you need for menu engineering
Take a closed period, a single location, and a comparable group of dishes. Analyze hot dishes, desserts, and beverages separately if guests choose them in different situations. For each item, you need the following data.
- Item code and category for the calculation.
- Number of portions sold and revenue after agreed adjustments.
- Discounts, returns, sales channel, and availability dates.
- Current recipe card, portion yield, and ingredient costing.
- Modifiers, portion size options, and combo meal contents, if present in the data.
Verify several lines manually. Compare dish sales with the cash register report, recipes with the kitchen, and food costing with accounting records. For iiko, r_keeper, or 1C, verify available fields and history in your setup. Pull a sample first, then decide which part of the calculation you can update regularly.
If a recipe card changed during the period, split sales by the effective dates of each recipe. Check prep batches, processing yield losses, and units of measurement. An outdated recipe can shift a dish's position in the matrix. Formulas and usage verification are covered in detail in our article on how to calculate food cost.
A Kaspi payment on its own does not reveal order contents. To assess popularity, you need POS sales line items. When combining locations in Almaty and Astana, first match codes, recipes, and portion sizes. Keep different dishes with the same name as separate entries until you resolve the differences.
Compare your sales export against the full list of available menu items. An item with zero orders may be missing from receipt line items even though it stayed on the menu the entire period. Add it with zero sales and confirmed availability. If an item lacks a recipe card, display the gap explicitly: silently dropping the row alters the composition of the analyzed group and masks data preparation issues.
How to build a popularity and contribution matrix
In a simple calculation, popularity reflects a dish's share of portions within the selected group. For an identical period, you can also compare portion counts directly. Document which sales are included in the denominator, how you account for returns, and how many days the dish was available.
Contribution after food cost reflects the remaining revenue after paying for the ingredients of sold portions. This is an intermediate metric. For the restaurant's bottom line, you must also factor in labor, rent, and other operating expenses. A high food contribution still requires verifying overall economics.
Total contribution = revenue of selected sales minus their food cost.
Contribution per portion = total contribution / number of corresponding portions.
Use contribution per portion in tenge for the vertical axis. Use popularity for the horizontal axis. If you place metrics that heavily depend on sales volume on both axes, you will partially duplicate the same measurement. Keep total contribution nearby as an indicator of overall scale.
Agree on the boundaries for high and low values within the category. You can start with the average portion count per item and the chosen average contribution, then test whether this split makes sense. Specify the averaging method and the rule for values equal to the threshold. If you adjust thresholds, recalculate the comparison period.
| Matrix group | Popularity | Contribution per portion | First question |
|---|---|---|---|
| Stars | High | High | How do you maintain availability and quality? |
| Plowhorses | High | Low | What limits contribution when demand is steady? |
| Puzzles | Low | High | Why do guests rarely choose this item? |
| Dogs | Low | Low | What role does this dish play on the menu? |
What to do with stars, plowhorses, puzzles, and dogs
Stars. Check ingredient availability, cooking consistency, and availability during peak hours. Keep the description clear. Before active promotion, evaluate kitchen workload and your capacity to fulfill orders. Record reasons when an item is unavailable: lost demand can stay outside the sales report.
Plowhorses. Review purchase costs, yield, discounts, and the use of expensive add-ons. Check with the kitchen whether recipe cards are being followed. Consider adjusting the offer or pairing it with other items. If you change portion sizes, first evaluate quality and guest expectations, then check sales and contribution.
Puzzles. Review the name, description, menu placement, and server explanations. Check whether the dish was available throughout the entire period. Choose a single change to test, such as clarifying the description, and keep all other conditions the same. After a follow-up period, compare volume, contribution, and the impact on neighboring items.
Dogs. Identify the role of the item: a specific guest need, an add-on to an order, or the use of a shared ingredient. Check related write-offs and labor intensity. If the role is not confirmed, discuss changing the recipe, replacing the item, or removing it. Before deciding, see what guests can choose instead of this dish.
Assign an owner, a start date, and a review date for each action. Keep the original matrix. If you change the description, recipe, and sales conditions all at once, explaining the result will be much harder. In meetings, discuss verified causes and specific team actions.
Before testing, record the expected change and related metrics. For example, when clarifying a puzzle's description, also look at sales of its closest alternative. Growth in one item may be accompanied by a shift in demand within the group. After testing, compare the category's total contribution, availability, and kitchen workload. Based on the results, decide whether to keep the change, refine it, or revert to the previous version. Save the explanation alongside the decision so that the next review starts from verified conditions.
