Menu Engineering for a Wine List

Menu engineering sorts the references on a list along two axes: margin per unit and sales volume. Four quadrants come out of it — stars, high margin and high volume; puzzles, high margin and low volume; plowhorses, low margin and high volume; and dogs, low on both — each calling for a different move. Applied to a wine list, the grid is a genuinely effective steering tool. Here is how it works in practice.

Menu engineering in two minutes

The method was formalised in American restaurants in the 1980s, and it puts structure on something good operators have always done by instinct: not every item on a list is worth the same. Some earn well and sell well, and belong at the top. Some earn little but bring people through the door, and stay for the traffic. Some earn well and barely sell, and need better promotion rather than a lower price. And some earn little and sell little, and should go.

The two axes, applied to wine

• Margin per unit. Calculated per bottle, or per pour if that is what you are measuring, this is the gross profit each sale produces. Draw the median of your list: everything above it is high margin, everything below is low.
• Sales volume. Measured over a period long enough to mean something, three to six months, this is how often the reference is ordered. Draw the median again: above is high rotation, below is low.

The four categories

Stars — high margin, high rotation

Your champions. Good price, steady movement, and the backbone of your beverage profitability. Protect them and feature them. Top of the section, an occasional sommelier's pick, and a floor team trained to recommend them first. Watch your supply: losing a star for three weeks hurts more than it looks like it should.

Puzzles — high margin, low rotation

Bottles that earn well when they sell and rarely sell. Usually the exceptional references, the serious growths, the niche producers. Improve the promotion without touching the price. A richer tasting note, the story of the producer, a specific dish pairing. Let the virtual sommelier put them in front of guests whose stated budget supports them. If rotation has not moved after three months, accept that these are image references with a real role, and size the stock accordingly.

Plowhorses — low margin, high rotation

They sell well and earn little per unit. Guests like them, you move volume, and the total margin they produce is lower than the volume suggests. Test a gradual reprice, five to ten percent at a time, and read the elasticity — a guest rarely walks over a dollar or two. If the margin is genuinely too thin, look for a comparable wine with a better cost from another distributor.

Dogs — low margin, low rotation

They earn nothing and they do not move. Take them out. Clear them through a promotion — a discovery pour, a wine-included menu — or send them to the kitchen. The shelf space they free has to go to something with more potential. A cellar is not a museum, it is an asset that has to produce.

Sorting your own references

The grid needs two figures per reference: margin per unit, which follows from your purchase price and list price, and sales volume over a representative period. Without software that is a laborious spreadsheet exercise. With analytics in place, the matrix builds itself from your sales and your entered costs, and refreshes daily.

What to do with each, in one place

• Stars — protect, put at the top of the section, train the team to lead with them.
• Puzzles — enrich the promotion, and leave the price alone except in extreme cases.
• Plowhorses — test a gradual reprice, and look for better-cost alternatives.
• Dogs — clear them, and free the space for something better.

Where the data comes from

The link between your cellar, where you enter purchase costs, and your POS, which reports sales, builds the matrix on its own. You arrive each morning with a current list of your stars, puzzles, plowhorses and dogs, and you make your decisions on promotion, pricing and purchasing from numbers rather than instinct.