Menu engineering

A high food-cost percentage isn't the whole story. An item at 34% food cost that sells 400 a week out-earns a 22% item that sells 12 — but per-item margin alone never shows that, because it's never multiplied by how many you sell. The Menu tab in Reports does that multiplication, so you can see which dishes actually pay the bills.

The four quadrants

Every item on the menu is placed on two axes over the period you pick:

  • Popularity — the item's share of all units sold (its menu mix), measured against 70% of an equal share. If you had 10 items and each sold the same amount, each would hold 10%; an item is "popular" once it clears 70% of that fair share.
  • Profitability — the item's contribution margin per unit (selling price − food cost), measured against the average across everything you sold.

Crossing the two gives four groups:

  • Star — popular and profitable. Your winners; protect them and feature them.
  • Plowhorse — popular but below-average margin. People love it, but it earns thin — re-cost the recipe or nudge the price.
  • Puzzle — high margin but unpopular. It makes money when it sells; it just doesn't sell — promote it, move it up the menu, or rename it.
  • Dog — unpopular and low margin. A candidate to rework or retire.

Reading the tab

The scatter plots every item by units sold (across) and contribution per unit (up), with the bubble sized by revenue and dashed lines marking the two averages, so each item falls into one of the four corners. Below it, the table lists every item sorted by total contribution dollars — revenue minus food cost over the period — which is the number that tells you what a dish is worth to the kitchen, not just its percentage.

  1. Use the date range at the top. Menu engineering defaults to the last 30 days, so menu-mix shares have enough sales behind them to mean something.

  2. These are where the money is. A Plowhorse is a pricing or portioning fix; a Puzzle is a merchandising fix.

  3. The top rows are the dishes carrying your menu. Tap any row to open its full menu item report.

Combos count as their own item here — a bundle has its own price and its cost is the sum of its components — because menu engineering rates what the customer actually buys and pays for. Archived items still appear if they sold in the period, labeled so you can tell them apart.

Where the numbers come from

Food cost per sale is the same figure the rest of the app uses — weighted-average receipt cost, recipe-explored, portion-scaled — see cost and pricing. Units sold come from your uploaded sales. Contribution is simply revenue minus food cost; nothing here fires an alert or changes your data.

Common mistakes

With only a handful of items selling, the popularity bar sits low and the margin average swings on a few dishes. Give it a wider date range and more sales history before reading too much into the quadrants.

Items with no cost basis at all — no recipe ingredients or combo components — are left out, because a zero-cost item would read as pure profit and distort the averages. Add the recipe or components and it appears.

A low food-cost % on something nobody orders earns nothing. Sort by contribution dollars, not percentage, to see what actually funds the kitchen.