--- name: menu-engineer description: "Analyze and optimize a restaurant menu for profitability. Food cost calculations, menu matrix (Stars/Plowhorses/Puzzles/Dogs), pricing strategy, description rewrites that increase average check, and strategic placement recommendations. Use when the owner says 'my menu isn't profitable,' 'what should I charge,' 'rewrite my descriptions,' or 'what should I cut.' Uses Perplexity for food cost benchmarks and competitor pricing." --- # Menu Engineer Skill > **Purpose:** Turn your menu from a list of food into a profit engine. Every item classified, priced right, described to sell, and placed where eyes go first. --- ## PREREQUISITE - **Perplexity MCP** — food cost benchmarks, competitor pricing, ingredient cost trends - **Market Recon data** (optional but helpful) — competitor pricing for context --- ## WHAT THIS SKILL DOES - Calculates **food cost percentage** for every menu item - Builds the **Menu Matrix** — classifies every item as Star, Plowhorse, Puzzle, or Dog - Identifies **pricing opportunities** — where to raise prices without losing volume - Rewrites **menu descriptions** using sensory language that increases perceived value - Recommends **strategic placement** — what goes in the "golden triangle" and eye-tracking sweet spots - Calculates **contribution margin** per item — not just cost %, but actual dollars of profit - Flags **menu bloat** — items that add complexity without adding profit - Provides **what-if scenarios** — "if you raise pasta $2 and cut 3 dogs, here's what happens" ## WHAT THIS SKILL DOES NOT DO - Does not replace your accountant or POS system data - Does not track real-time ingredient prices (uses market averages) - Does not design the physical menu layout (gives placement recommendations) - Does not account for waste, spoilage, or prep labor — pure ingredient cost basis --- ## INPUTS NEEDED | Input | Required | Example | |-------|----------|---------| | Menu items with prices | Yes | "Margherita Pizza $14, Chicken Parm $22" | | Food cost per item | Ideal | "$3.20 ingredient cost" per item | | Monthly item sales count | Ideal | "Margherita: 180/month, Chicken Parm: 95/month" | | Cuisine type | Yes | "Italian," "Mexican," etc. | | Location | Helpful | For market pricing comparison | | Current average check | Helpful | "$28 per person" | | Target food cost % | Optional | Default: 28-32% | > **Minimum viable input:** Menu items with prices + cuisine type. Everything else improves accuracy. --- ## GOAL - Primary: Increase per-guest profitability by 15-25% through pricing and mix optimization - Secondary: Simplify the menu — fewer items, better execution, lower waste - Tertiary: Rewrite descriptions that increase perceived value and order rate of high-margin items --- ## THE MENU MATRIX Every menu item falls into one of 4 quadrants based on two axes: - **Popularity** (sales volume — high or low) - **Profitability** (contribution margin — high or low) ``` HIGH PROFIT │ PUZZLES │ STARS (High margin, │ (High margin, low volume) │ high volume) → Fix or feature │ → Protect + promote │ ──────────────────────┼────────────────────── │ DOGS │ PLOWHORSES (Low margin, │ (Low margin, low volume) │ high volume) → Cut or rework │ → Raise price or │ reduce cost │ LOW PROFIT ``` ### Star Items (High Profit + High Volume) - Your best items. Protect these. - Don't change the recipe. Don't raise the price aggressively. - Feature them: menu placement, server recommendations, social media - These are your identity — customers come FOR these ### Plowhorse Items (Low Profit + High Volume) - Popular but not making you money - Two options: raise the price (test $1-2 increase) or reduce food cost (smaller portion, cheaper protein swap, adjust garnish) - Never cut a plowhorse — customers love them. Just fix the economics. ### Puzzle Items (High Profit + Low Volume) - Great margins but nobody's ordering them - Fix the description (make it sound irresistible), move it to a better menu position, have servers recommend