Coffee Robot ROI Benchmark: Real Payback Data From Three European Deployments
We published the complete eight-month profit-and-loss curve for one Madrid coffee robot — then set it against two more European deployments. One has verified operator data. Two do not, and we label that on the table instead of hiding it. Here is what the numbers actually show about how fast these machines pay for themselves.

Key findings
Figures for the Madrid deployment are reconstructed from operator profit-and-loss records and machine telemetry across eight trading months. Figures for the Türkiye and Bulgaria deployments are the projected values published in those case studies and are not yet verified. See Data status and method.
Why we are publishing this
Almost every coffee robot ROI claim is unverifiable. A vendor quotes a payback period — eight months, twelve months, "under a year" — and nobody shows the monthly curve that produced it. There is no cost stack, no revenue ramp, no month-by-month reconciliation. The number simply appears, usually in a headline, usually without a source.
Buyers have learned to discount those numbers. They are right to. A payback figure without an underlying curve is not a forecast; it is a marketing asset. And in an industry where a single machine costs more than a family car, that is an expensive gap to leave open.
This benchmark is our attempt to close it for our own deployments. We took the one deployment with a complete operator record — a coffee robot in a Madrid shopping centre, deployed in ten weeks and documented in full atour Madrid case study — and published the entire eight-month curve: cups per day, revenue, monthly operating cost, net profit, and cumulative net profit, month by month.
Not the average. Not the projection. The curve.
We then set that against two further European deployments: a sealed-cabinet unit at a shopping mall entrance near İzmit, Türkiye, and a fully branded robot at a Kaufland supermarket entrance in Sofia, Bulgaria — the city's first robotic barista.
Only one of those three deployments currently has verified financial data. The other two are running on the projections published in their own case studies — which those pages label, correctly, as representative figures shown for illustration.
We mark that distinction directly on the comparison table rather than burying it in a footnote. A benchmark you cannot audit is just marketing with a spreadsheet attached — and we would rather publish a smaller number we can defend than a bigger one we cannot.
Data status and method
Before any numbers, the disclosure. This is the section that makes the rest of the article worth reading, so it comes second rather than last.
| Deployment | Data status | What the figures are based on |
|---|---|---|
| Madrid, Spain | Verified | Operator profit-and-loss records for eight trading months, cross-checked against machine telemetry (cup counts, drink mix, uptime). Reconciled by us; not independently audited. |
| İzmit, Türkiye | Projected | Projected figures published in the İzmit case study. Operational details (machine model, placement, install method) are confirmed; the financial figures are not yet verified. |
| Sofia, Bulgaria | Projected | Projected figures published in the Sofia case study. Operational details (client, placement, branding, launch) are confirmed; the financial figures are not yet verified. |
Definitions
ROI comparisons go wrong when everyone means something different by "payback". These are the definitions used throughout this article.
- Payback — the month in which cumulative net operating profit first equals or exceeds total invested capital. It excludes the owner's own time and any financing cost.
- Net operating profit — revenue, minus ingredient cost, minus location commission, minus all fixed monthly operating costs. It is stated before tax and before any loan repayment.
- Gross margin per cup — (selling price − ingredient cost) ÷ selling price. It excludes commission and all fixed costs, which is why it is always a flattering number.
- Run-rate — the volume the machine sustains at the end of the measurement period, after the ramp has flattened.
The cost model behind the Madrid curve
We rebuilt the Madrid monthly curve from the published cost structure and reconciled it against the operator's reported eight-month cumulative net profit. The model uses three inputs:
- Location commission — 18% of gross revenue, with no fixed rent.
- Ingredient cost — averaged 8.7% of revenue across the eight months, equivalent to a 91.3% gross margin. This sits inside the operator's reported 84–93% per-cup range; the weighted average is toward the upper end because the drink mix skews to espresso and Americano, which carry no milk cost.
