LatteMatchaDripChaiMocha
Sales & Costs

Recipe Costing 101: Know What Every Drink Actually Costs You

Parly Team·February 12, 2026·4 min read

The $5.50 latte question

Cost breakdown donut

You charge $5.50 for a latte. What does it cost you to make? If your answer starts with "probably about..." then you are guessing. And guessing on cost means guessing on margin, which means guessing on profitability.

Recipe costing is the practice of calculating the exact ingredient cost for every item on your menu. It sounds tedious, but it is the single most important exercise for understanding your cafe's financial health.

Why estimates fail

Estimated vs actual cost gauge

Most cafe owners estimate drink costs in one of two ways:

The ratio method: "Industry standard says coffee shops should hit 25-30% cost of goods. My latte sells for $5.50, so it probably costs about $1.50."

The problem: this works backward from a target rather than forward from actual costs. If your real cost is $2.10, you would never know.

The gut method: "Espresso beans are expensive, milk is cheap, cups are pennies. It's probably around $1.50."

The problem: this misses the cumulative impact of every ingredient. A latte is not just beans and milk. It is beans, milk, a cup, a lid, a sleeve (for hot), a straw (for iced), a pump of syrup (for flavored), and the electricity and water to pull the shot and steam the milk.

How to cost a recipe

Recipe cost breakdown screen

The process is straightforward. For each drink, list every ingredient and its cost per unit of measure, then multiply by the amount used.

Step 1: Establish unit costs

For every ingredient, calculate the cost per smallest unit of measure you use in recipes.

Unit Cost Breakdown

Ethiopian beans

5 lb bag$65.00

$0.81/oz

Oat milk

6-pack carton (32 oz each)$28.50

$0.15/oz

Whole milk

1 gallon$4.80

$0.04/oz

Simple syrup

House-made, 1 batch$3.20

$0.08/oz

16 oz hot cup

Sleeve of 50$12.00

$0.24/cup

Hot lid

Sleeve of 50$5.50

$0.11/lid

The key detail: use your actual purchase costs, not list prices. If you buy oat milk through Metro Supply Co at a different price than retail, use the Metro Supply Co price.

Step 2: Map ingredients per drink

List every ingredient that goes into a specific drink variant. A hot 16 oz oat milk latte is a different recipe than an iced 16 oz oat milk latte (different cup, different milk quantity, potentially different ice displacement).

Hot Oat Milk Latte (16 oz)

Ethiopian beans

$0.57

20g (0.71 oz) oz @ $0.81

Oat milk

$1.80

12 oz oz @ $0.15

16 oz hot cup

$0.24

1 each @ $0.24

Hot lid

$0.11

1 each @ $0.11

Cup sleeve

$0.05

1 each @ $0.05

Total$2.77

At a $5.50 menu price, that is a 50% cost ratio before tax. If you assumed it was $1.50, you were off by 85%.

Step 3: Account for modifiers

Modifiers change the cost. When a customer swaps whole milk for oat milk, the drink cost changes. When they add an extra shot, it goes up by the cost of 20g of beans.

For accurate costing, you need modifier-level pricing:

  • Milk swap to oat: +$1.40 per drink (oat at $0.15/oz vs whole at $0.04/oz for 12 oz)
  • Extra shot: +$0.57
  • Flavored syrup: +$0.24 (1 oz of house-made syrup)

This is why POS modifier data matters. If 60% of your latte orders swap to oat milk, your average latte cost is significantly higher than the whole milk base recipe.

The modifiers that kill your margins

Modifier P&L table

Some modifiers are margin-positive (customer pays more than the ingredient cost). Others are margin-negative (the ingredient cost exceeds the upcharge, or there is no upcharge at all).

Common margin traps:

Free milk alternatives. If you do not charge for oat milk swaps, every swap costs you an extra $1.40 per drink with zero additional revenue. At 100 oat milk drinks per day, that is $140/day or $4,200/month in unrecovered cost.

Generous "extra" portions. An "extra drizzle" of honey that turns into a heavy pour. A "light ice" request that means more milk. These small adjustments add up when they happen on every other order.

Bundle discounts that ignore COGS. A pastry-and-coffee combo priced at $8 sounds good for the customer. But if the pastry costs $3.50 from your supplier and the coffee costs $2.77, your margin on the bundle is $1.73 (22%). That might be fine, or it might be eating your profitability.

What good margins look like

Target COGS bands by drink type

Target cost ratios vary by drink type:

  • Espresso drinks: 25-35% COGS is healthy
  • Drip coffee: 10-15% COGS (low cost per cup, high margin)
  • Specialty drinks (matcha, chai): 30-40% COGS (premium ingredients, premium pricing)
  • Pastries: 50-60% COGS (typical for outsourced pastries)
  • Blended menu average: 28-33% COGS

If your blended COGS exceeds 35%, either your prices are too low or your costs are too high. Recipe costing tells you exactly which drinks are the problem.

Keeping costs current

Recipe costs are not static. They change every time a supplier adjusts pricing. The difference between a useful cost model and a stale one is update frequency.

Update unit costs when you receive invoices. If your oat milk price went from $28.50 to $30.00 per case, update the unit cost immediately. Every recipe that uses oat milk is now more expensive.

Review full recipe costs monthly. Even if you update individual ingredients, a monthly review catches any items you missed and shows trends.

Compare actual COGS to recipe-predicted COGS. Your recipe says you should spend $X on oat milk this month based on sales. Your invoices say you actually spent $Y. The difference is waste, over-portioning, or recipe inaccuracy. This comparison is one of the most powerful diagnostic tools in cafe operations.

Start with your top 5

You do not need to cost your entire menu in one sitting. Start with your five highest-volume drinks. Calculate their real costs. Compare to what you assumed. Adjust prices or portions if needed.

Then work through the rest of the menu over the next week. Once every drink is costed, you will never look at your menu the same way. Every sale becomes a data point. Every modifier becomes a margin decision. And every pricing conversation is grounded in reality instead of estimates.