Restaurant Menu Engineering: Sell More with Data
What menu engineering is, how to do it and which metrics matter: the Star/Plowhorse/Puzzle/Dog matrix, price psychology, product positioning and data-driven decisions with a QR menu.
Menu engineering is the discipline of designing a restaurant menu scientifically to maximise profit. Unfortunately, most restaurant owners run their menu on intuition — yet businesses that decide based on data earn noticeably more profit from the same menu. In this article, we look at the core menu engineering matrix, price psychology and how QR menu analytics fits in.
The Menu Engineering Matrix: Four Categories
Every item on a menu has two dimensions: profitability (profit margin) and popularity (sales volume). Plotted on a matrix, these two dimensions create four categories:
| Category | Profitability | Popularity | Strategy |
|---|---|---|---|
| ⭐ Star | High | High | Keep it in the spotlight, feature the image, recommend it |
| 🐎 Plowhorse | Low | High | Lower the cost or raise the price |
| 🧩 Puzzle | High | Low | Market it, promote it, strengthen the description |
| 🐕 Dog | Low | Low | Remove it from the menu or reinvent it |
1. Stars — Your Showcase Items
Both profitable and popular. These are your menu's flagships. Strategies:
- Place them at the top of their category (top = first in the eye's path)
- "Chef's Recommendation" and "Most Popular" badges
- Your best images
- Detailed, appetising descriptions
- A permanent fixture on the staff recommendation list
Stars lift your average order value.
2. Plowhorses — Big Sellers, Thin Margins
Guests love them, but the margin is low. Removing them is a risk (you lose sales); keeping them as they are costs you too. Strategies:
- Cost optimisation — review portion sizes and bring the cost down
- Raise the price by 5–10% — grow the margin while keeping the popularity
- Add a cross-sell — pair the Plowhorse with a high-margin side dish
- Add a premium variant — a high-margin version (Köfte → Köfte Special)
3. Puzzles — Profitable but Overlooked
High margin, but guests don't choose them. Usually the problem is a lack of promotion. Strategies:
- Enrich the description (sensory words, how it's prepared)
- Strengthen the image (make it professional with AI)
- Place it next to a Star (to catch attention in the eye's path)
- Add a "New" or "Chef's Recommendation" badge
- Introduce it on the staff recommendation list
Once promoted, Puzzles turn into Stars — the most profitable move in menu engineering.
4. Dogs — Low Sales, Low Profit
They take up space on your menu without creating value. Strategies:
- Remove — simplifies kitchen operations and reduces stock
- Reinvent — change the recipe or presentation and turn it into a Star
- Make it seasonal — activate it only at certain times of year
The Data Behind Menu Engineering: QR Menu Analytics
For the matrix to work, you need two pieces of data:
- Sales volume — from your POS system (how many were sold)
- Profit margin — from your cost calculations (how much profit each makes)
A QR menu adds a third dimension: taps / views. This is critical:
- If an item gets many taps but few sales = the description or image works; the price may be wrong
- Few taps = poor menu position, a weak image or a thin description
- High taps + high orders = a Star
Two more signals round out the matrix:
- Items viewed together — which items guests look at in the same visit. If many guests who open the slow-roasted lamb also look at a cocktail in that session, placing the two closer on the menu makes a natural cross-sell.
- Hour and table heatmaps — when and at which tables demand peaks. Direct evidence for staff planning and campaign timing.
One note: the QR menu on the table is mostly a viewing channel, and the waiter takes the order. That is why QrMenum does not claim to know "who ordered what" from a scan — order data comes from your POS, interest data from the menu.
Price Psychology
An often-overlooked dimension of menu engineering is how prices are presented. Its impact is significant:
- Drop the currency symbol — "150" hurts less; "₺150" reminds guests they are spending money
- Round numbers — ₺150, not ₺149; charm pricing reads as the opposite of luxury
- Decoy effect — a ₺350 item raises the perceived value of a ₺250 item
- Order — expensive at the top, cheaper below → a higher average
Category Order
Guests read the menu from top to bottom. What they see first is what they are most likely to choose. The order should be strategic:
- Stars at the top
- High-margin categories first (usually mains)
- Seasonal items highlighted
- Desserts and drinks last (add-on sales)
Automated Menu Engineering in QrMenum
The QrMenum AI Menu Engineering module:
- Analyses tap, order and co-view data
- Builds the Star/Plowhorse/Puzzle/Dog matrix automatically
- Suggests which action to take for each item
- Recommends description improvements
- Flags seasonal fluctuations
A Practical Path: The 30-Day Menu Engineering Cycle
- Weeks 1–2: Data collection (once analytics is active)
- Week 3: Build the matrix and assign categories
- Week 4: Take action (new images for Puzzles, price adjustments for Plowhorses, remove Dogs)
- Weeks 5–8: Collect new data and measure the impact
- The following month: Start a new cycle
After a few cycles, your menu becomes measurably more profitable.
Conclusion
Menu engineering runs on data, not intuition. Without a QR menu there is no data; without data, every decision is a guess. QrMenum's analytics and AI Menu Engineering module make this cycle automatic.
QR menu basics: What Is a QR Menu? The Complete Guide →
Smart menus with AI: Smart QR Menu: Transforming Your Menu with AI →
