Menu engineering

The only two pieces of data you need

How many times each dish is ordered in a period, and how much margin each leaves in euros. The first comes from the POS. The second comes from the costing: selling price excluding VAT minus raw ingredient cost.

One month of service is enough to see the pattern. You do not need a year or a new system: you need to put the two columns next to each other, which is the thing that almost never gets done.

The four groups

You draw the average of sales and the average of margin, and every dish falls into a quadrant. The point is that each quadrant calls for a different decision, and only one of them is "leave it alone".

                 HIGH MARGIN          LOW MARGIN
               ┌────────────────────┬────────────────────┐
  SELLS        │  STARS             │  PLOUGH HORSES     │
  A LOT        │  Do not touch.     │  Cut the cost or   │
               │  Protect them and  │  raise the price.  │
               │  put them where    │  If neither works, │
               │  the eye lands.    │  off it goes.      │
               ├────────────────────┼────────────────────┤
  SELLS        │  PUZZLES           │  DOGS              │
  LITTLE       │  Good dish, bad    │  Sells nothing and │
               │  placing or bad    │  leaves nothing.   │
               │  name. Move it     │  Takes up kitchen, │
               │  and rename it.    │  stock and space.  │
               └────────────────────┴────────────────────┘

Plough horses: the dish costing you money while it flies

This is the finding that surprises people most. The dish everyone describes as "that one goes out constantly" is usually the one that leaves least, and that is exactly why it goes out so much: it is cheap for what it costs to make.

Before raising its price, look at the cost: often it gets fixed by adjusting the portion, changing the garnish or renegotiating the produce, without touching the menu or startling the regulars.

Puzzles: good dish, wrong place

A dish with a good margin that almost nobody orders is rarely a bad dish. It is usually a problem of placing, of name or of description. It is last in its section, it is called something that says nothing, or it sits next to the star.

Moving a dish to first or second place in its block — which is where the eye lands — and giving it a name people understand changes its sales more than you would expect. It costs nothing.

• Move it into the first two positions of its section.
• Give it a concrete name: "Grilled octopus with potato" sells better than "Octopus special".
• Describe what is in it, not how good it is.
• Avoid lining every price up in a column: it invites comparison by price.

Dogs: taking dishes off is a management decision too

A dish that neither sells nor leaves margin still costs: it takes a line in the kitchen, it forces you to hold produce that turns slowly, and it lengthens the menu. Every dish you take off cuts wastage, speeds up service and makes what you do want to sell easier to read.

The usual resistance is "but there is one customer who always orders it". Worth working out what keeping that dish for that customer costs before deciding.

How often to do it

Twice a year is enough in a venue with a stable menu, plus every menu or season change if you move it. What does not work is doing it once and considering it solved: the matrix moves on its own as soon as purchase prices change.

And it is worth doing per service: the lunch menu and the evening menu can have different stars, and mixing them gives an average that describes neither.

Frequently asked questions