- Fixed monthly costs — €965, comprising restocking labour €550, maintenance reserve €150, insurance €120, electricity €110 and connectivity €35.
The resulting curve reproduces the operator's reported cumulative net profit of €76,940 to within 0.3%. We state that margin of error rather than presenting the reconstruction as exact.
What is excluded, and why it matters
Three things are deliberately left out of every figure in this benchmark, because including them would make payback look better than it is.
The owner's time. The Madrid operator spends roughly one hour a day on restocking and basic checks. That time is paid for inside the €550 monthly labour line, but a solo operator doing it themselves is not paying themselves a market wage. If you value that hour at €20, add roughly €600 a month to the cost base.
Financing. All three deployments were purchased outright. If you finance €43,450 over three years at 8%, you add roughly €1,360 a month in the first year — which pushes Madrid's payback from 5.4 months to about 8.4.
Tax. Every net profit figure here is pre-tax.
Currency: Madrid figures are stated in euros. The İzmit and Sofia case studies publish in US dollars. We have not applied an exchange-rate conversion anywhere in this article, and we do not present a combined multi-deployment total, because converting operator figures at a spot rate would manufacture a precision the underlying data does not support.
The benchmark
Here is the comparison in full. Read the data-status column first — it changes how much weight the rest of each row deserves.
| Deployment | Machine & placement | Data status | Cups / day | Avg ticket | Monthly revenue | Cumulative net | Payback |
|---|---|---|---|---|---|---|---|
| Madrid, Spain | Open-layout unit, shopping-centre corridor between metro exit and food court | Verified | 185 | €3.50 | €19,425 | €76,940 (8 mo) | 5.4 mo |
| İzmit, Türkiye | Sealed-cabinet unit, shopping-mall entrance; internal water and waste tanks, no plumbing | Projected | 200+ | — | — | — | ~8 mo |
| Sofia, Bulgaria | Fully branded unit, Kaufland supermarket entrance; city's first robotic barista | Projected | 180+ | — | — | — | ~9 mo |
Cells marked "—" are not published as verified figures. We are collecting reconciled profit-and-loss data from both operators and will update this table when it is available; until then we would rather show an empty cell than a modelled one dressed up as measured.
What the table does and does not tell you
Two things are worth drawing out. The first is that the ordering is stable: Madrid pays back fastest, İzmit second, Sofia third. That ordering comes from the cup volumes and payback figures that each case study publishes, so it holds regardless of whether the underlying figures are verified.
The second is that the gap between first and third is smaller than the headline difference suggests — 5.4 months against roughly nine. All three deployments sit in the same broad band. That is itself a finding: the machine is not the variable. A sealed cabinet in a Turkish mall entrance and an open unit in a Spanish shopping centre produce broadly comparable payback periods. What differs is the market, the placement and the deal structure around them — and those are what the rest of this article is about.
Why Madrid pays back fastest
The occupancy deal is the single biggest factor
Madrid runs on a straight 18% revenue commission with no fixed rent. That one clause did more for the payback period than any hardware specification.
In month one, at 87 cups a day, the operator paid €1,134 in commission. A fixed-rent deal at a comparable shopping-centre pitch — and comparable pitches in Madrid run €1,800–2,500 a month — would have cost more than double that, whether the machine sold one cup or two hundred.
That is the difference between a ramp and a crisis. A revenue-share structure means your worst month is also your cheapest month. Fixed rent means your worst month is your most expensive one.
The second reason: a €3.50 average ticket
Madrid's average selling price is €3.50. That is not an accident of the machine; it is a function of what the local market will pay for coffee. In cities where the café price anchor sits closer to €2.00, the same machine, serving the same number of cups, produces roughly 40% less revenue per unit of footfall.
This is why volume alone is a poor predictor of payback. Two hundred cups a day at €2.00 generates less revenue than 120 cups a day at €3.50 — and the cost of the machine is identical in both cases. We come back to this in the sensitivity analysis below, because it is the most useful planning table in the article.
