SKU Reduction and Short Menus: Before vs After, and Who Actually Wins

For independent restaurants under 15 tables: aggressive cuts to 18-24 dishes

If you run fewer than 15 tables with a brigade of three to five people, cut down to 18-24 dishes built on a maximum of 45 unique ingredients, and that profile usually gains between 2 and 5 points of food cost within a quarter. That 42-seat steakhouse in Medellín carried 71 dishes and 214 ingredients: 14 dishes produced 78% of revenue and 31 dishes never reached two units per week, dragging along 63 ingredients that turned once every ten days. At that turn rate, waste is not an accident, it is the menu design collecting its bill. Outside pressure reinforces the case: full-service menu prices peaked at 9.0% year over year in 2022 according to the National Restaurant Association, and no small operator absorbs that while buying 214 references in absurd volumes. A short menu is measured in INGREDIENTS, not in printed lines, and that distinction decides whether the cut gives you money back or merely takes options away from your guest.

What does a short menu actually measure: lines or ingredients?

Twenty dishes where each one carries its exclusive protein, its exclusive sauce and its exclusive side run worse than thirty-four dishes assembled on 40 shared ingredients, because the first buys little of a lot and the second buys a lot of little.

The ruling ratio is dishes per ingredient: below 0.45 you own an expensive menu disguised as a varied one; above 0.60 you start negotiating real prices with your supplier. Diego F. Parra insists on that number before touching a single line of the menu, because cutting without measuring it produces the worst of both worlds: fewer sales and the same purchase cost. When kitchen turnover runs above 60% a year —a common figure in independent operations— the best option is not the shortest possible menu but the one with the most shared families. The same red wine reduction feeds three dishes, the same protein cut appears in two preparations at different prices, and a new cook learns 40 ingredients in two weeks instead of 214 in three months.

Best for kitchens with high staff turnover: shared ingredient families

Savings there do not come only from food cost: they come from the learning curve and from portioning mistakes that get paid in waste. A badly built short menu does the opposite, deleting lines while keeping the same 200 ingredients bought in smaller volumes each, so the unit purchase price gets WORSE and the owner wrongly concludes that reduction does not work. Three scenarios exist where dropping to 18-24 dishes takes cash away instead of adding it. First: operations weighted toward delivery, where the basket is built by combination and losing lines shrinks average ticket before purchase savings ever show up. Second: locations that live on limited-time offers, and the data here is blunt, 52% of consumers consider an attractive LTO important when choosing a restaurant according to Technomic 2024, so a frozen 20-dish menu with no seasonal rotation loses traffic. Third: kitchens with demand concentrated in a declining category; seafood lost menu penetration in the United States during 2024 according to SeafoodSource/Technomic, and cutting everything else to stay anchored to it concentrates risk rather than reducing it.

Cut by margin in dollars, never by the chef's taste

Cut by marginal profitability measured in DOLLARS per unit sold, crossed with the real sales mix, not by which dishes bore the kitchen. The mistake showing up in almost every menu I review is killing the dish with the ugly food cost percentage —a 38% that looks bad on the sheet— when that dish leaves 14,000 pesos of absolute margin per unit and sells 90 units weekly, while the 24% dish leaves 5,200 and sells 11. The arithmetic is grade-school: 1,260,000 against 57,200 a week. Cutting the first to protect a percentage means giving away 1.2 million monthly for accounting aesthetics. Sort the list by total contribution margin, mark the cut where the cumulative reaches 92% of revenue, and argue afterward. Four signals tell you the short-menu proposal on your table was built wrong. If unique ingredient count did not drop at least 40% while dishes dropped 60%, the cut was cosmetic and your purchase list stays just as expensive.

Red flags when comparing reduction options

If nobody recalculated minimum orders with the supplier after the cut, you will buy less volume at the same unit price and lose margin. If the proposal removes dishes without touching the sales mix analysis of the last 90 days, it is guessing. And if the new menu has no high-turn ingredient shared between starters and mains, you never built families, you only pruned. That fourth signal is what most often turns a savings project into a sales drop. When your bottleneck is kitchen space rather than the dining room, a short menu buys you capacity without construction. With 45 ingredients you free up cold storage, shrink inventory counts from two hours to forty minutes, and eliminate dead prep for dishes that left the pass twice a week. Fast-casual, which works with tight menus, shows 64.7% plant-based penetration against 31.6% in fine dining according to Plant Based Foods Association/Datassential 2024, and that gap is not ideological: short menus absorb a new line without disrupting service because they have the mise en place room to do it.

Best for operations that want volume growth without more kitchen square meters

The long menu, by contrast, hides the dishes that subtract profitability behind the general food cost average and blinds you to where the next thing fits. Assume you keep those 71 dishes twelve more months. The 63 slow-moving ingredients keep getting purchased below the profitable minimum order, so you pay between 8% and 15% in surcharge per reference; waste on those products eats another point and a half of food cost; and every passing month your supplier list becomes harder to renegotiate because you carry no volume to offer on any line. By year end that menu cost a 42-seat business something close to three points of operating margin that never appeared on a P&L under the right name: they showed up as "high cost of sales". Open your last 90 days of sales mix this week and sort it by absolute margin. A financial cut works on marginal profit per dish —contribution margin in currency, never in percentage— crossed against units sold.

What separates a cut that makes money from one that only annoys customers?

An instinctive cut works on kitchen preferences and ends up killing dishes that delivered high absolute margin even when their percentage food cost looked ugly.

A properly built short menu shares ingredients across dish families: the same red wine reduction serves three plates, and the same protein cut appears in two preparations at different prices. A badly built short menu simply deletes lines while keeping the same 200 ingredients, now purchased in smaller volume each, which makes the unit purchase price WORSE. Long menus hide the dishes that hurt profitability behind the blended food cost average. Drop to 22 dishes and every one of them stands exposed: if a plate breaks the 32% we set as the per-dish ceiling, there is nothing left to dilute it with, and the decision becomes obvious at the next price review. Classic menu engineering sorts dishes into stars, plowhorses, puzzles and dogs. SKU reduction adds a layer that matrix lacks: the INVENTORY COST each dish imposes on the business, including locked capital and the spoilage risk of its exclusive ingredients.

What separates a cut that makes money from one that only annoys customers — in practice?

Two dishes with identical margin can carry inventory costs that differ by a factor of four. In delivery the arithmetic partly inverts.

Cutting too deep sinks the average ticket, because a customer building an order for two or three people runs out of options, and the fixed cost of the drop gets spread across less money. That is why the delivery profile does not win at 12 dishes: it wins with a 22-dish core plus a layer of high-margin add-ons that lift the ticket without adding new purchase references. Diego F. Parra keeps repeating inside Masterestaurant that restaurant menu design starts on the purchase sheet, not in the graphic design file. If the ingredient list does not shrink, the menu was not shortened: it was only restyled.