Restaurant menu engineering: the card as a sales tool

Restaurant menu engineering decides whether a venue survives or thrives. The menu is the most powerful commercial document a restaurant holds, and an optimised version can raise profit by up to 15% without increasing prices. With Italian operating margins between 5% and 10%, a single percentage point matters. Most operators still build the menu the way they did thirty years ago: culinary creativity, a chef's advice and prices estimated from ingredient costs.

The Italian market numbers

The sector is under strain. In 2024, 12,368 bars closed, a net loss of 8,409 units, and the total number of businesses in the sector fell to 328,000. Average net operating margin sits between 5% and 7%. Food cost absorbs 25% to 35% of revenue, and staff is the heaviest item at 30% to 40%. Average monthly electricity costs reach EUR 2,470 for restaurants and about EUR 1,160 for bars. In that context, menu optimisation is a survival question rather than a refinement.

The Smith-Kasavana matrix

Menu engineering classifies each dish by two objective parameters. Popularity measures how much a dish sells relative to others in its category. Margin measures the profit it generates after ingredient cost. Crossing the two produces four groups. Stars sell well and earn well, and they deserve promotion and protection. Plowhorses sell well but earn little; they carry volume and need rebalancing through a small price change or a lower ingredient cost that keeps perceived quality intact. Puzzles sell poorly but earn well; they need repositioning, a better description or a professional photograph. Dogs sell poorly and earn little, and they should leave the menu to free space in the kitchen and the card. A risotto at EUR 12 margin and a sea bass at EUR 17.50 illustrate the difference: pushing the risotto over the bass loses EUR 5.50 per plate.

Food cost targets by category

Not every dish behaves the same. A tomato pasta can hold a margin above 30% at a fair price, while a raw fish starter will typically sit below it. Realistic food cost benchmarks are 25% to 30% for first courses, 28% to 35% for meat mains, 32% to 38% for fish mains, 15% to 20% for sides and 18% to 22% for desserts. When a dish moves significantly outside its range, the operator decides whether to keep it, adjust it or replace it.

Pricing psychology

Knowing which dishes to promote is half the work. The other half guides the customer toward them. Charm pricing, which ends prices in 9 or 99, raises sales by at least 24%: EUR 9.90 feels clearly cheaper than EUR 10 even though the gap is ten cents. The technique suits mid and lower price bands. For premium proposals, round prices such as EUR 18, EUR 25 or EUR 35 signal quality. The decoy effect works on the same logic: with three steaks at EUR 18, EUR 28 and EUR 45, the middle option looks like the sensible compromise, and most customers choose it. Anchoring completes the picture: when the first price a guest sees is EUR 38, a EUR 22 dish begins to look reasonable.

Layout and descriptions

Placement and wording decide how much of the analysis reaches the guest. High-margin dishes belong in high-visibility areas, such as the upper-right corner or the top of each section, where eye-tracking studies locate the golden triangle. Removing the euro symbol, so that a dish reads with a bare number, reduces the mental association with spending. Sensory descriptions raise sales by 27%; a plain grilled fillet becomes a fillet cooked over a wood fire, described in a few precise words rather than a long sentence. Stars earn a discreet visual signal, a box or a slightly larger type, and nothing more.

Tools and a practical case

Dedicated menu engineering platforms start around EUR 89 per month, while MenuCalc is available at about EUR 49 per month; both automate the calculations and produce ready reports. A well-structured spreadsheet with the right formulas covers most of the work: the full list of dishes with food cost, sales per dish over at least 90 days, gross margin per dish and the classification into the four groups. Modern restaurant management systems from vendors such as TeamSystem or Syrve already include menu engineering modules, and it is worth checking whether the function is included before choosing one.

A real case shows the effect. A traditional trattoria with 50 covers and 28 dishes analysed 90 days of transactions. Average margin was 32%, and three dogs represented 12% of the menu but only 4% of sales, with negative margin once waste was counted. The operator removed the three dogs, added professional photographs and rewritten descriptions to two high-margin puzzles, and raised the price of one very popular plowhorse from EUR 9.50 to EUR 11 while improving its presentation. After four months, gross profit rose 22% without changing the products, and the average ticket grew by EUR 2.80 per cover. The ideal mix is around 60% plowhorses for volume, 25% stars for profit and image and 15% puzzles for testing, reviewed every three months with the seasonal menu.

The card works inside a wider demand curve, and event pricing in Italian tourism shows how Jubilee and the Olympics move that curve.

Before the table, though, the customer searches: voice search for local businesses is where a menu becomes discoverable.