Hotel Budgeting: The Real Cost of Running the Season Blind
It is mid-August. Occupancy is high, the tills are working, everyone is on their feet. Hotel budgeting is not on anyone's agenda in these weeks, because a budget is the document that looks least necessary while money is moving. Which is precisely where its justification starts: a budget earns its keep not when the cash stops, but by telling you which decision you were making while it flowed.
The cost of managing without one never appears as a line item. No closing statement carries a row called "no budget." The cost is incurred all the same — inside decisions that arrive late, unit costs that go uncorrected, financing arranged in a hurry, and negotiating leverage handed away. This piece prices those four separately.
Every figure below is hypothetical. None of it is an industry average, a benchmark or a measured finding; all of it is an assumption, left open so you can rerun the arithmetic with your own property's numbers.
The season closed, you have the numbers, and nothing to compare them to
The closing statement lands. Rooms revenue this much, food cost that much, operating profit this. The figures are right, the accounting is clean, nobody is hiding anything.
Then comes the question: is that good or bad?
Without a budget there is no answer. Your only reference point is last year, and last year closed under the same uncertainty. Comparing a number to its own history is recall, not measurement: revenue up 6% is a poor result if capacity went up 9%.
A budget is the only structure you can set opposite the close. Without it, the closing statement stays an accounting document and never becomes a management one.
A budget is not a forecast, it is a record of decisions
The most common answer to "what is a hotel operating budget" is the wrong one. A budget is not next year's forecast. A forecast is what you think will happen. A budget is what you have decided to do.
The difference between budget and forecast shows up in practice. A forecast gets revised; a budget gets locked. A forecast is a probability distribution; a budget is a commitment. When a forecast misses, you update the model; when a budget misses, you interrogate a decision. Load both jobs onto one spreadsheet and you lose both — in most properties the file labelled "budget" is a forecast, which is exactly why it binds no decision.
One hypothetical property carries the whole piece. Its decisions, written down:
| Assumption (hypothetical) | Value | How it was derived |
|---|---|---|
| Rooms | 200 | Decision: capacity fixed |
| Operating days | 184 | Decision: 1 May – 31 October |
| Sellable room nights | 36,800 | 200 × 184 |
| Budgeted occupancy | 80% | Decision: hold rate, do not chase occupancy |
| Budgeted room nights | 29,440 | 36,800 × 80% |
| Budgeted ADR | TRY 5,000 | Decision: hold the channel mix |
| Rooms revenue | TRY 147,200,000 | 29,440 × 5,000 |
| Non-rooms revenue (TRY 2,500 per room night) | TRY 73,600,000 | 29,440 × 2,500 |
| Total revenue | TRY 220,800,000 | 147,200,000 + 73,600,000 |
Every row is a decision, not a prediction. Behind "occupancy 80%" sits the decision not to discount into 86%. The budget puts that in writing, and when somebody discounts mid-season the variance is measured against that decision.
First cost: decisions arrive late
In an unbudgeted property, decisions are not made wrongly. They are made late. The gap is the entire cost.
How do you learn that an expense line is running above plan? In a budgeted property, at the monthly close, against the budget row. In an unbudgeted one, only once the line has grown large enough to notice — usually at year-end. Through every month in between, the overrun recurs.
In our hypothetical, average in-season volume is 29,440 / 6 = 4,907 room nights a month. Take three overruns and assume the same size in both properties, with only detection lag differing:
| Overrun (hypothetical) | Monthly impact (TRY) | Derivation | Budgeted: 1 month | Unbudgeted |
|---|---|---|---|---|
| Channel mix drifting into commission (2 points of ADR) | 490,700 | 4,907 × TRY 100 | 490,700 | 3 months → 1,472,100 |
| Twelve unplanned seasonal hires | 720,000 | 12 × TRY 60,000 | 720,000 | 2 months → 1,440,000 |
| Maintenance contract renewed at a higher rate | 210,000 | Assumption | 210,000 | 4 months → 840,000 |
| Total | 1,420,700 | 3,752,100 |
Hypothetical gap: TRY 2,331,400. Not one lira of it is attached to a contract, a loss or a mistake. It is the price of detection lag, nothing else.
The mechanism matters: a budget does not improve the decision. It installs the mirror in front of the decision a year in advance. Where there is no mirror there is still a variance — just no image of one.
Second cost: you cannot correct a variance you cannot see
A budget is necessary to see a variance, but not sufficient. Comparing against a static budget understates variances systematically.