The third reason: 35,000 people a day, already walking past
The Madrid unit sits in a centre reporting 93% commercial occupancy in 2025 and more than 35,000 daily visitors, positioned in the corridor between the metro exit and the food court. The machine does not have to generate traffic. It only has to convert traffic that is already moving in the right direction, at the right time of day.
Deployment took ten weeks from contract to first cup. That timeline matters for payback in a way that is easy to miss: every week between payment and launch is a week of zero revenue against a fixed capital cost.
The fastest ROI lever is not price, and it is not ingredient cost. It is time-to-volume — how quickly the location learns the machine is there.
Moocoo ResearchOperator benchmark, September 2026
Where the €43,450 actually goes
| Capital item | Amount |
|---|---|
| Moocoo coffee robot unit | €38,500 |
| Shipping & customs (Shenzhen → Valencia → Madrid) | €2,200 |
| Installation & on-site technician | €1,500 |
| Initial inventory (beans, milk, cups, syrups) | €800 |
| Electrical minor works | €450 |
| Payment terminal setup | €0 (included) |
| Total initial investment | €43,450 |
The machine is 88.6% of the total. The remaining €4,950 — shipping, customs, installation, initial stock and electrical work — is the part that tends to arrive late in a budget conversation.
Our guidance, based on this deployment and the others in this benchmark, is to budget 11–13% above the quoted machine price to get to a trading unit. Operators who budget only for the machine are typically surprised by the customs invoice and the electrician, in that order.
Two lines deserve specific attention. Shipping and customs on a unit this size is not trivial, and it varies with route and destination country. Electrical work is small in euro terms but it is on the critical path — if the site needs a dedicated circuit, that is a scheduling dependency, not a line item. Both are covered in our deployment process.
The monthly cost stack, and the operating-leverage trap
This is the published monthly operating cost for the Madrid unit, as it stood in the first trading month. Look at which lines scale with volume and which do not — the pattern is the whole story of the payback curve.
| Cost line | Monthly (EUR) | Scales with volume? |
|---|---|---|
| Location commission (18% of revenue) | €1,134 | Yes — with revenue |
| Restocking labour (part-time, 1 hr/day) | €550 | No — fixed |
| Coffee beans (Arabica blend, 18 kg/mo) | €430 | Yes — with cups |
| Milk & dairy alternatives | €320 | Yes — with cups |
| Cups, lids, napkins, stirrers | €180 | Yes — with cups |
| Maintenance reserve | €150 | No — fixed |
| Insurance (equipment + liability) | €120 | No — fixed |
| Electricity (dedicated circuit) | €110 | No — fixed |
| Syrups & flavourings | €95 | Yes — with cups |
| Internet (dedicated 4G router) | €35 | No — fixed |
| Total monthly operating cost | €3,124 | €965 of it is fixed |
Published month-one cost structure for the Madrid deployment. At this revenue level the commission line reflects roughly 87 cups a day at a €3.50 average ticket.
The pattern: €965 fixed, and everything else follows the cups
Strip the table back and the structure is simple. Labour, maintenance reserve, insurance, electricity and connectivity total €965 a month and do not move. Commission tracks revenue at 18%. Ingredients track cups, and averaged 8.7% of revenue over the eight months.
That means the operating cost ratio falls as volume rises — but the absolute cost still grows, and the fixed base is brutal at low volume. Here is the same cost structure applied across three points in the ramp:
| Month | Monthly revenue | Operating cost | Cost as % of revenue | Net margin |
|---|---|---|---|---|
| Month 1 | €9,135 | €3,404 | 37.3% | 62.7% |
| Month 4 | €13,650 | €4,610 | 33.8% | 66.2% |
| Month 8 | €19,425 | €6,152 | 31.7% | 68.3% |
Vendors quote net margin at run-rate, because that is where it looks best. At Madrid's month-eight volume the net margin is 68.3% — an excellent number for any retail format, and one that makes the investment case look obvious.
