The controller in a hotel: what the job actually is, and who does it when there isn't one
An independent hotelier today has more data available than ever before. The PMS records every room movement, the channel manager tracks where every booking came from, electronic invoicing files away in structured form every euro spent with every supplier. And yet the simplest question a business owner can ask — how much did I really make on rooms last month, and why — often goes unanswered until the following spring, when the accounts arrive from the accountant’s office and describe a year that can no longer be corrected.
The problem, almost never, is a lack of data. It is the lack of someone whose job is to turn that data into decisions, at a regular cadence and with assigned responsibility. That job has a name: the controller. It is worth looking at it closely, not as an abstract category but as daily work: what this person does on a Tuesday morning, what they produce, and what happens in a hotel when nobody does it at all.
The invisible work that comes before the numbers
The first surprise, for anyone who pictures the controller as a producer of reports, is that the most important part of the job has nothing to do with producing numbers. It is building the conditions that make them comparable.
Before any margin can be calculated, somebody has to have decided that laundering the room linen is a rooms department cost and not a general expense; that the OTA commission is a distribution cost and not a reduction in revenue; that the seasonal barman’s contract weighs on the months he worked and not on the month he was paid; that energy is allocated between rooms, restaurant and spa according to a written and stable rule. These are tedious decisions, taken once and then defended for years. Without them two identical months produce two incomparable P&Ls, and every analysis that follows is an act of imagination.
This is precisely what UNI 11618 — the Italian standard defining the professional requirements of the management control specialist — places among the role’s primary tasks: to put in place the organisational, informational and IT arrangements of the control process, and to design, implement and keep up to date the tools that feed it. Not to interpret the numbers: to build the infrastructure that generates them. The standard’s list goes on to include support for strategic planning, a contribution to the early detection of business distress, support for the preparation of the statutory accounts and — an item worth as much as all the others — internal training and communication on control matters.
That last point deserves attention, because it is the one hoteliers underestimate most often. A departmental P&L is of little use if the front office manager cannot read it and the head housekeeper does not understand why she is being asked for hours worked per occupied room. A good part of the controller’s job consists of translation: making it comprehensible, to people with no accounting background, what the numbers are saying about their department, and what they are being asked to change.
The calendar, not the dashboard
A controller is recognised by the cadence at which they work, far more than by the tools they use. The distinction is a practical one: anyone can produce a brilliant analysis once a year, and that analysis will change nothing. The effect comes from regularity.
The daily rhythm is the lightest and involves very few numbers: rooms sold, average rate, the day’s revenue against forecast. It is not there to correct prices — that is the revenue manager’s responsibility — but to notice that the forecast on which spending commitments were made, from purchasing to rota planning, has stopped holding. A hotel that discovers at month end that it traded ten per cent below expectations has already spent as though those expectations had been correct.
The weekly rhythm is almost entirely devoted to labour cost, and not by chance: it is the heaviest line and at the same time the only one still governable within the week. Scheduled hours are set against expected rooms, overtime is read department by department, productivity is measured in elementary terms — housekeeping hours per occupied room, covers served per hour of restaurant labour. Sector estimates put labour cost in an Italian four-star hotel within a wide range, around thirty per cent of revenue, but these come from consultancy sources and with no stated methodology: the useful number is not the national average, it is your own property’s, measured the same way for twelve consecutive months.
The monthly rhythm is where the craft really shows. It comprises a fast management close, which does not wait for the tax accounts but works on the data available within the first few days of the following month; a P&L broken out by department, with each department’s margin; variance analysis against budget, which is the least interesting part; and finally the reforecast, which is the decisive part. A controller who confines themselves to explaining why March came in worse than expected is writing history. The value lies in rewriting, in the light of March, what is expected of April, May and June, and in saying which spending commitments that revision makes untenable.
The annual rhythm is the budget, which in an independent hotel is less an arithmetical exercise than an internal negotiation: it means putting in writing how much each department commits to produce and to consume, and turning the owners’ aspirations into numbers that somebody has agreed to defend.
Why the general manager cannot do it alone
There is a structural reason, before any technical one, why this function struggles to sit with the person running the hotel: whoever takes the decisions is rarely also the person best placed to verify their effects. It is not a question of honesty, it is a question of position. The general manager who pushed for opening the restaurant in the evening has a legitimate interest in reading the results in the most favourable way, and in the absence of an independent reading that way becomes the only one available.
The controller’s contribution is therefore partly technical and partly positional. They do not decide: they make decisions verifiable, and they put the price of each option on the table before it is chosen.
From this standpoint it is worth clarifying two boundaries that Italian practice confuses continually. The first is with the accountant, whose work is by construction retrospective and geared to compliance: it reports what has already happened, using classifications that serve the statutory accounts and the tax return, not the running of a department. It is necessary work and cannot be replaced, it simply does not answer the same questions. Asking your accountant which department is eroding the margin is asking for information their tools were never designed to produce.
