Guides
How to rank coffee shop menu products for profitability by contribution per bar second
Profitable coffee shop products depend on your own numbers. This method shows contribution per bar second, the arithmetic, and how to rank your menu with it.
What to take away
- No product is profitable in the abstract. Profit is contribution per unit of the thing you have least of, and during the morning that is bar time.
- An item can have the best margin on the price list and still be the worst thing to sell at eight in the morning.
- The ranking changes by hour. Run it twice, once for the peak and once for the quiet stretch, and expect two different answers.
- Waste belongs in the calculation. A pastry ordered yesterday and thrown out today was a decision, not an accident.
Why no ranked product list appears here
A published ranking of profitable coffee shop products would have to assume your ingredient costs, your prices, your equipment, your staffing and your demand curve. Change any one and the order changes. Worse, most such lists rank by gross margin percentage, which is the measure least connected to what a busy shop actually earns in a day.
What transfers between shops is the method. So this page gives you the calculation and the procedure for running it, and you supply the numbers from your own records.
Contribution per bar second
Two measures, taken together, tell you almost everything.
Contribution per bar second
- Rprice customer pays
- Iingredients, packaging, waste
- Tbar seconds per unit
- (R minus I) / Trank by this
Let R be the price a customer pays for the item. Let I be everything consumed to produce one unit: coffee and milk, syrup, packaging, and the share of any batch thrown away rather than sold.
Let T be the seconds of bar time it consumes, measured with a stopwatch on a busy shift rather than a quiet one.
Contribution is R minus I.
Contribution per bar second is (R minus I) divided by T.
Take two typical drinks. A 12 oz latte: R is 5.00, I is 1.20, contribution is 3.80. T is 40 seconds, so the ratio is 0.095 per bar second. A 16 oz drip coffee: R is 3.00, I is 0.45, contribution is 2.55. T is 12 seconds, so the ratio is 0.2125 per bar second.
The latte earns more per cup. The drip earns more per second, so in the peak it goes first, while the latte still earns its keep when the bar is quiet.
Rank your menu by the second measure, not the first. An item with a smaller contribution that clears the bar in a third of the time earns more during a block that is time limited, because you can sell three of them.
Two cautions on I. Include packaging, which is easy to forget and scales with every unit. And include waste at the item level: if a batch is brewed and part of it is discarded, the discarded share belongs to the units that were sold.
Two rankings, not one
Run the calculation twice.
Peak versus quiet ranking
Peak
- Scarce resource
- bar time
- Rank by
- contribution per second
- Board action
- promote fast items
- Route
- pre order
Quiet stretch
- Scarce resource
- demand
- Rank by
- plain contribution
- Board action
- promote slow items
- Route
- sit down
During the peak, bar time is the scarce resource, so rank by contribution per bar second. This ranking tells you what to promote on the board, what to make visibly fast, and which items to move to a pre order route.
During the quiet stretch, bar time is not scarce at all. Rank by plain contribution instead, because the constraint has moved to demand. An item that is too slow for the morning may be exactly right in the afternoon, when a slow drink also gives somebody a reason to sit down.
Where the numbers come from
- Rcomes off the price list, after any discount you routinely give. Use the actual average, not the sticker.
- Icomes from a weighing session. Weigh the dose, weigh the milk actually poured rather than the milk intended, and count the packaging.
- Tcomes from a stopwatch on the busiest hour, timed from ticket to handoff, repeated enough times to see the spread. Typical figures from stopwatch work: about 10 to 15 seconds for drip, 30 to 45 seconds for a milk drink, and longer when two drinks share one ticket.
- Waste comes from a written count at close, not from memory. The discipline for that sits inside how a shop tracks what it holds and what it loses.
What to do with the ranking
- Take the top three items by contribution per bar second and make them easy to choose. Position on the board, staff suggestion, and a name a stranger understands.
- Take the bottom three that also sell well and ask what makes them slow. Often it is one step: a syrup stored out of reach, a second pitcher rinse, a lid that does not seat first time.
- Take anything with unique ingredients and low volume and decide whether it stays. Each one carries ordering, storage and rotation costs that never appear beside it on the price list.
- Look at what the top items load. If all three sit on the steam wand, the ranking is telling you to buy capacity rather than to promote harder. The station map behind that is in how items sit across the stations of a bar.
Bought in food and the day before problem
Pastry is the clearest case of a decision made in advance. You order against a forecast, it arrives, and whatever does not sell is a loss that was locked in yesterday.
Pastry contribution across the order
- High margin, poor sell throughlower real contribution
- Lower margin, clearshigher real contribution
That means the item's real contribution is not the margin on a sold unit. It is contribution across the whole order, including the units that go in the bin. Calculate it that way and the picture changes: a high margin pastry with poor sell through can earn less than a lower margin one that clears.
Order size is the lever, not price, so track sell-through by day of week for a month and order to the pattern rather than to the hope. The gap between the two is where the loss lives.
Order size also affects the labor plan, because a display that empties by ten shapes what the bar does for the rest of the day. Staffing is part of how hours are planned against a real demand curve.
What the calculation cannot tell you
It cannot tell you which items bring people in. A shop with a strong reputation for one drink may be carried by an item that ranks badly on both measures, because it is why somebody walks past two other shops.
It also cannot tell you whether the demand is there in the first place. That is a question about your area rather than your menu and belongs with how local demand is read and tested.
And it cannot tell you what you are permitted to sell or how you must describe it. Claims on a board about origin, freshness or preparation must be truthful and supportable.
Disclosure duties for made to order drinks come from your local health department, not from a national page. The paperwork behind what you are licensed to make and sell is a separate exercise from this one.
Ask your state or local health department, which adopts the FDA's questions and answers on food allergens, the FDA Food Code and the menu and vending machine labeling rules in varying forms.
Common questions
Should I just raise the price of the slow items?
Sometimes, and it is worth testing. But a slow item raised in price is still occupying the constraint, so the better first move is usually to find the step that makes it slow and remove it.
Where do labor costs go in this calculation?
Bar time is the labor cost, expressed in the unit that is actually scarce. If you also want a currency figure, multiply T by your fully loaded cost per bar second, but for ranking purposes the seconds alone are enough.
How often should the ranking be redone?
Whenever an input moves: a supplier price change, a recipe change, a new machine, or a shift in your demand curve. Twice a year as a floor, and immediately after anything that changes I.