But in month one the same machine returned 62.7%, and that is the month you have to survive on cash flow. The gap between those two figures — 5.6 percentage points — is the entire reason a ramp-up needs to be planned for, and it is invisible in a single-number ROI pitch.
One structural note on the commission
The 18% revenue share is not a small number. Over eight months it consumed roughly €1 in every €5.50 of gross revenue — more than ingredients, more than labour, more than any other single line in the cost stack.
It is still the right deal, for one reason: it converts the location's rent from a fixed obligation into a variable one. On a machine with a €43,450 upfront cost, the thing most likely to kill the business is not a thin margin. It is a fixed monthly bill during a slow first quarter.
The ramp curve
This is the table we have not seen published anywhere else in this category, and it is the one that changes how you should read every payback claim you are given.
| Month | Cups / day | % of run-rate | Monthly revenue | Monthly net profit | Cumulative net profit |
|---|---|---|---|---|---|
| 1 | 87 | 47% | €9,135 | €5,731 | €5,731 |
| 2 | 96 | 52% | €10,080 | €6,424 | €12,155 |
| 3 | 113 | 61% | €11,865 | €7,732 | €19,887 |
| 4 | 130 | 70% | €13,650 | €9,041 | €28,928 |
| 5 | 148 | 80% | €15,540 | €10,426 | €39,354 |
| 6 | 164 | 89% | €17,220 | €11,657 | €51,011 |
| 7 | 180 | 97% | €18,900 | €12,655 | €63,666 |
| 8 | 185 | 100% | €19,425 | €13,274 | €76,940 |
Reconstructed from operator records and machine telemetry; reproduces the operator's reported eight-month cumulative net profit of €76,940 to within 0.3%. Revenue is calculated at a €3.50 average ticket across a 30-day month. The cumulative net profit first passes the €43,450 invested-capital threshold during month six — which is the "5.4 months" quoted throughout this article.
Three phases, three different problems
The curve is not a straight line, and treating it as one is the most common modelling error we see. It breaks into three phases, each with its own constraint.
Discovery
Footfall is already high, but conversion is low. People do not yet know the machine is there, and those who notice it are still deciding whether to trust a robot with their coffee. Growth here is driven by visibility, not by the product.
Habit formation
The steepest part of the curve. Repeat customers appear — the same commuters, the same shift workers, the same shoppers on a Saturday routine. This is the phase where the business actually gets built, and it is the phase a revenue-share deal protects.
Saturation
Growth flattens to roughly 3–5% a month. This is the real run-rate, and it is the number any honest ROI model should be built on — not a projection made before the machine was switched on.
Month one was 47% of the eventual run-rate
That single ratio is worth more than any payback claim. In the first month of trading the Madrid machine sold 87 cups a day against a run-rate that would eventually reach 185. It took eight months to close the gap.
If a vendor quotes you a payback period based on run-rate volume, they are quoting a number that does not exist until month eight — and they are quietly assuming the ramp costs nothing. It does. The next section puts a figure on it.
Steady-state payback vs actual payback
The table below models steady-state payback in months — that is, how long €43,450 takes to return if the machine starts at full run-rate on day one — across a grid of daily volumes and ticket prices. The cost structure is the Madrid one: 18% commission, 8.7% ingredient cost, €965 fixed.
| Cups / day | €2.50 ticket | €3.00 ticket | €3.50 ticket | €4.00 ticket |
|---|---|---|---|---|
| 80 | 12.6 mo | 10.1 mo | 8.4 mo | 7.1 mo |
| 100 | 9.6 mo | 7.7 mo | 6.4 mo | 5.5 mo |
| 120 | 7.7 mo | 6.2 mo | 5.2 mo | 4.5 mo |
| 150 | 6.0 mo | 4.9 mo | 4.1 mo | 3.6 mo |
| 185 (Madrid) | 4.7 mo | 3.9 mo | 3.3 mo | 2.8 mo |
Steady-state payback: months required for cumulative net operating profit to recover €43,450, assuming the machine trades at the stated volume from month one. Pre-tax, excluding financing and the owner's time.
