Tools / Menu Intelligence
Free tool · Menu Intelligence

Menu Engineering Matrix

Every menu has dishes that quietly fund the business and dishes that quietly drain it, and sales reports alone don't tell you which is which. Menu engineering plots each item on two axes, how often it sells and how much margin it makes, and sorts your menu into four boxes: stars, plowhorses, puzzles and dogs. Enter your dishes with units sold, selling price and plate cost, and the matrix does the sorting.

Menu Intelligence — Menu Engineering Matrix
In short

A dish is a Star if both its popularity and its margin are above the menu average, a Plowhorse if popular but low-margin, a Puzzle if high-margin but slow-selling, and a Dog if it is below average on both.

Margin per dish = price − plate cost. Popularity threshold = 70% of the average share (the Kasavana-Smith rule). Margin threshold = the sales-weighted average margin across all dishes.
DishUnits soldPrice (₹)Plate cost (₹)MarginQuadrant
₹309.00Star
₹239.00Plowhorse
₹429.00Puzzle
₹184.00Dog
Total units sold
675
Avg. margin (weighted)
₹280.59
Popularity threshold
17.5%
Stars
Butter Chicken (47.4%)

High margin, high popularity. Protect and feature them, never discount.

Plowhorses
Dal Makhani (41.5%)

Popular but thin margin. Nudge the price up or re-engineer the cost.

Puzzles
Lamb Shank (5.9%)

Profitable but slow. Reposition, rename, or have staff recommend them.

Dogs
Veg Manchurian (5.2%)

Below average on both. Strong candidates to cut from the menu.

How to use the Menu Engineering Matrix

  1. Add a row for each dish with units sold in the period, selling price and plate cost.
  2. The tool computes each dish's margin and share of total sales.
  3. Read the four quadrants: Stars, Plowhorses, Puzzles and Dogs, with a suggested action for each.

Frequently asked questions

What do I actually do with each quadrant?

Stars: protect them, feature them, never discount them. Plowhorses: popular but thin, raise the price slightly or re-engineer the recipe cost. Puzzles: profitable but ignored, reposition them on the menu, rename them, or have staff recommend them. Dogs: cut them, or keep one or two only if they serve a purpose like a kids option.

What period of sales data should I use?

A full month is the practical minimum, it smooths out weekday and weekend patterns. Avoid festival weeks or unusual periods, they distort popularity. Re-run the matrix each quarter or after any menu or price change.

Why is the popularity cutoff 70% of the average share, not the average itself?

That is the standard Kasavana-Smith convention. Using the raw average would mark nearly half the menu unpopular by construction. The 70% factor gives dishes a fair band around the mean before they get flagged as slow movers.

More Menu Intelligence tools

Want this run for you? Book a free auditExplore the platform →