Menu Engineering for Bars: Stars, Puzzles, Plow Horses, Dogs

The cocktail menu at your bar has 14 drinks on it. Three of them pay the rent. Four lose you money every time a bartender builds one. The rest just sit there. Do you know which is which?

Menu engineering for bars is how you find out. It sorts every drink into one of four buckets based on how much money it makes and how often it sells, then tells you what to do with each bucket. The framework has been in kitchen management textbooks for 40 years. Almost nobody runs it on a drink program, which is exactly why the bars that do get an edge.

Real cocktails, real margin math, a step-by-step you can run this week.

What is menu engineering and where did it come from?

Menu engineering plots every item on your menu against two axes: how popular it is and how much contribution margin it generates. Contribution margin is the dollar profit left after you subtract product cost from menu price. Not a percentage. Dollars.

The framework was formalized in 1982 by Michael L. Kasavana and Donald I. Smith, professors at the Michigan State University School of Hospitality Business, in their book Menu Engineering: A Practical Guide to Menu Analysis. They built on earlier work by Jack Miller and renamed his “Winners, Marginals, and Losers” into the Stars, Plowhorses, Puzzles, and Dogs vocabulary every hospitality textbook uses today. Plenty of blogs credit Kasavana alone. The original book is co-authored.

The core insight: stop obsessing over cost percentages and start managing the dollar amount each item contributes to profit. A drink with a 28% pour cost that contributes $9 in margin beats a drink with an 18% pour cost that contributes $4. Percentages lie. Dollars pay the electric bill.

Why menu engineering matters for a bar program

Every published menu engineering guide is written for kitchens. Search “menu engineering” and you get clam chowder, vegan pasta, and burger upsells. None of that helps the manager figuring out whether their Negroni program is making money.

Bar programs have their own economics. A cocktail has 4 to 8 ingredients, each with its own unit cost and shelf life. Garnish waste distorts drink cost in ways a food line item never has to worry about. BTG wine turns over one bottle at a time, so pricing a glass wrong compounds fast. Draft beer has line loss and foam built into its true cost. Bartenders free-pour, so spec and actual cost drift apart.

Run this on a bar menu with kitchen assumptions and you get wrong answers. Run it properly and you find out which two or three drinks are carrying the whole program.

What are the four quadrants?

Plot every drink on a 2x2 grid. Horizontal axis is popularity, measured as sales mix percentage (the share of total drinks sold this item represents). Vertical axis is contribution margin in dollars. Draw a line across the middle of each axis and you get four quadrants.

Stars: High popularity, high contribution margin. These drinks sell a lot and make a lot per unit. Protect them, feature them, never tinker with them.

Puzzles: Low popularity, high contribution margin. Strong money when someone orders them, but nobody orders them. Your job is to figure out why and fix it.

Plow Horses: High popularity, low contribution margin. They fly off the rail all night, but each one makes you less than it should. Re-cost or re-price.

Dogs: Low popularity, low contribution margin. Nobody orders them and the ones that do make nothing. Usually candidates for removal.

One note on terminology. The “high popularity” threshold Kasavana and Smith proposed is 70% of the expected average for the category. If you have 10 drinks and each would sell equally, that is 10% each. The popularity line is drawn at 7% (70% of 10%), not 10%. Items at or above 7% count as popular. Johnson & Wales’ food service finance chapter has the full math.

How do you calculate contribution margin on a cocktail?

The math is the simplest in the building.

Contribution Margin = Menu Price - Total Beverage Cost

“Total beverage cost” for a cocktail is not one number. It is every ingredient in the spec, plus garnish, plus a realistic waste allowance. You cannot eyeball this. Recipe costing is line by line.

Here is how it plays out on a classic Old Fashioned built with Woodford Reserve.

Ingredient	Spec	Cost
Woodford Reserve, 2 oz (from $35 / 750ml)	2.0 oz	$2.76
Simple syrup, 0.25 oz	0.25 oz	$0.02
Angostura bitters, 2 dashes	2 dashes	$0.04
Orange peel garnish	1	$0.15
Waste and spillage allowance	-	$0.10
Total beverage cost		$3.07
Menu price		$14.00
Contribution margin		$10.93

That Old Fashioned has a drink cost percentage of 21.9%, which is a clean number on its own. But the $10.93 contribution margin is the metric that matters. Run this for every drink and you have the vertical axis of your matrix.

For popularity, pull 30 days of POS data, get the unit count per drink, and divide by total drinks sold. That gives you sales mix percentage, the horizontal axis. No POS export? A month of tick sheets works fine.

