Municipal permits, unit format, real service capacity, diary of dates and viability, with eight by-laws read one by one and a reproducible calculation engine.
A decision report, not an overview of the food truck sector. Every figure comes out of a reproducible engine whose 99 assumptions are published one by one, and every decisive statement carries its source, its date and the confidence grade we give it, including the three gaps the report expressly says it does not know.
The report audits by name the venta ambulante (street trading) by-law and the public-domain occupation regime of Madrid, Barcelona, Valencia, Seville, Málaga, Zaragoza, Bilbao and San Sebastián/Donostia, with a cut-off of 4 August 2026, citing the specific article that decides in each one: which product the municipal catalogue admits, which selling arrangement is authorised and which is not, for how long and with what renewal. And it declares its own perimeter in every sentence, not once in the methodology: Spain has more than 8,100 municipalities and the others have not been read, so no conclusion is extended to them. For yours, the report delivers the exact enquiry to make before buying anything.
In the same tender and on the same day in Zaragoza, a site with a €23,900 reserve price was awarded at €72,014 after fifteen bids, while another at €5,600 closed at €6,000 with a single bid. The report publishes both files and draws the reading that inverts the usual reasoning: where the fee does not rise it is because there are no customers, so the price of the ground works as a thermometer of footfall, and using the reserve price as a budgeted cost is a methodological error. It also documents the tenders that squeeze the margin from both ends —maximum selling price and minimum turnover at the same time— and the case of a fee paid in kind.
The 750 kg maximum authorised mass —the boundary for registration, roadworthiness testing and driving licence— has to be weighed with the unit ready and the tanks full, not unladen: the model closes at 532.5 kg with 217.5 kg of headroom, line by line. And the electricity consumption of an espresso machine does not depend on the number of groups but on its architecture: in the manufacturer's own brochure, three lever models of one, two and three groups declare the same wattage, while six pump models declare a different figure, likewise identical among themselves. The report compares the seven formats the sector lumps together, with their CAPEX, their registration regime and what each one allows and prevents.
The typical business plan multiplies footfall by conversion and treats the sum as done, without checking whether anyone could physically prepare those drinks in the window in which they are ordered. The report separates three magnitudes that are confused daily —sustained rate per service hour, fifteen-minute burst and peak-hour capacity— and shows that, with demand concentrated, drinks served are a function of drinks demanded, not a fixed cap. It publishes the full curve: what gets served at 70, at 95, at 130 and with demand saturated, how much of the queue is lost at each point, and above what volume a second person stops costing money and starts earning it.
The report reads Regulation (EC) 852/2004 chapter by chapter and shows that the «physical killer» the sector takes for granted —pressurised water, double sink— lives in the chapter that expressly excludes mobile premises: the hypothesis the research started from was false, and its refutation is published. What does change the product is two regional rules: a municipal guide that prohibits using tank water as an ingredient when an espresso is more than ninety per cent water, and a regional technical sheet requiring an industrial dishwasher at 80 °C or single-use tableware, which decides whether you can serve in a cup. With the food-business registration circuit resolved article by article: what is notified, to whom, and what suffices to be entered on the register.
The service session lasts six hours, but the real day —loading, driving, setting up, serving, breaking down, returning, cleaning at base and cashing up— lasts eleven door to door. That ratio governs the economics of the format and the engine applies it without exceptions: it deducts the eleven hours at the gross employer cost of a barista-operator under the collective agreement, before saying whether the capital earns anything. The report also explains why the role does not exist in the agreement under that name —it is the aggregation of three trades that in a fixed site are shared between three people— and what that means the day you want to replace yourself.
There is no Spanish data point for this format, and the report explains why that is a finding and not an omission: there is no Spanish coffee cart franchise, so there is no franchise prospectus, no documented case and no operator declaring volume. Faced with that, there are two possible behaviours: take a figure from another channel, another country or another link in the chain and present it as your own, or build demand up from conversion while declaring every assumption. The report does the second, marks the magnitude as low confidence with its band stated explicitly, puts it into the tornado, and delivers a four-variable counting protocol so that you measure your own location instead of believing ours.
