You are running a restaurant in 2026 where beef is up another 14%, your labor floor is $18–22/hour, third-party delivery is skimming 28–30% off every ticket, and your menu prices were last touched in a panic during the 2024 inflation spike. You know your food cost percentage because your accountant tells you every month. What you don’t know is which of your 47 items is quietly bleeding you — the $24 entrée that sells 300 times a month at a $6 contribution margin while the $19 pasta nobody orders throws off $13. Every operator you compete with is guessing the same way you are: raise everything 5%, hope nobody notices, watch traffic soften. Menu engineering is the last profit lever nobody is pulling with real math, because pulling it used to require a consultant, six weeks, and $12,000.
This is written for restaurant owners, multi-unit operators, and consultants who want to sell menu audits as a service. You should be comfortable exporting a sales report from your POS, opening a spreadsheet, and pasting text into a chat window — that is the technical bar. You do not need to code, use an API, or build anything. Out of scope: full P&L rebuilds, labor scheduling, inventory management systems, franchise-level pricing compliance, and anything requiring integration work with your POS vendor. This is menu-level profitability, not restaurant-wide consulting.
Honest assessment of where AI earns its keep here and where it will burn you: AI is genuinely excellent at classification, pattern-finding across hundreds of line items, rewriting descriptions at volume, and running contribution margin math faster than you can open Excel. It is unreliable at recipe yield calculations when units get mixed (a 12-count case of avocados versus 12 pounds will silently wreck a costing), at reading badly scanned vendor PDFs where a decimal shifts, and at knowing your local market’s price ceiling. Human review is non-negotiable on three things: every recipe cost before it drives a pricing decision, every vendor invoice number the AI extracts from a PDF, and every final price against what your neighborhood will actually bear. The AI proposes. You verify the arithmetic and own the price.
What This Guide Covers
- How to identify which items on your menu are actually making money — and which high-volume “winners” are the ones sinking you
- A classification framework that sorts every item into four buckets, each with a different, specific action
- Getting clean item-level sales data out of Toast, Square, Clover, and Lightspeed without vendor support tickets
- Turning messy food cost inputs — MarginEdge exports, Craftable reports, and unreadable vendor PDFs — into usable numbers
- A repeatable prompt sequence for contribution margin analysis, with side-by-side notes on where Claude and Gemini each outperform the other
- The specific guardrails that keep AI from botching recipe costing and yield math — the failure mode that kills most DIY attempts
- Re-pricing using psychological anchoring, so increases land without triggering guest resistance
- Rewriting item descriptions to lift attachment rate and steer traffic toward your high-margin plates
- Layout and eye-path redesign specs you can hand to a designer or execute yourself in Canva or Figma
- Handling delivery and catering menu variants, where third-party fees quietly erase the margin you just engineered
- A tested pricing model for selling this as a service: what to charge for an audit, what to charge monthly, and what each tier includes
- A cold outreach approach built around a free single-item teardown that gets operators to take the meeting
- Scope language, legal boundaries, and the case math that justifies your fee before the client asks
- Where this goes next: dynamic pricing, agentic re-costing, and what changes when you scale to multi-unit accounts
Instant online access after checkout — your guide is available immediately, in full. No upsell, no course pitch, no membership. You bought the guide; you get the guide.
Reviews
There are no reviews yet.