What an automated bar really costs, and when it pays for itself

Most articles about bar automation talk about the wow factor. That is not why anyone buys one twice. The second machine gets bought because the first one moved a number on a spreadsheet.

This is a worked example of which numbers move, and by how much. Every assumption is stated so you can replace it with yours - the structure of the calculation matters more than our figures.

The four numbers that decide it

Ignore everything else until you have these:

1. Drinks served per hour at peak. Not average - peak. Bars lose money in the twenty minutes when the queue is longest, not over the evening average.
2. Labour cost per hour of service, fully loaded with charges.
3. Pour cost variance. The gap between the recipe and what actually goes in the glass.
4. Walkaways. Guests who look at the queue and do not order.

Automation moves all four, but not equally, and which one dominates depends entirely on your venue.

Where the money actually comes from

Pour consistency, the boring one that wins

A bartender pouring free-hand is typically 10 to 20% over recipe, and that overpour is invisible because it never appears as a loss - it just shows up as stock disappearing faster than sales explain.

A calibrated machine pours the recipe. If your spirits cost is 25% of the drink price and you were overpouring by 15%, the recovered margin is roughly 3.75% of revenue - straight to the bottom line, with no change to your prices or your menu. On a venue doing 200 000 EUR a year in cocktails, that alone is around 7 500 EUR.

This is the least glamorous benefit and, in our experience, the largest one.

Labour, but not the way people expect

The naive pitch is “replace a bartender”. That is usually wrong, and venues that buy on that promise are disappointed.

What actually happens is that one person stops pouring and starts doing the things that need a person: greeting, upselling, handling the unusual request, keeping the room moving. Throughput per staff member goes up; headcount often stays. The gain is real but it shows up as more revenue per shift, not as a smaller payroll.

The exception is events. A festival bar running 12 hours with a self-service machine genuinely needs fewer people per serving point, and that is where the labour line moves hard.

Throughput and walkaways

A self-service kiosk serving in parallel with a bartender roughly doubles peak capacity for the drinks on the machine. If your peak is queue-limited - and if you have ever watched people leave a queue, it is - that capacity converts directly into sales that were previously walking away.

This is the number most operators underestimate, because walkaways are invisible. They never appear in the till.

A worked example

A single-venue cocktail bar, one machine, six pumps. Assumptions written out:

Assumption	Value
Cocktails sold per week	700
Average price	11 EUR
Spirits/ingredient cost	25% of price
Current overpour	12%
Peak-hour walkaways	5% of potential sales
Software	Pro licence, 99 EUR/month

Annual cocktail revenue: 700 × 11 × 52 = 400 400 EUR.

• Recovered overpour: 400 400 × 25% × 12% ≈ 12 000 EUR/year
• Half the walkaways captured: 400 400 × 5% × 50% × 75% margin ≈ 7 500 EUR/year
• Software cost: 1 188 EUR/year

Net effect before hardware: roughly 18 300 EUR/year.

Against that, the hardware. A machine is a capital purchase that varies enormously with how it is built and finished - a self-built unit and a bespoke installed bar are not in the same bracket. Take your quoted price, divide by the 18 300 above, and you have your payback period in years. For most of the venues we work with, that lands somewhere between a few months and about two years, and the single biggest swing factor is volume, not price.

Run the same arithmetic with your own numbers on the pricing page - the ROI calculator there is the same model with your inputs.

Where automation does not pay

We would rather you not buy the wrong thing:

• Low volume, high-touch cocktail bars. If your proposition is a bartender who talks to you and builds something bespoke, automating the pour removes the product.
• Constantly rotating menus. Every machine ingredient is a bottle occupying a line. If your menu changes weekly and never repeats, you will spend more time reconfiguring than you save.
• Under about 100 cocktails a week. The maths above simply does not reach a sensible payback, whatever the hardware costs.
• No table for it. This sounds trivial. It is the most common practical blocker: machines need a footprint, a power outlet and somewhere for the bottles.

What to measure before you decide

For two weeks, before you talk to anyone selling machines:

1. Count cocktails sold per hour, and record the peak hour separately.
2. Weigh your opening and closing spirit stock, and compare consumption to what sales say it should be. That gap is your overpour.
3. During your busiest hour, have someone count how many people approach the bar and leave without ordering.

Those three measurements tell you more than any vendor calculator, including ours. If they point to automation, the bars and restaurants page covers how a machine fits into an existing service, and the live demo lets you put the guest interface in front of your staff before committing to anything.