The second boundary, less discussed and today more material, is with revenue management. The revenue manager answers for rooms revenue and is measured on revenue indicators, typically RevPAR and rooms revenue against budget. It is an incentive system consistent with their mandate and increasingly misaligned with the owners’, who look at operating profit. The international debate on this point is by now explicit: growth in labour, energy and distribution costs has broken the correspondence between revenue growth and profit growth, to the point that a positive movement in RevPAR can sit alongside a flat or falling GOPPAR, depending on how that revenue was composed. The sharpest analyses of this imbalance come from revenue management technology vendors, which counsels a degree of caution about the magnitudes, but the direction of the argument is confirmed by academic and industry sources too.
Translated into practice: if the increase in occupancy came from shifting volume out of direct channels and into the OTAs, with higher commissions, more rooms to service every day and higher consumption, the hotel can close the year fuller and less profitable than the year before. The controller is the person whose job is to ask that question while the year is still running.
A role that, historically, comes from outside
A digression is worth the space, because it helps explain why this role struggles to establish itself in Italy’s independent hotels. A piece published on Hospitality Net by Justin Taillon, a lecturer at Highline College, traces the evolution of the role through surveys by HFTP, the leading international association of hospitality finance and technology professionals.
In 1989 members were asked which skills the controller of the future would need, and named, in order, information technology, reading the P&L, revenue management, food and beverage and night audit. The same question, put again in 2018, produced a list with almost no overlap: interpreting the P&L, building and controlling budgets, management and leadership skills, analysing financial situations, reading the balance sheet. Thirty years of forecasting, and the only surviving item is the ability to read a P&L. The same text ventures a projection to 2033 built on a conversation with five futurists during an industry conference: advanced data analysis, technological competence, sustainability, international finance, risk management and cybersecurity. It is material for reflection, not a structured survey, and should be taken as such.
More interesting is a side observation in the same piece: hotel controllers rarely come from hospitality. They are predominantly accountants applying to the sector a training acquired elsewhere, and they tend to see themselves as outsiders to the hotel world. At the same time, and apparently at odds with their reputation as conservatives, it was they who brought computers into hotels: the industry’s computerisation in the 1980s started in the accounting offices and only later spread to the rest of the property.
For an independent hotelier the operational lesson is twofold. There is no point looking for a profile of the future, because the future of skills is historically unpredictable; the thing to look for is somebody who can read a departmental P&L and explain it to people who do not read numbers. And it is worth accepting that this person will, in all likelihood, have to learn the hotel, because the hotel is not where they come from.
The function without the person
There remains the point that makes this discussion either concrete or academic, depending on how it is approached. According to the Cassa Depositi e Prestiti paper on the Italian hotel system, chain hotels average one hundred and ten rooms, against thirty-three for everyone else, and among domestic operators the overlap between owning the building and running the business as a family remains the prevailing model. The European average sits at around sixty rooms. A full-time in-house controller, in a thirty-room hotel, is not a debatable choice: it is an unaffordable cost.
But UNI 11618 describes a function with defined content, not necessarily a full-time job. The useful question, for an independent, is not whether they can afford a controller. It is who exercises that function, at what cadence, and with what documentary evidence of having done so.
There are essentially three workable arrangements, each with a stated limitation. The first is internal: someone in administration is given the mandate, the training and the protected time to do it, with a periodic external review. It works if the time really is protected, which in practice happens rarely, because the first operational emergency takes it back. The second is external: a professional on a fixed monthly cadence, who closes, analyses and reforecasts. It works if the cadence is contracted rather than on call, because a controller summoned when the numbers start to worry always arrives too late. The third is that management itself takes it on, accepting the conflict of position and offsetting it with written rules and a second, independent reading at least quarterly. It is the most common and the most fragile, but it is preferable to nothing at all.
Something has recently entered this territory that Italian hoteliers would do well to know about. In April 2025 UNI published the reference practice UNI/PdR 167, on adequate organisational, administrative and accounting arrangements for small and medium-sized enterprises, with requirements and operational guidance for conformity assessment. The document translates into verifiable criteria an obligation that Article 2086 of the Italian Civil Code has placed on company directors since 2019, and it explicitly introduces the possibility of certification issued by an accredited third-party body.
The practical consequence is precise. The adequacy of the arrangements is not demonstrated by one person’s diligence, it is demonstrated by a process that leaves a trail: written procedures, cadences kept, dated documents. If the general manager changes tomorrow, the process must remain. It is the soundest reason, today, why the control function in an independent hotel should be formalised even when it is carried out part time by somebody who also does something else.
What software can do, and what it cannot
All three arrangements described above share the same constraint: one person’s time. And it is precisely on time, not on the nature of the function, that the technology of recent years has changed the arithmetic of the problem.