Worked example: a 10-drink bar menu plotted

A neighborhood cocktail bar with a focused list. Here is the menu, the recipe-costed beverage cost, the menu price, the contribution margin, and the 30-day units sold. Total drinks sold that month: 2,000. Weighted average contribution margin across the menu: $8.65. Average popularity per drink: 10%. Popularity threshold at 70% of average: 7%.

#	Drink	Bev Cost	Price	CM	Units	Mix %	Pop?	CM?	Category
1	Vodka Soda (Tito’s)	$1.18	$10	$8.82	360	18.0%	Yes	Yes	Star
2	House Margarita	$2.05	$12	$9.95	320	16.0%	Yes	Yes	Star
3	Old Fashioned (Woodford)	$3.07	$14	$10.93	240	12.0%	Yes	Yes	Star
4	Domestic Draft (16 oz)	$1.25	$6	$4.75	380	19.0%	Yes	No	Plow Horse
5	House Chardonnay BTG (5 oz)	$2.20	$9	$6.80	180	9.0%	Yes	No	Plow Horse
6	Mezcal Negroni (Del Maguey)	$3.90	$16	$12.10	100	5.0%	No	Yes	Puzzle
7	Barrel-Aged Manhattan	$3.45	$17	$13.55	80	4.0%	No	Yes	Puzzle
8	Aperol Spritz	$2.60	$11	$8.40	220	11.0%	Yes	No	Plow Horse
9	Espresso Martini	$2.95	$13	$10.05	100	5.0%	No	Yes	Puzzle
10	Amaro of the Week	$2.80	$10	$7.20	20	1.0%	No	No	Dog

Popularity threshold: 7%. Margin threshold: $8.65. Anything at or above either line clears it.

Three Stars. Vodka Soda, House Margarita, and Old Fashioned. All three clear both lines. The Old Fashioned is the quiet workhorse, lowest unit volume of the three, but every one drops almost $11 into the till. Defend these with your life.

Three Plow Horses. Domestic Draft, House Chardonnay BTG, and Aperol Spritz. They print volume but margin per drink sits below average. The domestic pint at $4.75 of contribution is ugly next to the Old Fashioned at $10.93. A guest who orders two domestics instead of one cocktail is costing you real money.

Three Puzzles. Mezcal Negroni, Barrel-Aged Manhattan, and Espresso Martini. Hidden winners. The Barrel-Aged Manhattan throws off $13.55 per drink, the best number on the menu, and only sells 80 a month. Double it to 160 and you add about $1,080 to monthly gross profit without expanding the menu.

One Dog. The Amaro of the Week. 20 orders a month at $7.20 of margin, $144 of monthly contribution. That slot is worth more to something else.

What do you do with a Star?

Protect them. Feature them. Do not touch them.

The instinct when a drink is selling and making money is to raise the price. Sometimes that works. Sometimes it tips the drink out of its quadrant. Before you touch a Star, check three things. Is the spec consistent shift to shift? Is it placed upper-right or center on the physical menu, where eyes land first? Is the bartender building it in under 90 seconds so it does not clog the rail?

If all three are yes, leave it alone. If any are no, fix that first. Star margin math works because volume is already high. Protect the volume.

One price test that does work: raise the price by $1 for two weeks and watch volume. If it holds, keep the price. Drop more than 10%, roll it back. Most top Stars absorb a $1 raise without flinching.

How do you fix a Plow Horse?

Three moves: re-cost, re-price, or re-engineer.

Re-cost. Audit the recipe. A Plow Horse margin erodes because the ingredient mix has quietly gotten more expensive. Maybe the house tequila went up at the last price book. Maybe you switched lime juice suppliers. Pull the last three invoices and rebuild the pour cost per drink. If it has drifted up 15% or more, swap the ingredient or raise the price.

Re-price. The domestic draft in our example is $6 with $4.75 of contribution margin. A $7 draft with the same cost pushes contribution to $5.75, a 21% bump per pint. You will lose some volume. In most neighborhood bars, a $1 raise on domestic costs 5 to 8% of volume and nets 10 to 15% more gross profit. You are ahead.

Re-engineer. Sometimes the answer is a spec change. The House Chardonnay BTG at $6.80 contribution could move to a different wine at the same glass price if a supplier comes in 20% cheaper per bottle. You are not changing the guest experience. You are changing the math behind the glass. BTG pricing discipline matters: margin on wine-by-the-glass is tight enough that a single supplier switch can move the category from Plow Horse to Star.

One move you should almost never make on a Plow Horse: deletion. These drinks drive traffic and fill cover counts. Kill one and the guest usually orders nothing at all, not the Star you hoped they would upgrade to.

What about Puzzles?

Puzzles are the most interesting quadrant because a small intervention can generate big dollars. Margin is already strong. All you have to do is raise the popularity.

