The matrix: two axes, four boxes
Everything rests on a two-dimensional matrix:
- Popularity: how many times a dish is ordered per 100 orders
- Gross margin: the profit the dish generates once food cost is deducted
Cross the two axes (high/low) and you get four categories, traditionally called stars, plowhorses, puzzles and dogs.
⭐ Stars: popular and profitable
These are your champions. Ordered constantly, with a comfortable margin. The house burger at £18 that costs £4 in ingredients, or the house tartare at £22.
Strategy: push them forward. A photo, a highlight box, top position in their section, a "house speciality" tag. Above all, leave them alone. Raise the price by £2 and you risk killing the popularity. Change the recipe and you risk killing the margin.
🐄 Plowhorses: popular but barely profitable
Ordered all the time, but on a thin margin. Usually the classics people expect: steak and chips with an expensive cut, Caesar salad with proper Parmesan, margherita pizza with mozzarella di bufala.
Strategy: three options, in order of preference.
- Work quietly on the food cost (a cheaper supplier at equal quality, a portion 10% smaller)
- Raise the price by £1 to £1.50, staying under the psychological threshold
- Shrink the dish's footprint on the menu to make room for a star you want to push
"We turned our rib of beef from a plowhorse into a star just by changing the knife we used to slice it in the kitchen. Better presentation pushed the price people accepted up by £4." — Vincent, chef in Toulouse
🐎 Puzzles: profitable but rarely ordered
Strong margin, but customers walk right past. Usually dishes with a dull description, a bad spot on the menu, or an unappealing name. A cep risotto might cost £3 to produce and sell for £24, but call it "Mushroom risotto" with no further detail and nobody notices it.
Strategy: put them in the spotlight.
- Rewrite the description with words that make people hungry ("creamy", "slow-cooked", "ceps from the Périgord")
- Add a good photo
- Move them to the top of their section
- Have the floor staff recommend them as the dish of the day
🐶 Dogs: neither popular nor profitable
Nobody orders them, and when someone does, you make nothing. These dishes clutter the menu, complicate the kitchen and tie up stock that goes to waste.
Strategy: cut them. No sentimentality. The one exception: a dish with real emotional weight that is part of your identity (the "grandad's special" that three customers a week order, but that defines the place). Keep it if you must, just don't kid yourself about it.
How to classify your dishes with QRMENUPRO
Historically, the hard part of menu engineering has been measuring popularity. On a paper menu you knew what rang through the till, but not what customers looked at without ordering.
With a digital menu, you now have three indicators:
- View rate: how many customers opened the dish's page
- Reading time: how many seconds they spent on it
- Order rate (linked to your till): orders divided by views
A dish viewed often but rarely ordered means the description or the price needs work. A dish rarely viewed needs repositioning on the menu. See our features page for more detail on the statistics module.
The right rhythm
Menu engineering is not a one-off exercise. Redo the analysis every three months. A star can slip into plowhorse territory if food cost rises; a puzzle can become a star once you change its name and add a photo. The goal is to pull your whole menu upwards, step by step.
Key points
- 4 categories: stars, plowhorses, puzzles, dogs
- Stars: leave them alone, put them front and centre
- Plowhorses: cut the food cost or raise the price gently
- Puzzles: improve the description, the photo and the position on the menu
- Dogs: cut them (unless they carry real identity value)
- Repeat the exercise every three months to track how things shift
- A digital menu gives you the data paper never could