The part of the controller’s work that absorbs the most hours is not the analysis, it is the feeding of the system: retrieving invoices, classifying them, assigning them to the right department, reconstructing the accounting period of something paid in one month and consumed in another. In Italy that raw material already exists in structured form, because electronic invoicing delivers every month a complete and readable archive of costs, and it is why systems that automatically classify invoice lines and map them to a departmental framework — in hotel practice, to USALI — make such a difference. They do not eliminate the invisible work described at the start: the allocation rules still have to be decided by someone. They eliminate the cost of applying those rules over and over, which is what stops that work, in most independent hotels, from being done every month rather than once a year.
The next step, and the more interesting one, comes when cost data is crossed with forward-looking sales data rather than actuals. A system that knows the occupancy expected over the coming weeks and the property’s cost structure can estimate the cost of that forecast before it materialises: flagging that housekeeping hours scheduled are not consistent with the rooms expected, that the channel mix building for a given period compresses margin even as turnover rises, that a department is consuming its quarterly budget ahead of time. It is the controller’s weekly and monthly cadence, compressed into a continuous flow, with artificial intelligence suggesting where to intervene in the allocation of resources. It should be added that it also produces the documentary trail that UNI/PdR 167 makes a condition of adequacy: dated, repeatable reports that do not depend on the memory of whoever produced them.
That said, it is worth being precise about what software does not do, because the opposite promise circulates with a certain casualness. A system does not negotiate a budget with a head of department, does not decide to close the restaurant in November, and does not take responsibility for a choice. It answers the question of what is happening, not the question of what would be wise to do. And above all it applies the rules it has been given: if the chart of accounts is badly set up or the allocation criteria are arbitrary, automation reproduces that error with a consistency and a speed it never had before, making it harder to notice.
The realistic conclusion, for an independent hotelier, is therefore not that software replaces the controller. It is that it reduces the function from a full-time job to a few hours a month of interpretation — a threshold a general manager can genuinely sustain, provided a periodic external second reading remains, to offset the conflict of position that no technology eliminates.
The question that remains
The risk with an article like this one is that it sounds like yet another invitation to hire one more person in a property already struggling to cover its shifts. That is not the point.
The point is that in every hotel, every month, somebody answers for turnover and somebody else answers for the annual accounts, and in between there is a gap of ten or twelve months in which nobody answers for the margin. It is in that gap that the decisions shaping the year are taken: which channels to feed, how many shifts to schedule, which department to keep open in November, which supplier to renew with. The controller’s job is to occupy that gap with timely information and with a reading that has no interest in being optimistic.
If nobody does it, the hotel does not stop taking those decisions. It takes them without knowing what they cost.
Frequently asked questions
What does a hotel controller actually do? The controller designs and maintains the system that makes the property’s financial data comparable — chart of accounts, revenue and cost centres, allocation criteria — and produces, at a regular cadence, the management close, the departmental P&L, variance analysis and the reforecast for the following months. A significant part of the job also consists of making that data comprehensible to department heads with no accounting background.
What is the difference between the controller and the hotel’s accountant? The accountant works on final figures with classifications geared to statutory and tax obligations; their output arrives after the financial year has closed and serves to report. The controller works on in-year data with management classifications by department and responsibility centre, and their output serves to decide while the year is still running. The two functions are complementary and cannot be substituted for one another.
Can the revenue manager also act as controller? The two functions answer different questions and are measured on different indicators. The revenue manager is typically assessed on rooms revenue indicators, whereas the controller answers for the overall margin, which also depends on distribution costs, labour cost and the consumption generated by the volume sold. Merging the two responsibilities removes precisely the cross-check that makes the second one useful.
Does a thirty-room independent hotel need an in-house controller? In most cases a full-time in-house controller is not economically sustainable for a property of that size. The function, however, remains necessary, and it can be exercised by an internal resource with a specific mandate and training, by an external professional on a contracted monthly cadence, or by management itself with written rules and a periodic independent review — today with technological support that appreciably reduces the workload.
Can artificial intelligence replace the controller in a hotel? AI-based systems automate the most onerous part of the function, namely the classification of costs by department and the recurring production of reporting, and they can anticipate the financial impact of an occupancy forecast. They do not replace the interpretation of results, the negotiation of the budget with department heads, or decision-making responsibility, and they faithfully replicate any errors in how the chart of accounts was set up.
Is there a recognised qualification for controllers in Italy? The profession falls among the unregulated activities under Italian Law 4/2013 and involves no registration with a professional body. The requirements of knowledge, skill, autonomy and responsibility are defined by UNI 11618:2022, which allows voluntary certification by independent third-party bodies.
How often should a hotel produce management reporting? The useful cadence is layered: a few daily revenue indicators against forecast, a weekly check on labour cost and productivity by department, a monthly management close with a departmental P&L and a reforecast of the remaining months, and an annual budget cycle. It is the regularity of the cadence, more than the sophistication of the tools, that produces the management effect.