Start with menu layout. Puzzles usually fail because they are buried. Move the Barrel-Aged Manhattan out of the “Whiskey” section into a “Barkeep’s Picks” spot near the top. Put the Mezcal Negroni next to the regular Negroni so the guest who knows a Negroni sees the upgrade. People order what they see first.

Next, staff education. Every bartender should pitch each Puzzle in one sentence. “If you like a Manhattan, our barrel-aged one is rested in a rye cask for 30 days, two bucks more, best one in the neighborhood.” That line, said three times a shift by three bartenders, moves a Puzzle toward Star status in about 60 days. No new ingredients, no menu reprint.

Third is description. Does the Puzzle have two words next to it (“Rye, vermouth, bitters”) or a 15-word description that tells a story? Puzzles need words. Stars sell themselves.

One Puzzle you probably should delete: the one that is a Puzzle because the spec is broken. If the Espresso Martini is low volume because it takes 4 minutes to shake and bartenders actively push people toward easier drinks, the problem is the build. Fix it or accept that the drink does not work in your bar.

When should you kill a Dog?

Most of the time. Not always.

The Amaro of the Week contributes $144 per month. Replace it with a second margarita variant at the House Margarita’s economics ($9.95 CM, 16% sales mix). If that new drink cannibalizes even 5% of the original Margarita’s volume while adding its own 5% fresh sales, the slot goes from $144 to closer to $1,600 a month.

Keep a Dog if:

1. It is a signature. The drink named after the bar, on the menu since opening, on every regular’s Instagram. Brand equity has dollar value even when per-unit margin does not.
2. It moves a slow-moving bottle. Bought a case of Fernet Branca two years ago and the only thing that pours it is the Dog? Kill it and the bottle becomes dead stock. The loss on dead inventory can exceed the drink’s lifetime contribution.
3. It is a loss leader with a purpose. A $3 beer-and-a-shot special at 6pm on Tuesday is a Dog on the matrix and the reason the bar is full by 7. The drink is not the product. Door traffic is.

If your Dog fails all three tests, kill it. Rotate it off at the next menu reprint and put something with real gross margin potential in its slot.

How often should you redo the analysis?

Quarterly at minimum. Monthly is better for high-volume rooms. Any time a distributor raises prices on a core ingredient (2 to 3 times a year right now), re-run contribution margins on the affected drinks.

Distributor pricing, seasonal shifts, and bartender drift on specs compound in about a 90-day cycle. A drink that was a Star in January can turn into a Plow Horse by April if the house whiskey went up 9% and nobody adjusted the price. Run the matrix once a year and you are making decisions on data that is already 6 months stale.

Pragmatic cadence: run a full recipe cost update every time a new price book drops, then a full popularity analysis at the end of each quarter. Two hours of work every three months. It pays for itself in the first decision you make.

Track menu engineering automatically with PourIQ

Doing this by hand works, which is why we walked through the math. It also eats a full afternoon every quarter, lives in a spreadsheet that breaks when your best bar manager quits, and requires discipline most operators lose by quarter three.

PourIQ’s recipe costing module builds the contribution margin side of the matrix automatically. Every drink is recipe-costed against current invoice prices, updated every time a new price book lands. Pair it with the POS integration for unit sales and the dashboard shows you the four quadrants without a spreadsheet. Plow Horses get flagged the week they drift below the margin line. The matrix runs itself.

To start with the napkin version, the drink cost calculator runs recipe costs in a minute and gives you contribution margin for any single cocktail.

Start with one drink, not the whole menu

The most common reason operators do not do menu engineering is that it feels like a big project. It is not. Pick the one drink on your menu you are most unsure about. Recipe cost it properly. Pull the last 30 days of sales. Calculate the contribution margin. Compare it to your menu average.

One drink, one number, 15 minutes. Do it today. Do the next one tomorrow. By the end of the week you have five drinks sorted. By the end of the month you have the whole matrix. Bars that run on gut feel are losing to bars that run on dollars of contribution margin. The math is not hard. It is just work nobody is doing.

Next step: Book a 15-minute demo to see how PourIQ sorts your cocktail list into Stars, Puzzles, Plow Horses, and Dogs automatically, or run one drink through the pour cost calculator.

The PourIQ Team

Sources

1. Kasavana, M. L. and Smith, D. I. (1982). Menu Engineering: A Practical Guide to Menu Analysis. Hospitality Publications, Lansing, MI.
2. Michigan State University Broad College of Business, Michael L. Kasavana faculty profile
3. Johnson & Wales University, Food Service Financial Systems: Menu Engineering chapter
4. AHLEI, The Power of Menu Engineering, Part One
5. FIU Hospitality Review, menu engineering framework history
6. Craftable, Menu Engineering 101