What the calculation looks like using a fictional example
Suppose a hypothetical cafe analyzes four hot dishes over a single closed period. All names, quantities, and amounts below are fictional and rounded for explanation purposes. This is a training scenario. The table shows aggregated sales and food costs; dish prices are not listed here.
| Dish | Portions | Revenue, tenge | Food cost, tenge | Contribution, tenge | Contribution per portion, tenge |
|---|---|---|---|---|---|
| Pasta | 400 | 360 000 | 160 000 | 200 000 | 500 |
| Cutlet | 300 | 180 000 | 120 000 | 60 000 | 200 |
| Fish | 200 | 200 000 | 80 000 | 120 000 | 600 |
| Vegetables | 100 | 60 000 | 40 000 | 20 000 | 200 |
| Total | 1 000 | 800 000 | 400 000 | 400 000 | 400 |
First, let's run an ABC analysis by total contribution. For this training calculation, we set group A up to a cumulative 80%, B up to 95%, and C for the remainder. Pasta accounts for 50% of the total contribution. Fish adds 30%, making 80% combined. These two items belong to group A. Cutlet adds 15% and belongs to group B. Vegetables provide the remaining 5% and belong to group C.
Now let's build a separate matrix. The high popularity threshold in this example is 250 portions: 1 000 / 4. The high contribution threshold is 400 tenge per portion: 400 000 / 1 000. Here we use the average contribution of a sold portion, volume-weighted. Values equal to the threshold are treated as high.
We get four different outcomes. Pasta becomes a star: 400 portions and a contribution of 500 tenge. Cutlet lands in plowhorses: 300 portions with a contribution of 200. Fish turns out to be a puzzle: 200 portions with a contribution of 600. Vegetables fall into dogs: 100 portions with a contribution of 200.
Fish belongs to group A and remains a puzzle. It delivers a noticeable share of total contribution, but within the selected group it is rarely ordered. Check availability and the description before changing your product mix. For pasta, check consistency; for cutlet, food costs; for vegetables, their role in orders. Further decisions require reviewing the cafe's specific conditions.
What mistakes distort ABC analysis and the menu matrix
One dish in multiple categories. An item can appear in different sections of the menu. For the calculation, assign it a single primary category or explicitly define an allocation rule. Make sure portions and revenue are included in the grand total only once. Save marketing collections as separate attributes.
Seasonality and out-of-stock periods. Compare like-for-like periods. For seasonal dishes, check the actual days they were available. Short-term sales cannot be directly compared with the full period of other items. If historical data is insufficient, note the limitation and postpone a final decision.
Combos. Choose the unit of analysis: the combo as a whole or its individual components. When analyzing components, allocate revenue according to an agreed rule and account for the food costs of each dish. Keep the original set for reconciliation. Double-counting combos and individual components will inflate quantities and results.
Modifiers. Extra ingredients, substitutions, and portion sizes change the calculation. Link the add-on to the original order. Verify both revenue and food cost. If there is not enough data, label the averaging and show which items require separate verification.
Mixing sales channels. For delivery, evaluate packaging and contractual deductions separately. Delivery aggregator commissions depend on the contract. After the product matrix, you can build a calculation with agreed channel expenses. Clearly label the updated contribution definition so that you compare identical cost structures.
Where to manage menu analysis and how to verify changes
Start with a spreadsheet or an available POS report. Lock down formulas, the dish catalog, and the update date. Verify total revenue and portions. To regularly combine POS, accounting, and multiple locations, consider automation. Criteria are available in our comparison of Power BI and POS analytics.
Yashcher implements Power BI. If the calculation being verified is already available to the team and updates on time, use your current tool. For a chain, first compare identical categories within locations, then consolidate the results. Specifics of this analysis are described in our article on coffee chain analytics.
Frequently asked questions
How does menu ABC analysis differ from menu engineering?
ABC ranks items by a single selected metric. Menu engineering connects assortment analysis with decisions. The popularity and contribution matrix serves as a separate step: it shows how demand combines with portion product economics.
Can you run the analysis based on revenue alone?
You can use revenue to build an ABC analysis and see your sales structure. For a product matrix, you also need portion counts and food cost. Without recipe cards or another verified ingredient valuation, note this limitation and restore this data first.
Do you need to remove all dishes from the dogs group?
First, check availability, seasonality, and role in orders. Discuss alternatives with the kitchen and the manager. Record the reason for changing the menu mix and verify the result. The group name alone does not replace a decision.
Why can't you evaluate a dish by food cost alone?
Food cost shows the ingredient share. To make a decision, you also need contribution in tenge, sales volume, and prep requirements. Compare the percentage with the monetary amount, then check other expenses and the item's role on the menu.
How often should you recalculate the menu matrix?
Choose a period that accumulates enough comparable sales. Repeat the calculation after agreed data reconciliation and significant menu changes. Keep previous thresholds or explain their revision separately so your team understands the reasons for shifts between groups.