it, run it as a special - If puzzles still don't sell after 60 days of promotion → consider cutting ### Dog Items (Low Profit + Low Volume) - Nobody orders them and they don't make money - Cut them. Every dog on your menu adds complexity, inventory, waste, and training burden - Exception: if a dog serves a strategic purpose (kids menu item, allergen-friendly option) — keep it but don't feature it --- ## FOOD COST CALCULATION ``` FOOD COST % = (Ingredient Cost / Menu Price) × 100 TARGET BY CATEGORY: Appetizers: 25-30% Entrees: 28-35% Pasta/Rice: 22-28% (carbs are cheap — these should be your margin heroes) Seafood: 30-38% (acceptable to run higher) Desserts: 20-28% (flour + sugar + butter = cheap) Non-alc drinks: 10-20% Cocktails: 18-22% Wine (glass): 25-35% Wine (bottle): 200-300% markup standard Beer: 20-28% OVERALL RESTAURANT TARGET: 28-32% IF overall food cost > 35% → You have a pricing or waste problem IF overall food cost < 25% → You might be underportioning (check reviews for "small portions") ``` **Contribution Margin (the number that matters more):** ``` CONTRIBUTION MARGIN = Menu Price - Ingredient Cost Example: Margherita Pizza: $14 - $3.20 = $10.80 margin (77%) Lobster Ravioli: $28 - $11.50 = $16.50 margin (59%) The lobster has a WORSE food cost % but BETTER contribution margin. You make $5.70 MORE per order of lobster. Contribution margin > food cost % for decision-making. ``` --- ## DESCRIPTION REWRITING Bad menu descriptions leave money on the table. Good descriptions increase item order rates by 27% (Cornell research). ### Rules: 1. **Use sensory words** — "crispy," "slow-braised," "wood-fired," "hand-pulled," "charred" 2. **Name the source when premium** — "San Marzano tomatoes," "grass-fed beef," "locally-sourced" 3. **Keep it to 2 lines max** — Nobody reads a paragraph 4. **Don't describe the obvious** — "served with a side of bread" adds nothing 5. **Lead with the cooking method** — "Wood-fired" > "Our pizza" 6. **Create scarcity or origin** — "Nonna's recipe," "seasonal," "house-made daily" ### Before/After Examples: **Before:** "Chicken Parmesan — Breaded chicken breast with marinara sauce and mozzarella. Served with spaghetti." **After:** "Chicken Parmigiana — Hand-breaded, pan-fried until golden, layered with house-made marinara and melted mozzarella. Over al dente spaghetti." **Before:** "Caesar Salad — Romaine lettuce with Caesar dressing, croutons, and parmesan." **After:** "Classic Caesar — Crisp romaine, shaved Parmigiano-Reggiano, garlic croutons, house-made Caesar." **Before:** "Mushroom Risotto — Risotto with mushrooms." **After:** "Wild Mushroom Risotto — Arborio rice, slow-stirred with cremini, shiitake, and oyster mushrooms, finished with truffle oil and aged Parmesan." --- ## PRICING STRATEGY ### The Anchor Effect - Place your highest-margin item at the top right of each section (eye tracking sweet spot) - Your most expensive item makes everything else look reasonable - Your second-most-expensive item is where most people land (the "decoy effect") ### Price Ending Psychology ``` $14.00 → Feels precise, premium $13.95 → Feels like a deal (retail psychology — works for fast-casual) $14 → Clean, modern (no cents — trending in upscale casual) Rule: Match your positioning. Fast-casual: .95 or .99 endings are fine Casual dining: whole dollar or .00 Fine dining: no dollar signs, no decimals ("Filet Mignon 52") ``` ### How Much Can You Raise Prices? ``` IF item is a STAR → Raise $0-1 max (don't mess with what works) IF item is a PLOWHORSE → Raise $1-3 (popular enough to absorb it) IF item is a PUZZLE → Don't raise — focus on selling more of them IF item is a DOG → Don't raise — cut it IF you're 15%+ below market → Raise to market average in one move IF you're at market → Raise $1-2, monitor volume for 30 days IF you're above market → Don't raise unless you're a Star with no competition ``` --- ## WHAT-IF SCENARIOS Always provide impact projections: ``` SCENARIO: Raise pasta prices $2 + cut 3 dogs + promote 2 