How Many Drinks Per Hour Can a Commercial Espresso Machine Make?
Coffee Business Resource Center
“How many drinks per hour?” is a reasonable equipment question. It needs a more useful answer than one large number on a sales sheet. The result depends on what counts as a drink, how it is made, and what else the people using the machine must do.
For a café owner, capacity means keeping up with the intended demand while maintaining drink quality and completing normal operating tasks. Work backward from that requirement when comparing machines.
Separate brewing capacity from complete drink service
An espresso extraction is one part of a milk drink. Someone still has to grind and prepare the coffee, handle milk, steam, finish the cup, identify it, and keep the station ready for the next order.
Some tasks overlap. Others compete for the same person or working position. A second group creates another brewing position, but it does not create another barista or automatically increase steam performance in the same proportion.
Also define the recipe. A double-spout portafilter does not automatically mean two full menu drinks from every extraction. The dose and yield assigned to each drink determine what that extraction supplies.
Read the exact configuration
Compare group count, boiler arrangement, heating input, controls, cup clearance, and the installation requirements. Ask which options are included in the quoted model rather than assuming every feature described for the product family is present.
La Marzocco's Linea PB page, for example, describes different controls and configurations, including ABR functionality. Those options affect the operator's work, but the existence of an automatic stop does not establish a universal finished-drink rate. Linea PB manufacturer information
Confirm that the proposed electrical and water supplies support the actual machine configuration. A demonstration on one installation does not establish identical performance on a different supply or mobile power arrangement.
Give the supplier a useful test brief
| Test input | What to specify |
|---|---|
| Drink mix | The share of straight espresso, milk drinks, iced drinks, and other machine uses |
| Recipes and sizes | Coffee dose and yield, milk routine, and the actual cup range |
| Staff | Number of people and which tasks each person performs |
| Grinder | Exact grinder, coffee, preparation method, and normal adjustments |
| Utilities | The electrical, water, treatment, and drainage arrangement |
| Duration | Enough operation to observe recovery, replenishment, and routine cleaning |
| Acceptance | Your drink-quality standard and the definition of a completed drink |
Ask for a demonstration that resembles the business. A string of identical espresso extractions can answer a narrow brewing question while leaving the milk-drink service question unresolved.
Understand why a cycle-time calculation is incomplete
Suppose, purely for illustration, someone assumes a group can repeat a complete cycle every 45 seconds. Dividing 3,600 seconds by 45 gives 80 cycles per hour for that assumed pattern.
That arithmetic does not prove 80 finished drinks per hour. It depends on the assumed cycle including all the necessary work, the recipe's relationship between cycles and drinks, and the rest of the station keeping pace. Doubling the groups does not remove those assumptions.
Use such a calculation to understand a proposed claim and ask questions about it. Do not present it to a customer as tested capacity.
Test sustained service, not just the first few orders
Start with equipment prepared according to its instructions and a dialed-in recipe. Run a representative sequence with the staff and supplies you intend to use. Record accepted finished drinks, remakes, pauses, and the reason for each pause.
Observe brewing and steaming through the run. If steam performance changes, document the conditions and discuss them with the supplier or technician. Avoid guessing that every slow milk drink proves the steam boiler is too small; technique, pitcher choice, access, cleaning, or an equipment fault can also affect the work.
Include restocking and the normal cleaning routine. A demonstration that stops before those tasks occur may overstate the rate you can maintain in the shop.
Our commercial grinder guide explains why the grinder and its duty requirements belong in this assessment too.
Keep the observed result tied to its conditions
If a hypothetical test completes 52 accepted drinks in 30 minutes, its observed equivalent rate is 104 per hour. Label it as that test's result, with the menu and conditions attached. It is not a promise that every hour will look the same.
Repeat relevant tests and compare them with actual service records when available. Look at the slower intervals and the longest waits as well as the average. A setup that only keeps up under its easiest conditions leaves little flexibility for real customers.
Do not apply an arbitrary universal allowance to turn a test into a guarantee. Base the service promise on the variability you observed and the consequences of a delay in your business.
Match the machine to peak demand
Compare the tested service with the busiest relevant interval, including orders from every channel. A shop's daily cup total can conceal a concentrated morning rush.
If the machine is a demonstrated constraint, consider a different configuration or additional capacity. If it is often idle while orders wait, investigate the workflow first. Moving the delay from the espresso machine to the grinder or handoff counter will not necessarily improve the customer's experience.
For a cart or small kiosk, also check whether the installation and working space can support the proposed upgrade. Larger equipment only helps when the complete station can use it effectively.
Sources and further reading
- La Marzocco USA: Linea PB configurations and specifications.
- Espresso Outlet: Commercial grinder duty cycle and workflow.
- Espresso Outlet: Two-group machine capacity and workflow.
All cycle-time and service-rate numbers in this guide are illustrative calculations, not manufacturer throughput ratings.