Madrid's steady-state payback is 3.3 months. It actually took 5.4.
Both numbers are correct, and the difference between them is the most important thing in this article.
At 185 cups a day and a €3.50 ticket, the Madrid machine generates €13,274 in monthly net operating profit. Against €43,450 of invested capital, that is a 3.3-month payback — a figure that would look extraordinary in any pitch deck.
The operator actually reached payback in 5.4 months, because the machine did not open at 185 cups a day. It opened at 87.
What the ramp cost
Run the counterfactual. Had the Madrid machine traded at its run-rate of €13,274 net profit from month one, eight months of trading would have produced €106,192 in cumulative net profit. It actually produced €76,940.
The ramp cost the operator €29,252 — roughly two-thirds of the price of the machine itself, given up not to inefficiency but simply to the time it takes a location to notice a new machine.
If the ramp is worth €29,000 on a single machine, then the highest-value decisions in this business are not ingredient sourcing or ticket price. They are the ones that shorten the ramp: picking a site with existing, repeated footfall; negotiating a revenue share instead of fixed rent; and launching fast.
A €0.10 improvement in ingredient cost is worth about €660 a month at Madrid's volume. Getting to run-rate two months earlier is worth roughly €26,000. They are not the same size of decision, and most ROI conversations treat them as if they were.
How to read the sensitivity grid
Three things fall out of the table that are worth holding on to when you evaluate a site.
- Ticket price buys you more than volume. Moving from an 80-cup site at €2.50 to a 120-cup site at €3.50 takes payback from 12.6 months to 5.2. Moving from 80 cups to 185 cups at the same €2.50 ticket only gets you to 4.7 — real, but it requires more than doubling throughput to achieve what a better market achieves for free.
- The low-volume, low-price corner is where projects die. Below roughly 100 cups a day at a €2.50 ticket, payback stretches past nine months. That is outside most operators' tolerance, and it is where the fixed €965 starts to bite.
- Above 150 cups, the returns compress. The difference between 150 and 185 cups at a €3.50 ticket is 4.1 months against 3.3. Worth having, but the marginal gain per cup falls as volume rises — because commission scales with every additional euro of revenue.
What actually goes wrong
A benchmark that only reports successes is a brochure. These are the patterns that turn a defensible payback projection into a machine sitting idle in a corridor.
Footfall without dwell time
A site can have enormous traffic and still fail. Commuter concourses, transit interchanges and building lobbies move people who are already late. Coffee purchase needs a pause — a moment where stopping is not costly. The Madrid site works partly because a shopping-centre corridor contains people who are deliberately not in a hurry.
Fixed rent on a machine that has not ramped
The most expensive mistake in the list. Signing a fixed monthly rent of €1,800–2,500 against a machine that opens at 87 cups a day converts a survivable first quarter into a cash-flow problem. If you cannot negotiate a revenue share, negotiate a rent-free ramp period instead.
A low-ticket market with a high-cost machine
Machine capital cost is broadly fixed across markets. The price the local market will pay for coffee is not. Deploying a €38,500 unit into a market with a €2.00 coffee anchor means paying Madrid's capital cost against roughly 57% of Madrid's revenue per cup — and no amount of operational discipline closes that gap.
The wrong drink mix
Milk-heavy drinks carry the lowest margin. Madrid's mix is deliberately weighted toward espresso and Americano — 40% of volume between them — precisely because those drinks carry no milk cost. An operator whose location demands large lattes and frappés will run a materially lower gross margin than the 84–93% range quoted here.
