Menu engineering with sales data

Menu engineering pairs how often a dish sells with how much it earns, then sorts every item on your menu into one of four quadrants. It's a classic restaurant-consulting technique, MenuSquared runs the math for you from a sales export.

• Stars: Popular and high margin. Protect these and keep them visible.
• Plowhorses: Popular but low margin. Good candidates for a small price nudge or a cheaper build.
• Puzzles: High margin but overlooked. Reposition them or promote them harder.
• Dogs: Low margin, low sales. Rework the dish or retire it.

Menu engineering is available on every plan, including Starter. The quadrant math runs locally in your browser and doesn't use AI credits. Only the optional AI-written summary of your results uses a credit.

You'll need an item-level sales report from your point of sale, not just a daily total. Most POS systems can export this as a CSV:

• Square: Reports > Items, choose a date range, then export as CSV.
• Toast: Reporting > Sales Summary, pull an item-level breakdown for your date range, then export.
• Any other POS: Any CSV works as long as it has a column for item name and a column for quantity sold. A revenue or net sales column is optional but improves the results.

MenuSquared automatically detects the item name, quantity, and revenue columns from common header names, so you don't need to reformat the file first.

1. Open the Sous Chef tab (chef hat icon in the editor sidebar), then switch to the Sales tab.
2. Upload your CSV. If MenuSquared can't confidently detect your columns, you'll see a quick mapping screen: pick which column is the item name, which is the quantity sold, and, optionally, which is revenue. A preview of the first row helps you check your picks.
3. Review the matches. Each row in your CSV is matched to a menu item by name. Strong matches are accepted automatically; anything uncertain lands in an "Unmatched" list where you can assign it to the right menu item or ignore the row, useful for modifiers, discounts, or voided-item lines that shouldn't count.
4. Confirm the import. MenuSquared saves the matched quantities and revenue and shows your results immediately.

The results view opens with your top Star and top Dog, tap either to jump straight to that item in the editor. Below that, a breakdown lists every analyzed item under its quadrant, with unit counts and a share of total units sold.

• Popularity is based on each item's share of total units sold, weighed against how many items are on your menu, so a 10-item menu and a 60-item menu are scored fairly against their own size.
• Margin uses the High Margin flag from the item inspector when you've set one. Otherwise, MenuSquared estimates it from price: anything priced at or above the median of your analyzed items counts as high margin. A note at the bottom of the results tells you how many items are using the estimate.

Results are recalculated live from your current menu every time you open the tab, so editing a price or toggling the High Margin flag reclassifies an item instantly, no re-import needed. If you rename or delete items that were part of your last import, you'll see a prompt to re-import a fresh CSV.

From the results view, you can click Narrate with AI to get a written summary: a headline and a few sentences per quadrant, with the specific items behind each one linked so you can jump to them. On the Starter plan, each summary draws from your shared pool of 10 monthly AI credits; Chef's Table and Business plans include unlimited AI. You can regenerate the summary at any time.