--- title: "Costing a cup of coffee: where a café's margin actually goes | GoSufra" description: "Why the coffee is the cheapest part of a latte, how to work out a true cost per cup including dial-in waste and packaging, and the number that matters more than drink margin in a café." url: "https://gosufra.com/en/blog/cafe-drink-costing/" language: "en" source: "https://gosufra.com" --- # Costing a cup of coffee: where a café's margin actually goes Why the coffee is the cheapest part of a latte, how to work out a true cost per cup including dial-in waste and packaging, and the number that matters more than drink margin in a café. 2026-09-24 · 9 min read ![Illustration of a coffee cup beside a scale and a cost breakdown](https://gosufra.com/blog/cafe-drink-costing.svg) ## Key points - Work the cost per shot from your real yield, not from the theoretical one — dial-in, purging and stale shots take a slice off every kilo. - On a takeaway drink, the cup, lid and sleeve often cost more than the coffee inside them. - Milk waste hides in the pitcher. Steaming more than you pour, every single drink, is a leak nobody logs. - In a café, revenue per seat-hour tells you more than drink margin. A high-margin latte on a table occupied for three hours is not a good sale. Ask a café owner what a cappuccino costs to make and you will usually get a shrug and a small number. Coffee is cheap, milk is cheap, the margin is obviously enormous — and yet the month closes tight, again. The shrug is the problem. Drinks feel too small to cost properly, so nobody costs them, and a business selling four hundred small things a day is run on guesswork. ## Start with the shot, and be honest about yield Everything downstream depends on one number: what a shot of espresso actually costs you. The theoretical version is easy. A kilo of beans, at an 18 g dose for a double, gives you: ``` 1000 g ÷ 18 g = 55 doubles per kilo ``` The real version is lower, and the gap is the point: - **Dial-in.** Every morning, and again whenever the grind moves or a new bag is opened, your barista pulls shots to calibrate and throws them away. Three to six doubles a day is normal. - **Purging.** Grinders retain grounds. Staff purge before the first drink of a session. - **Stale and remade shots.** A shot that sat too long, a drink returned, a mistake during a rush. If you lose five doubles a day and you use two kilos a day, you have lost roughly 5% of your coffee before a single guest was served. Build that into the cost: ``` effective yield = 55 × 0.95 ≈ 52 doubles per kilo cost per double = bean cost per kilo ÷ 52 ``` The correct move is not to *estimate* this. Weigh your waste for one week. Most cafés discover their effective yield is worse than they assumed, and the fix is often a grinder setting rather than a supplier change. ## Milk is where the quiet money goes A latte takes a certain volume of milk. A barista steams more than that, because a pitcher needs a minimum depth to texture properly, and the surplus goes down the drain. Steam 250 ml to pour 180 ml and you waste 70 ml. That is 28% of the milk in every single milk drink. Across a day of three hundred drinks, that is a crate of milk you bought, paid to chill, and poured away. This is not fixed by scolding anyone. It is fixed by having the right pitcher sizes for your common drink sizes, and by a rule about pouring leftover steamed milk away rather than re-steaming it. But you cannot manage it until it is in the cost, and it almost never is. ## The packaging shock Here is the part that changes how people think about takeaway. | Line | Share of a takeaway latte’s cost | | --- | --- | | Coffee | Small | | Milk | Moderate | | Cup, lid, sleeve, carrier | Often the largest single line | | Sugar, napkin, stirrer | Small but not zero | On a takeaway drink, the container frequently costs more than what goes inside it. That single fact should influence your pricing, your decision on whether takeaway and dine-in are priced the same, and whether that “free” second napkin-and-stirrer bundle handed to every customer is a habit worth keeping. Cafés that price takeaway identically to dine-in are often selling their highest-volume format at their lowest margin without knowing it. ## The lines that never make it onto a recipe A drink recipe usually stops at coffee and milk. Real cost does not. - **Syrups and sauces.** Measured in pumps, and the pump is the only thing standing between you and a free-pour habit that doubles the cost. - **Whipped cream, toppings, chocolate dust.** Small per drink, significant per month. - **The complimentary glass of water** served with every coffee. It has a glass, a wash cycle and a breakage rate. - **Water filtration.** Filters are consumables with a replacement schedule. Skip them and you pay in machine repairs instead. - **Machine maintenance and grinder burrs.