Underestimating the restocking routine
An hour a day sounds trivial until it is every day, including weekends, for the life of the machine. That hour is the €550 labour line in the Madrid cost stack, and it is a real operational commitment, not a rounding error. It is also the reason remote monitoring matters: a machine that runs out mid-afternoon on a Saturday is a machine that loses its best trading hours.
No telemetry, so downtime goes unnoticed
A failed grinder at 09:00 that nobody notices until Monday has cost you a weekend of peak trading. At Madrid's run-rate, a single lost Saturday is roughly €650 of revenue. The case for cloud monitoring is not convenience; it is that unmonitored downtime is invisible until the month-end report, by which point the revenue is unrecoverable.
What we still do not know
We would rather state the weaknesses of this benchmark than have you find them.
The sample is three, and only one is verified
Three deployments is not a dataset; it is an anecdote with a table around it. One verified case and two projections cannot establish an industry benchmark, and we are not claiming they do. What it can do is show the shape of a payback curve — and the shape is consistent enough, across three different markets and two different machine formats, to be worth publishing.
Eight months is short
The Madrid data covers eight trading months. That is long enough to see the ramp flatten and payback achieved, but not long enough to observe the things that matter at year three: major component failure, changes in the shopping centre's footfall, or a competitor installing a machine two corridors away. Long-run maintenance cost is a reserve in this model, not an observation.
One cost structure, applied broadly
The sensitivity analysis applies Madrid's cost structure — 18% commission, 8.7% ingredients, €965 fixed — to every combination of volume and price. Real deployments differ. Sofia and İzmit may carry different commission rates, different labour costs and different ingredient pricing. Where a deployment's structure differs materially from Madrid's, the grid will overstate or understate its payback.
Operator-reported, not audited
The Madrid figures come from the operator's own records, cross-checked against machine telemetry. They have not been reviewed by an accountant or an auditor. We reconciled them, and our reconstruction lands within 0.3% of the operator's reported cumulative figure, but reconciliation is not audit.
Europe only, and not all of Europe
All three deployments are in Europe. Nothing here should be read as evidence about payback periods in North America, the Gulf or Southeast Asia, where ticket prices, labour costs and site economics differ substantially. We will publish that data when we have it, with the same labels attached.
Because the alternative — another vendor quoting a payback period with no curve behind it — is worse. A benchmark with a stated sample size and a labelled data-status column lets you discount it appropriately. A single confident number does not.
Frequently asked questions
How long does a coffee robot actually take to pay for itself?
In the one deployment in this benchmark with verified data — a Madrid shopping centre — the machine recovered its €43,450 investment in 5.4 months, reaching €76,940 in cumulative net operating profit by the end of month eight.
The other two deployments in this benchmark publish projected payback of roughly eight months (İzmit, Türkiye) and roughly nine months (Sofia, Bulgaria). Those figures are not yet verified.
Across the three, a realistic planning range for a well-placed European deployment is five to nine months, pre-tax, excluding financing and the owner's time.
Why is the projected payback different from the actual payback?
Because projections are usually calculated at run-rate volume — the throughput the machine reaches once the location has adjusted to it — while actual payback has to absorb the ramp.
At Madrid's run-rate of 185 cups a day, steady-state payback would be 3.3 months. The machine actually opened at 87 cups a day and took 5.4 months. The two-month difference is the ramp, and it cost the operator roughly €29,252 in foregone profit over eight months.
Any payback figure you are quoted should be checked against the volume assumption behind it. If that assumption is run-rate, the real number is longer.
What is the biggest ongoing cost of running a coffee robot?
In the Madrid deployment the largest single monthly line is the location commission at 18% of revenue. Across the full cost stack it consumed roughly €1 of every €5.50 of gross revenue — more than ingredients, more than labour, more than any other line.
Total monthly operating cost in the first trading month was €3,124, of which €965 is fixed (labour, maintenance reserve, insurance, electricity, connectivity). The rest scales with sales. By month eight the operating cost had risen to €6,152 but fallen as a share of revenue, from 37.3% to 31.7%.
