If you run a restaurant, you already know margins are razor-thin. The National Restaurant Association puts the average full-service restaurant at a 3-5% net profit margin. Quick-service does a bit better at 6-9%.
Here’s what most owners don’t realize: shifting just 2-3% of your sales from low-margin items to high-margin ones can effectively double your net profit. And the fastest way to figure out which items to push and which to rethink? Let artificial intelligence (AI) do the math for you.
The Framework You’re Probably Not Using
Menu engineering is a well-established discipline that sorts every item on your menu into four categories based on two things: how profitable it is and how often it sells.
– **Stars** — High profit, high popularity. Protect these. Promote these.
– **Plowhorses** — Popular but low profit. Renegotiate ingredients or nudge the price up.
– **Puzzles** — Profitable but nobody orders them. Rename, reposition, or feature them.
– **Dogs** — Low profit, low popularity. Cut them or redesign them completely.
Doing this analysis by hand means pulling point-of-sale (POS) data into spreadsheets, cross-referencing ingredient costs, and spending hours on math. Most owners do it quarterly at best. AI can do it continuously.
What AI Actually Does Here
AI tools process your POS sales data, ingredient cost databases, and seasonal pricing fluctuations all at once. The value isn’t magic — it’s speed. Instead of a quarterly spreadsheet marathon, you get ongoing answers to questions like: “Which items have the highest contribution margin?” and “What’s dragging my food cost percentage up?”
Quick refresher on that metric: Food Cost % = (Cost of Goods Sold ÷ Food Revenue) × 100. Industry benchmarks generally fall in the 28-35% range for both full-service and quick-service restaurants, though exact figures vary by concept and region. (Quick-service restaurants tend to achieve better net margins through lower labor costs and higher volume, not necessarily lower food costs.) An $18 dish with $5.40 in ingredients hits 30% food cost and gives you a $12.60 contribution margin. AI helps you see that across your entire menu instantly.
Tools at Every Budget
**If you’re spending $20/month:** Upload your POS export as a CSV file into ChatGPT Plus or Claude Pro (both around $20/month). Ask it to build a menu engineering matrix, calculate contribution margins, and flag your dogs and plowhorses. It’s surprisingly effective.
**If you already use a POS system:** Check what you’ve got. Toast includes menu performance reports and product mix analysis. Square for Restaurants offers item-level sales reports (free plan available; paid tiers with additional features — check Square’s pricing page for current rates). Lightspeed Restaurant offers food cost tracking features, though pricing and feature availability vary by plan tier — verify directly with Lightspeed what’s included at each level.
**If you want a dedicated platform:** MarketMan integrates with Toast, Square, and Clover for inventory and menu costing — pricing isn’t listed publicly, so contact their sales team for a quote. xtraCHEF by Toast handles invoice processing and recipe costing (included with some Toast plans). ClearCOGS is a newer AI-driven option focused on demand forecasting and food cost, though pricing is custom.
Note: This market is evolving fast. Verify current pricing directly with vendors before committing.
The Actionable Takeaway
This week, export your last 90 days of sales data from your POS system as a CSV. Upload it to ChatGPT Plus or Claude Pro and ask: “Categorize these menu items into Stars, Plowhorses, Puzzles, and Dogs based on sales volume and contribution margin.” You’ll need to add your ingredient costs per item — even rough estimates work for a first pass.
You’ll likely find two or three dogs you should cut and a puzzle or two that deserves a better spot on the menu. That’s where your margin lives.
No consultant required. No dev team. Just your data and 20 minutes.