** Not a per-cup cost, but a monthly one that belongs in your beverage margin, not hidden in “repairs”. None of these individually justify a meeting. Together they are the difference between the margin you think you have and the one your P&L reports. ## A note on shisha, where it applies If your café serves shisha, cost it separately and properly. It behaves nothing like coffee: the consumables are tobacco and charcoal, but the real cost is **labour and turnover**. A shisha needs tending, coals need changing, and the guest occupies a seat for a long time. Cost it as tobacco + charcoal + the staff minutes it consumes, then compare the result against what that seat would have earned serving drinks for the same duration. Many owners find shisha is either their best product or their worst, and very few know which. ## The number that matters more than drink margin Here is the thing food-cost thinking misses about cafés. A restaurant sells a plate and reclaims the table in an hour. A café sells a small item to someone who may stay for three. Margin per drink is almost irrelevant if the seat is occupied all afternoon by a single purchase. The metric that actually runs a café is **revenue per seat-hour**: ``` revenue per seat-hour = daily revenue ÷ (seats × open hours) ``` Track it, and different decisions become obvious: - A laptop-and-wifi crowd is not a problem if they buy something every ninety minutes. It is a problem if they do not. - A pastry case at the till is worth more than a discount on coffee, because it raises the value of an occupied seat rather than lowering it. - Peak-hour queues are a seat-hour problem, not a staffing problem, when people are waiting for tables rather than for drinks. - Two-seat tables beat four-seat tables in a café where most groups are pairs, because the fourth chair earns nothing all day. Look at your drink margin and your revenue per seat-hour together. One tells you whether the product makes money; the other tells you whether the room does. ## Pricing: cost tells you the floor, not the price Costing a drink does not set its price. Nobody pays for a flat white based on your bean cost; they pay based on the neighbourhood, the seat, the wifi, the speed and the alternatives on the same street. What cost does is give you a **floor** and a **comparison**. It tells you: - Which drinks on your menu are quietly unprofitable, usually the elaborate ones with four ingredients and three minutes of labour. - Which drinks carry the business, usually black coffee and simple milk drinks at volume. - Where a small price change has a large effect, because it lands on your highest-volume line. - Which “signature” drink is a marketing item rather than a profit item — worth keeping, but worth knowing. Sort your drinks by popularity and margin and the picture appears quickly. Our guide to menu engineering walks through the four boxes that fall out of that exercise. ## Making the numbers appear by themselves Doing all of this on paper once is useful. Doing it every month by hand is not going to happen, and it should not have to. When a drink has a recipe attached, selling it deducts the beans, the milk and the cup from stock automatically, and a weekly count tells you whether your theoretical cost and your actual cost still agree. That gap — between what your recipes say you used and what your shelves say you used — is the dial-in waste, the over-steamed milk and the untracked syrup, appearing as a number instead of as a feeling. That is what GoSufra’s recipe and inventory link is for: not a report you read once, but a variance you can watch move week to week. ## What to do this week 1. **Weigh your waste for seven days.** Dial-in shots, poured-away milk, remade drinks. 2. **Recalculate your cost per double** using effective yield, not theoretical. 3. **Cost your takeaway packaging** as a line, and compare it to the coffee. 4. **Write recipes for your top ten drinks**, including cup, lid and syrup pumps. 5. **Calculate revenue per seat-hour** for a normal week. 6. **Look at the two lowest-margin drinks** on your menu and decide: reprice, resize, or remove. None of this takes a consultant. It takes a scale, a week, and the willingness to find out that the cheapest thing in the cup was the coffee all along. - coffee cost - café - drink pricing - beverage cost ## In the product ### [Recipes & Plate Costing](https://gosufra.com/en/features/recipe-costing/) Explore ### [Inventory & Purchasing](https://gosufra.com/en/features/inventory-management/) Explore ## Run all of this from one system POS, kitchen, inventory, recipe costing, staff and accounting — connected, and free to start. [Create your free account](https://app.gosufra.com/register?lang=en) ## Related articles ![Illustration of a plate divided into recipe components with a price tag](https://gosufra.com/blog/plate-costing.svg) Costing 2026-06-01 · 8 min read ### [Plate costing: work it out from the recipe, not from a hunch](https://gosufra.com/en/blog/plate-costing/) A practical guide to costing every dish on your menu — from unit price to yield and waste, ending in a sell price that actually protects your margin. Read article ![Illustration of a percentage ring beside report lines](https://gosufra.com/blog/food-cost-percentage.svg) Costing 2026-06-15 · 7 min read ### [Food cost percentage: the one number that reads your health](https://gosufra.com/en/blog/food-cost-percentage/) How to calculate food cost percentage properly from stocktakes rather than invoices, what separates theoretical from actual, and what to do when the gap widens. Read article ![Illustration of a four-box matrix with one quadrant highlighted](https://gosufra.com/blog/menu-engineering.svg) Menu 2026-07-06 · 8 min read ### [Menu engineering: four boxes that decide your profit](https://gosufra.com/en/blog/menu-engineering/) How to sort your menu by popularity and margin into stars, puzzles, plowhorses and dogs, what to do with each, and how the menu layout itself changes what people order. Read article [← All articles](https://gosufra.com/en/blog/) [Contact us](https://t.me/GoSufraSupport)