How to Use the Menu Engineering Analyzer
For a 30-item menu, the first analysis takes about 30 minutes. Most of that is gathering your data, not using the tool. Here's exactly how to run the Menu Engineering Analyzer, from the numbers you need to the actions worth taking first.
What you need before you start:
- Sales count per menu item (past 30 days)
- Selling price per item
- Food cost per portion (estimates work fine)
What you'll get:
- Every menu item mapped to Stars, Plowhorses, Puzzles, or Dogs
- Contribution margin per item and section averages
- Recommended actions for each category
New to the framework? The Analyzer walks you through the math. If you want to understand the logic behind the four categories before you start, Menu Engineering Basics for Independent Restaurants covers the Stars, Plowhorses, Puzzles, Dogs framework and how to read each one. Not required reading before opening the tool, but useful context for your first analysis.
What data do you need before you open the tool?
Three numbers per item. Nothing more.
Sales count is how many times each item sold in the past 30 days. Your POS product mix report is the fastest source if you have one. No clean report? Count from order tickets, estimate from a typical week and multiply out, or use your best recall. Close enough is useful. Exact is not required to get a meaningful result.
Selling price is what you charge for the dish. Simple.
Food cost per portion is what it costs to make. Recipe cost cards are ideal. If you don't have them, estimate. A pasta dish that runs roughly $5 in ingredients is close enough. Don't let this step stop you from starting — the Analyzer works with good-faith estimates.
One prep note before you open the tool: organize your items by section. The Analyzer compares items within sections, not across them. Appetizers versus appetizers, mains versus mains. A $10 starter shouldn't be plotted against a $38 main — the relative performance comparisons only mean something within comparable groups.
Better guest experience. Bigger nights. $299. Once.
How do you enter your menu data?
Open the Menu Engineering Analyzer.
Create your first section and name it (Appetizers, Mains, Desserts — whatever you call it). Add items one at a time: item name, selling price, food cost per portion, and sales count for the past month. Work through each item in the section, then move to the next section.
You can analyze your full menu or focus on one section where you suspect a problem. A complete analysis of your mains is more useful than a thin analysis spread across everything.
When your items are entered, click Analyze.
How do you read your results?
The results page places every item on a two-axis matrix. Contribution margin (selling price minus food cost) on one axis, popularity (sales count) on the other. Each item lands in one of four quadrants based on how it compares to your section averages.
| Category | Margin | Popularity | What it signals |
|---|---|---|---|
| Stars | High | High | Best earners. Protect their placement. Don't experiment on them. |
| Plowhorses | Low | High | Volume movers with thin returns. Price up modestly, cut food cost, or bundle with a higher-margin item. |
| Puzzles | High | Low | Good margin, but guests aren't ordering them. A visibility or description problem before it's a price problem. |
| Dogs | Low | Low | Low volume, low margin. The honest question: does this item justify its place? |
The Analyzer also shows each item's contribution margin in dollars, your section averages, and specific recommended actions per category.
On Puzzles specifically: The instinct when an item isn't selling is to lower the price. Resist it. A Puzzle earns well on every plate it does sell. The problem is usually where it sits on the menu, how the description reads, or whether servers are mentioning it. A server recommendation alone can move a Puzzle toward Star territory. Try the free fixes before you cut into a high-margin item.
What should you act on first?
Don't try to move on everything at once. That's how this kind of analysis ends up filed and forgotten.
The highest-impact first move for most independents: find your top two Plowhorses by sales volume and calculate what a $1.50 price increase does to monthly contribution margin. If those two items sell 180 times a month combined, that's $270 in additional margin on a change most guests won't notice. Run the math on your actual numbers before deciding.
After that, look at each Puzzle and ask three things: Is it in a visible spot on the menu? Does the description communicate the dish clearly? Do your servers know to recommend it? These fixes cost nothing.
Dogs can wait for your next menu revision. Make a note in the Analyzer, close the tab, and come back when you're ready to rotate the menu.
How long does the first analysis take?
Data gathering takes the most time. Expect 20 to 30 minutes to pull and organize your numbers — longer if you're counting manually or estimating food costs for the first time. Entering items into the tool takes 5 to 10 minutes for a typical 30-item menu. The analysis is instant.
The second time you run it: about 15 minutes total. Your cost estimates exist, you know where your sales numbers are, and you're updating rather than starting from scratch.
Run it quarterly. Monthly if your menu is seasonal or if ingredient costs shift frequently. Set the next reminder before you close this tab.
Menu profitability and food waste are two sides of the same equation. A poorly engineered menu that keeps high-cost Plowhorses without scrutiny generates waste cost on top of the margin problem. The hidden costs affecting most Canadian independents covers how these connect.
Source: Menu Engineering Analyzer, Trudy's Table.
Frequently Asked Questions
What three numbers do I need for each menu item?
Selling price, food cost per portion, and sales count for the past 30 days. That's the full input. The Analyzer calculates contribution margin and plots each item against your section averages automatically.
What if I don't know my exact food cost for every item?
Estimates work. If you know roughly what a dish costs to make — ingredients, portion size — enter that number. The matrix is built on relative comparisons within your own menu, not absolute precision. A good-faith estimate is enough to surface your Stars and Dogs.
Why do I need to keep sections separate?
Because the Analyzer compares items against section averages, not against a universal benchmark. A $10 appetizer with a $4 food cost shouldn't be measured against a $36 main with a $14 cost — the margin and volume benchmarks mean different things across sections. Mains versus mains, appetizers versus appetizers.
Can I run the analysis on just one section instead of my full menu?
Yes. A focused analysis of your mains is more useful than a thin analysis spread across everything. Start with the section where you suspect a profitability problem and expand from there.
How often should I re-run the analysis?
Quarterly for most operations. Monthly if your menu is seasonal or if ingredient costs shift quickly. Once your items are in the tool, a quarterly update takes about 15 minutes.