← Back to library
OperationsOpen1 prompt · paste & run
The Menu Engineer: Paste Your POS Export, Get the 4 Menu Moves Worth Real Margin
Menu consultants charge $2,000+ for this analysis. Paste item-level sales from your POS and get every item bucketed by popularity and margin, plus exactly four moves with dollar impacts attached.
Copy & run
**What this is.** This builds play #2 from our list *8 AI Plays for a Restaurant or Multi-Unit Restaurant Group*: menu engineering, the analysis consultants charge $2,000+ to run. Paste item-level sales from your POS and get every menu item bucketed by popularity and margin, plus exactly four moves with dollar impacts. **How to use it.** Export item-level sales for the last 30–90 days; Toast, Square, and Clover all do this in a few clicks. Add food cost per item where you know it; the prompt handles gaps with labeled assumptions instead of stopping to ask. --- > **Role:** You are a restaurant profitability analyst who has run menu engineering for hundreds of independent restaurants. You work only from the numbers you're given; you never invent or smooth a figure. > > **Context:** My restaurant in one line: [e.g. "fast-casual Mexican, 80 seats, $28 average check"]. Item-level sales from my POS for the last [30/60/90] days (item name, units sold, menu price): [paste export]. Food cost per item where I know it: [paste, or write "unknown: use typical cost percentages for this cuisine and mark every assumption with a *"]. > > **Task:** > 1. For each item, compute popularity (units vs. the average item) and contribution margin (price minus food cost). Mark every assumed cost with a *. > 2. Bucket every item: **Stars** (popular + high margin), **Plow-horses** (popular + low margin), **Puzzles** (unpopular + high margin), **Dogs** (unpopular + low margin). > 3. Recommend exactly 4 moves: one reprice, one re-cost or re-plate, one menu repositioning, one cut-or-replace. Each move names the item, the specific change, and the estimated monthly dollar impact at current volume. > 4. Flag any item where a price increase under $1 is unlikely to move volume, and what that's worth per month. > > **Format:** One table (item / units / price / cost / margin / bucket), then the 4 moves as a numbered list with dollar impacts, then a 3-line summary I can read to my kitchen manager. > > **Constraints:** Use only the numbers I pasted. Label every assumption. No generic advice: every recommendation names an item and a number. If my export is missing something, make a labeled assumption and keep going; don't stop to ask questions. --- **What it's worth.** Menu engineering typically finds a 2–4 point margin swing on your top sellers. On $80,000/month of food revenue, two points is $1,600 a month, from one paste. **Caution.** Make one pricing move at a time and re-run this in 30 days to see what volume actually did. And if your export happens to include employee or customer names, strip them first or work in a private/business AI workspace.