puzzles Current monthly revenue (estimated): Pasta section: 450 orders × $18 avg = $8,100 Dog items: 45 orders × $16 avg = $720 Puzzle items: 30 orders × $24 avg = $720 After changes: Pasta section: 430 orders × $20 avg = $8,600 (+$500/month) Dogs cut: -$720/month revenue BUT -$540/month cost = -$180 net Puzzles promoted (2x volume): 60 orders × $24 avg = $1,440 (+$720/month) NET IMPACT: +$1,040/month (+$12,480/year) Plus: reduced waste, simpler inventory, faster kitchen, better execution ``` --- ## MCP USAGE ### Perplexity — Food Cost Benchmarks ``` perplexity_ask: "Average food cost percentages for [cuisine type] restaurants in [city/region] in 2026. Break down by category: appetizers, entrees, pasta, seafood, desserts, beverages. Also include average ingredient cost trends for [specific ingredients if relevant]." search_context_size: "high" ``` ### Perplexity — Competitor Menu Pricing ``` perplexity_ask: "Average menu prices for [cuisine type] restaurants in [ZIP] [city]. What do comparable casual [Italian/Mexican/etc] restaurants charge for appetizers, entrees, pasta, pizza, and desserts?" search_context_size: "medium" ``` --- ## DECISION LOGIC ``` IF owner provides full menu with costs AND sales data → Full Menu Matrix + pricing + descriptions + placement + what-if scenarios IF owner provides menu with prices but NO food costs → Estimate food costs using Perplexity benchmarks, flag as estimates Still build Menu Matrix using price-based margin estimates IF owner provides menu with prices but NO sales data → Can't do true Matrix (need popularity data) Do: pricing analysis, description rewrites, benchmark comparison Ask: "Which 5 items sell the most? Which 5 sell the least?" IF owner says "just rewrite my descriptions" → Skip Matrix. Rewrite all descriptions using sensory language rules. IF owner says "what should I charge" → Pricing analysis only. Pull competitor data. Benchmark by category. IF market-recon data exists → Use competitor pricing from market-recon instead of new Perplexity call DEFAULT → Ask for menu + prices at minimum. Estimate what's missing. ``` --- ## COMPLETE EXAMPLE ### Example Input: > Engineer my menu. Italian casual restaurant in Austin 78704. Average check $26. Here's my top items with costs: > - Margherita Pizza: $14, cost $3.20, sells 180/month > - Chicken Parm: $22, cost $7.80, sells 95/month > - Lobster Ravioli: $28, cost $11.50, sells 35/month > - Caesar Salad: $10, cost $2.80, sells 140/month > - Mushroom Risotto: $18, cost $4.50, sells 25/month > - Spaghetti Bolognese: $16, cost $5.60, sells 110/month > - Tiramisu: $10, cost $2.20, sells 60/month > - Bruschetta: $9, cost $2.50, sells 45/month ### Example Output: --- ## MENU ENGINEER **Restaurant:** Rosario's Italian Kitchen | **Location:** Austin, TX 78704 *Generated: March 12, 2026 | Average Check: $26* ### ITEM ANALYSIS | Item | Price | Cost | Margin | Food Cost % | Monthly Vol | Classification | |------|-------|------|--------|-------------|-------------|---------------| | Margherita Pizza | $14 | $3.20 | $10.80 | 23% | 180 | ⭐ STAR | | Caesar Salad | $10 | $2.80 | $7.20 | 28% | 140 | ⭐ STAR | | Tiramisu | $10 | $2.20 | $7.80 | 22% | 60 | ⭐ STAR | | Spaghetti Bolognese | $16 | $5.60 | $10.40 | 35% | 110 | 🐴 PLOWHORSE | | Chicken Parm | $22 | $7.80 | $14.20 | 35% | 95 | 🐴 PLOWHORSE | | Lobster Ravioli | $28 | $11.50 | $16.50 | 41% | 35 | 🧩 PUZZLE | | Mushroom Risotto | $18 | $4.50 | $13.50 | 25% | 25 | 🧩 PUZZLE | | Bruschetta | $9 | $2.50 | $6.50 | 28% | 45 | Monitor | ### MENU MATRIX MAP ``` HIGH MARGIN │ PUZZLES │ STARS Lobster Ravioli │ Margherita Pizza Mushroom Risotto│ Caesar Salad │ Tiramisu ──────────────────┼────────────────── │ DOGS │ PLOWHORSES (none currently)│ Spaghetti Bolognese │ Chicken Parm │ LOW MARGIN ``` ### RECOMMENDATIONS (by impact) **1. Fix Plowhorses — $1,840/month opportunity** - **Spaghetti Bolognese:** Raise from $16 → $18. At 35% food cost, you're losing margin on your 3rd most popular item. A $2 increase on 110 orders = $220/month. Market average for comparable pasta in 78704 is $18-22. - **Chicken Parm:** Raise from $22 → $24 OR reduce food cost. Your cost is $7.80 — switch from breast to thigh (same flavor, $1.50 savings per plate) OR raise price $2. Either way: +$190-285/month. **2. Promote Puzzles — $810/month opportunity** - **Mushroom Risotto ($13.50 margin, only 25/month):** This is a $13.50 margin item that nobody's ordering. Rewrite the description (see below), move it to a featured position, have servers recommend it. If volume doubles to 50/month: +$337/month. - **Lobster Ravioli ($16.50 margin, only 35/month):** Your highest-margin item by dollars. Run it as a weekend special, feature it on social media. If volume goes to 50/month: +$247/month. **3. Raise Appetizer Prices to Market — $225/month opportunity** - Caesar Salad $10 → $12 (market avg is $12-14): +$280/month, ~5% volume drop = net +$225/month - Bruschetta $9 → $11 (market avg is $10-14): +$90/month ### DESCRIPTION REWRITES | Item | Current | Rewrite | |------|---------|---------| | Mushroom Risotto | "Risotto with mushrooms" | "Wild Mushroom Risotto — Arborio rice, slow-stirred with cremini, shiitake, and oyster mushrooms, finished with truffle oil and aged Parmesan" | | Lobster Ravioli | "Ravioli stuffed with lobster in cream sauce" | "Maine Lobster Ravioli — Hand-made pasta filled with sweet lobster, tossed in a saffron cream with roasted cherry tomatoes" | | Chicken Parm | "Breaded chicken with marinara and mozzarella" | "Chicken Parmigiana — Hand-breaded, pan-fried golden, layered with house-made marinara and melted mozzarella. Over al dente spaghetti" | | Spaghetti Bolognese | "Spaghetti with meat sauce" | "Slow-Simmered Bolognese — House-made ragù, braised 4 hours with San Marzano tomatoes, served over fresh spaghetti" | ### WHAT-IF SCENARIO ``` IF you implement all 3 recommendations: Current estimated monthly food profit: $7,650 After changes: $10,525 Monthly increase: +$2,875 Annual increase: +$34,500 Assumptions: 5% volume drop on price increases, 2x volume on promoted puzzles, no menu cuts needed (no dogs identified) ``` ### HANDOFF SUMMARY - **For Local Buzz:** Promote Mushroom Risotto and Lobster Ravioli — highest margin items that need volume. Feature in Instagram posts and weekly specials. - **For Catering Builder:** Pasta dishes (Bolognese, Margherita) are your highest-volume items and scale well for catering. Build catering packages around these. - **For Review Engine:** If guests mention specific dishes in reviews, respond naming the dish + "glad you loved it." Reinforces menu items in review SEO. --- ## OUTPUT FORMAT 1. **Item Analysis** — table with price, cost, margin, food cost %, volume, classification 2. **Menu Matrix Map** — visual quadrant showing where each item sits 3. **Recommendations** — ranked by monthly dollar impact 4. **Description Rewrites** — before/after for items that need better copy 5. **What-If Scenario** — projected impact of implementing all recommendations 6. **Handoff Summary** — data for downstream skills --- ## CONSTRAINTS - [ ] Never recommend cutting a Star item - [ ] Never recommend price increases on Puzzle items (they need volume, not higher prices) - [ ] Always show contribution margin in dollars, not just percentage - [ ] Never rewrite descriptions with HARD NO words from brand voice - [ ] Food cost estimates are benchmarks — always flag as "estimated" without real cost data - [ ] Price increase recommendations should never exceed 15% in a single change - [ ] Always provide what-if math with assumptions stated --- ## QUALITY CHECKLIST - [ ] Is every item classified as Star, Plowhorse, Puzzle, or Dog? - [ ] Are recommendations ranked by dollar impact? - [ ] Do description rewrites use sensory language and stay under 2 lines? - [ ] Is there a what-if scenario with projected monthly impact? - [ ] Are food costs flagged as estimates if real data wasn't provided? - [ ] Would the owner look at this and know exactly which 3 things to change first?