Marketing gets people to the table; the menu decides what happens next. Menu engineering — analysing what sells and what earns, then designing the menu to steer orders — routinely shifts 5–10% of orders toward your most profitable dishes. It's the highest-ROI project in this entire blog, because it needs no new customers at all.
Sort every dish into four boxes
Pull two numbers per dish: how often it sells (popularity) and what it earns per plate (contribution margin, not percentage). Four quadrants emerge. Stars (popular + profitable), plough-horses (popular, low margin), puzzles (profitable, ignored) and dogs (neither). Every decision below follows from this one-hour exercise.
Promote the stars, shamelessly
Boxes, borders, a "chef recommends" tag, top-right placement, the best photo on the page — your stars earn the spotlight. These are also the dishes your reels and Google photos should feature, because they're the ones you most want strangers to order first.
Fix the plough-horses quietly
Your crowd-pullers with thin margins: raise the price a notch, trim a costly component, or pair them into a combo with a high-margin drink or dessert. Don't delist them — they're why people come. Just stop them from being why you don't profit.
Give puzzles one honest chance
High-margin dishes nobody orders usually have a description problem or a placement problem. Rename with appetite words, move next to a star, have servers suggest it for two weeks, feature it in one reel. If it still won't move, it becomes a dog.
Kill the dogs
Every unpopular, unprofitable dish costs inventory, prep time and menu attention. Cutting them shortens tickets, reduces waste and makes the kitchen faster. A tighter menu almost always outsells a sprawling one.
Write descriptions that sell
"Paneer tikka" is a label; "charred tandoor paneer, hand-ground mint chutney" is a craving. Use origin, technique and texture words. Drop currency symbols (₹ signs make diners price-focused) and avoid price columns that invite scanning for the cheapest item.
Re-run the numbers quarterly
Costs drift, tastes shift, seasons change. Re-sort the quadrants every three months and after any supplier price change. Menu engineering isn't a one-time redesign; it's a quarterly habit that compounds.
Want a profit-focused menu review?
GrowLift's audit includes a menu-engineering pass — your stars, dogs and quick wins — alongside your Google, Instagram and review channels. Free, and honest.
Book a free marketing audit Pick a time on CalendlyFrequently asked questions
What is menu engineering in simple terms?
Analysing each dish's popularity and per-plate profit, then using menu design — placement, boxes, descriptions, pricing — to sell more of what earns you the most.
Does menu design really change what people order?
Yes — placement, highlighting and appetite-led descriptions reliably shift 5–10% of orders. Diners decide in about 90 seconds; the menu's job is to guide that decision.
How many items should a restaurant menu have?
Fewer than you think. Tight menus order faster, waste less and cook better. If a dish is neither popular nor profitable for two quarters, cut it.
Keep reading: Food photography that sells · Filling weekday tables · Restaurant marketing budgets