With the article of the IVA (Spanish VAT) Act and a binding ruling from the tax authority, the report dismantles the doctrinal error that takeaway coffee is taxed at 21% «because it is a service», and points to the trap behind it: on a menu with coffee and soft drinks two IVA rates coexist, and a till configured to a single rate is eleven points out on those lines. Contributions to RETA (Spain's self-employed social security scheme) are closed against the order published in the BOE —with its rate, its fifteen brackets and its bases verified one by one—, without the range that circulates in the sector and does not exist in the legislation, and with the general contribution-system deduction cited under its correct article. It also explains what happened to the national street trading framework, repealed without a replacement, and what that means for anyone planning to operate across several cities.
If income depends on the calendar and not on footfall, the deciding asset is not the vehicle: it is the contracts. The roadmap therefore inverts the usual order —eight phases, four schedules and seven gates with a numerical threshold, in three of which the instruction is to walk away— and forbids buying a unit before phase 5: the earlier phases are walked in full with no CAPEX, with permits queried, counts done by hand and equipment hired for a real pop-up, which is the gate that decides the project and the cheapest of them all. The risks section turns eleven ways of losing the money into probability, indicator with its source and its alarm threshold, mitigation with an owner and a cost, and residual risk — starting with the two the buyer assumes are already solved.
The economic figures come out of a reproducible engine whose 99 assumptions are published in full with source, consultation date and confidence level, and the register states how many come from official primary sources and how many are estimates you should replace with your own quotations. The tornado moves twelve levers ±20% and publishes the range of each. The three gaps the research could not close are declared in the body of the report, not hidden in an annexe, together with what the report does instead. And the adversarial review starts with the most uncomfortable material: eight of the report's own statements that the evidence or an external audit knocked down, published with their correction.
Download a free sample with the cover and the full table of contents of the report.
The sample includes the cover and the contents, so you can see the full scope of the report. The analysis, the financial model and the verdicts are in the complete edition.
For the sole trader or small business about to turn their savings into a mobile coffee unit —a cart, a tricycle, a trailer, a van— and to give it their full working day: a barista wanting out of someone else's premises, a caterer extending their service, a roaster looking for a channel of its own, or a small investor who has seen the format on social media. It is written for whoever decides alone, with no committee and no legal department, and on the premise that they are risking their own money.
That it does not start from the vehicle or the budget, but from the article of the by-law that decides whether you are authorised to serve prepared coffee in each municipality: eight town councils read one by one with their rule cited, and the unaudited universe declared in every sentence rather than once in the methodology. And that it publishes its own limits: whatever could not be verified appears as a declared gap, not as an estimate dressed up as data.
It covers by name Madrid, Barcelona, Valencia, Seville, Málaga, Zaragoza, Bilbao and San Sebastián/Donostia, with fieldwork closed on 4 August 2026. Spain has more than 8,100 municipalities and the rest have not been read: the report says so expressly and extends no conclusion to them. What it does give you for yours is the method —which rule to look for, which article it examines, what to ask at the town hall and in what order— and the three by-laws that not even we could locate, declared as such.
Yes, and it is reproducible: all 99 assumptions are published in full with source, date and confidence level, so you can replace the ones you obtain yourself and redo the calculation. Five operating archetypes run through the same engine, with their CAPEX built line by line, their annual result, their net present value and the owner's euros per hour; unit economics per drink; and a tornado that moves twelve levers ±20% and publishes the range of each. The convention that governs the whole model is that the owner's eleven-hour working day is deducted at gross employer cost before any talk of return.
Magnitudes the report expressly says it does not know, instead of filling them with a plausible figure. There are three, and they are declared in the body of the report, not hidden in an annexe: the sub-letting price of a pitch at a large event —which is not published in any tender, tariff or organiser's website—, the HORECA price of milk, and the drinks per day of a cart in Spain. For the third, rather than borrowing a figure from another country or another channel, the report delivers a four-variable counting protocol so that you measure your own location.
Edition 7, with fieldwork closed on 4 August 2026. The report identifies which elements force a review —a municipal public consultation still open on the closing date, the terms and renewals of the authorisations, the self-employed contribution rates for the year, and the pitch awards, which are settled annually— and recommends a review horizon.
Eight municipalities audited one by one with the governing article cited and the unaudited universe declared, the pitch fee measured by actual award rather than reserve price, the three capacity magnitudes and the curve that turns demand into drinks served, five archetypes with CAPEX, annual result, net present value and the owner's euros per hour, 99 assumptions with source and confidence grade, twelve sensitivity levers, three declared gaps with their closing protocol, and a section written to destroy the report's own thesis.
Buy the report · €3,499