Is a revenue-share deal better than paying fixed rent?
For a machine that has not yet ramped, yes — decisively. Madrid runs on 18% of revenue with no fixed rent. In month one that cost the operator €1,134, against an estimated €1,800–2,500 for a fixed-rent pitch of comparable quality in the same city.
The advantage is not that a revenue share is cheaper. At run-rate, 18% of €19,425 is more than a typical fixed rent. The advantage is that a revenue share makes your worst month your cheapest month, which is exactly what a business with a large upfront capital cost and a slow first quarter needs.
If a revenue share is not available, negotiate a rent-free ramp period instead. The economics of the first three months are what decide whether a deployment survives.
How many cups a day does a coffee robot need to sell to be profitable?
Profitability arrives well before payback. At Madrid's cost structure, the machine covers its €965 of fixed monthly costs plus commission and ingredients at roughly 25–30 cups a day. Everything above that is contributing to the return of capital.
Payback is the harder threshold. At a €3.50 average ticket, recovering €43,450 takes about 8.4 months at 80 cups a day, 6.4 months at 100, 5.2 months at 120 and 3.3 months at 185 — assuming steady-state trading from month one.
Below roughly 100 cups a day at a €2.50 ticket, payback stretches beyond nine months, which is outside most operators' tolerance.
How long does deployment take from order to first cup?
The Madrid deployment took ten weeks from contract to first trading day, covering manufacture, shipping and customs from Shenzhen via Valencia, installation and commissioning, branding, and calibration of the drink menu.
Timeline matters directly to ROI: every week between payment and launch is a week of zero revenue against committed capital. Site assessment and lease negotiation should run in parallel with production, not after it. Our deployment process covers the sequence in detail.
Data appendix
| Item | Value / source |
|---|---|
| Madrid deployment period | Eight trading months, ending September 2026 |
| Madrid data source | Operator profit-and-loss records; machine telemetry (cup counts, drink mix, uptime) |
| Madrid capital investment | €43,450 |
| Madrid average ticket | €3.50 |
| Madrid gross margin per cup | 84–93% |
| Madrid drink mix | Café con leche 42%, espresso 22%, Americano 18%, Café Bombón 12%, specialty 6% |
| Location commission | 18% of revenue, no fixed rent |
| Ingredient cost (modelled) | 8.7% of revenue |
| Fixed monthly cost (modelled) | €965 |
| Month-1 operating cost | €3,124 |
| Month-8 revenue | €19,425 |
| Cumulative net profit, 8 months | €76,940 |
| Modelled cumulative net profit | €76,940 (reconstruction within 0.3%) |
| Payback | 5.4 months (during month six) |
| Deployment timeline | 10 weeks, contract to first cup |
| Site footfall | 35,000+ daily visitors; 93% centre occupancy (2025) |
| Türkiye deployment | Sealed-cabinet unit, mall entrance, İzmit, Kocaeli — Projected figures |
| Bulgaria deployment | Fully branded unit, Kaufland supermarket entrance, Sofia — Projected figures |
All figures are pre-tax and exclude financing costs and the owner's time. Madrid figures are reconstructed from operator records and machine telemetry and have not been independently audited. Türkiye and Bulgaria figures are the projected values published in those case studies and are not verified. This article is an analysis of operating data, not a forecast, a guarantee or a financial projection for any individual deployment. Site economics vary materially with location, market and deal structure.
Model your own payback before you commit capital
Send us a site — footfall, dwell time, local coffee pricing and the occupancy deal on the table — and we will work through the ramp curve with you, including the months that do not look good.
All figures in this article are pre-tax, exclude financing costs and the owner's time, and relate to the specific deployments described. Madrid figures are reconstructed from operator records and machine telemetry and have not been independently audited. Türkiye and Bulgaria figures are projected values published in those case studies and are not verified. Past performance of one deployment is not a forecast for another.