Menu Engineering: A Data-Driven Guide to Knowing Which Menu Items to Keep, Rework, or Cut

Why Menu Engineering Matters for Small Cafes

Menu engineering sounds fancy — like something only big restaurant chains do with a team of analysts. But really, it's just a systematic way to answer a simple question: which menu items are making you money, and which ones are just consuming resources?

Most cafe owners make menu decisions based on gut feeling. "This seems trendy." "Customers like it." "It'd be a shame to remove it — it's been on the menu forever." Gut feeling isn't wrong — but without data, you don't know whether the menu item that "customers like" is actually generating profit or quietly eroding your margins.

With sales data from your POS, you can analyze every item on your menu along two dimensions: how well it sells (popularity) and how much profit it generates (profitability). The combination of these two dimensions creates a framework you can actually act on.

The Menu Engineering Framework: 4 Quadrants

Every item on your menu falls into one of these four categories:

• Stars (High Sales + High Margin): Your flagship items. They sell well and each portion generates good profit. Strategy: protect them, don't tweak the recipe, and make sure they're always available. Place them prominently on your menu.
• Puzzles (Low Sales + High Margin): Good profit per portion, but few people order them. These have potential to become Stars if you can increase awareness. Strategy: try repositioning on the menu, make them a staff recommendation, or tweak the name and presentation.
• Plow Horses (High Sales + Low Margin): Lots of people order them, but profit per portion is thin. These are tricky because removing them disappoints customers. Strategy: try raising the price slightly, reduce portion size, or substitute cheaper ingredients without sacrificing taste.
• Dogs (Low Sales + Low Margin): Not popular and not profitable. Strategy: remove them from the menu. Seriously. Every item adds complexity to the kitchen, to inventory, and to training. Dogs are just deadweight.

Step 1: Gather the Data You Need

For each menu item, you need two numbers:

Sales volume: How many portions sold in a given period (ideally 30 days). This data lives in your POS sales reports.

Food cost per portion: How much the ingredients cost to make one serving. You calculate this from your recipe cards — if you don't have recipe cards yet, this is a great reason to start.

From there, calculate:

• Profit per portion = Selling price − Food cost per portion
• Total profit = Profit per portion × Sales volume
• Food cost percentage = (Food cost / Selling price) × 100%

You don't need special software — a simple spreadsheet is enough. One column per metric, one row per menu item.

Step 2: Set Your Benchmarks

To categorize items into the 4 quadrants, you need benchmarks:

Popularity benchmark: Calculate the average sales volume across all items. Items above average = "high sales." Below average = "low sales."

Profitability benchmark: Calculate the average profit per portion across all items. Items above average = "high margin." Below average = "low margin."

Example: if you have 20 menu items and the average volume is 50 portions per month, then items selling 60+ portions are "high sales" and those at 30 are "low sales." Same logic applies to profit per portion.

Step 3: Plot and Categorize

Once you have the numbers, categorize each item:

• Volume above average AND profit per portion above average → Star
• Volume below average AND profit per portion above average → Puzzle
• Volume above average AND profit per portion below average → Plow Horse
• Volume below average AND profit per portion below average → Dog

If you want a visual, create a simple scatter plot: X-axis = sales volume, Y-axis = profit per portion. But a simple table works just fine too.

Step 4: Take Action by Category

Stars — Protect and Promote

Don't mess with Stars without a very good reason. They're your profit engine. What you can do: make sure their supply chain is stable, place them in eye-catching positions on the menu, and train your staff to recommend them when customers are undecided.

Puzzles — Increase Visibility

Puzzles have good margins but low sales. Usually the problem isn't the product — it's awareness. Try:

• Repositioning on the menu to a more prominent area (research suggests customer eyes typically focus on the upper right section of a menu)
• Marking them as "Chef's Recommendation" or a daily special
• Having staff actively mention them when taking orders
• Renaming or rewriting the description to make them more appealing

Plow Horses — Optimize the Margin

These are the trickiest. Customers love them, but your margin is thin. Options:

• Raise the price by a small amount. For items customers already like, a modest increase usually doesn't drive them away.
• Review the recipe: can any ingredients be substituted with something cheaper without sacrificing taste?
• Reduce the portion slightly (10-15%). Customers usually don't notice if the plating still looks full.
• Bundle it: make it part of a combo where the overall margin is better.

Dogs — Remove or Overhaul Completely

This is the emotionally hardest part. Maybe you created the recipe yourself. Maybe it was the first item you launched. But if the data says it's neither popular nor profitable, keeping it means you're subsidizing something nobody wants.

Cut it. Or if you still believe in the concept, overhaul it completely — new recipe, new price, new presentation. But set a deadline: if it's still a Dog 30 days after the overhaul, remove it permanently.

Step 5: Regular Reviews — Not a One-Time Thing

Menu engineering isn't a one-time project. Do this review every month or at least every quarter. Why? Because customer preferences change, ingredient prices fluctuate, and new items you've added need evaluation.

Each review, compare with the previous month's data. Did that Puzzle you repositioned become a Star? Is the Plow Horse you raised prices on still selling? Are any Stars starting to decline?

This data also informs decisions about seasonal items or new additions. Before adding a new item, check: do you have room? Is there a Dog you can replace? Adding without removing = adding complexity without guaranteed return.

A Lean Menu Beats a Long Menu

The biggest temptation for cafe owners is to keep adding to the menu for fear of losing customers. But the reality is that an overly long menu confuses customers, slows down the kitchen, creates inventory chaos, and makes training take longer.

A data-optimized menu — where every item has a reason to exist — almost always generates higher profit than a long but uncurated one. And all you need is sales data from your POS, recipe cards for food cost, and a simple spreadsheet.

Start this month. Pull 30 days of sales data, calculate food cost per item, plot them into the 4 quadrants. The results might surprise you — and the decisions you make from there will show up directly on your bottom line.