Say occupancy was budgeted at 80% and came in at 77%: 36,800 × 77% = 28,336 room nights. Food cost is budgeted at TRY 900 per room night.
| Comparison base | Calculation | Amount (TRY) |
|---|---|---|
| Static budget (29,440 room nights) | 900 × 29,440 | 26,496,000 |
| Flexed budget (28,336 room nights) | 900 × 28,336 | 25,502,400 |
| Actual (hypothetical) | — | 27,100,000 |
| Variance vs static budget | 27,100,000 − 26,496,000 | +604,000 (2.3%) |
| Variance vs flexed budget | 27,100,000 − 25,502,400 | +1,597,600 (6.3%) |
Same actual, two different stories. The static budget says "2.3% over, immaterial." The flexed budget asks the right question instead: did you spend more on food while serving fewer guests?
Take it to unit cost and the answer is unambiguous: 27,100,000 / 28,336 = TRY 956.4 per room night against TRY 900 budgeted, or TRY 56.4 over. That is not a volume effect, it is unit-cost decay — and unit-cost decay persists. Finding a volume variance at year-end is irritating. Finding a unit-cost variance at year-end is expensive.
The working rule: flex the budget to volume, then compare per unit. An unflexed budget is a budget that hides variances.
Third cost: cash flow stays exposed mid-season
A profitable season does not mean an easy one for cash. In a seasonal property much of the spend lands before any revenue does: maintenance, pre-opening, deposits, hiring, opening stock. That is a working-capital problem, and it has no direct relationship to profitability.
The hypothetical property's annual cash profile:
| Month | Cash in (TRY) | Cash out (TRY) | Net | Cumulative |
|---|---|---|---|---|
| February | 3,000,000 | 8,000,000 | −5,000,000 | −5,000,000 |
| March | 6,000,000 | 12,000,000 | −6,000,000 | −11,000,000 |
| April | 11,000,000 | 19,000,000 | −8,000,000 | −19,000,000 |
| May | 22,000,000 | 21,000,000 | +1,000,000 | −18,000,000 |
| June | 34,000,000 | 26,000,000 | +8,000,000 | −10,000,000 |
| July | 44,000,000 | 30,000,000 | +14,000,000 | +4,000,000 |
| August | 46,000,000 | 30,000,000 | +16,000,000 | +20,000,000 |
| September | 30,000,000 | 22,000,000 | +8,000,000 | +28,000,000 |
| October | 18,800,000 | 15,936,000 | +2,864,000 | +30,864,000 |
| November–January | 6,000,000 | 2,000,000 | +4,000,000 | +34,864,000 |
| Total | 220,800,000 | 185,936,000 | +34,864,000 |
In the hypothetical the year closes with an operating surplus of TRY 34,864,000 and still needs TRY 19,000,000 of funding at the end of April. Both sentences describe the same property, and both are true.
What separates the two properties is when that 19 million becomes known. The budgeted one knows it the previous October and negotiates the facility out of season. The unbudgeted one discovers it in the last week of April and borrows under pressure. Assume the spread between the two is 10 percentage points a year — an assumption, not a market rate:
19,000,000 × 10% × 4/12 = TRY 633,333
Same money, same term, same property. The only variable is notice.
Fourth cost: purchasing loses its negotiating position
The part of a supplier negotiation that earns money is not the price, it is the volume commitment. A property without a budget cannot commit volume, because it does not know its volume. A commitment you cannot make turns into spot buying.
| Item | Annual volume (hypothetical) | Contract unit | Spot unit | Gap (TRY) |
|---|---|---|---|---|
| Laundry | 29,440 room nights | TRY 62 | TRY 74 | 353,280 |
| Dry goods, main group | 220 tonnes | TRY 38,000/t | TRY 41,500/t | 770,000 |
| Cleaning chemicals | 18,000 litres | TRY 145/L | TRY 168/L | 414,000 |
| Fixed-price energy tranche | 4,200,000 kWh | TRY 3.10 | TRY 3.38 | 1,176,000 |
| Total | 2,713,280 |
In the hypothetical that is 7.8% of the operating result. No operational failure anywhere in it — the discount simply could not be claimed, because no number could be quoted.
The insidious part is that the loss appears nowhere. A spot price is still an invoice, and it is booked correctly as one. A discount you never obtained does not generate an expense line. In hospitality the cost lines that never reach the P&L run heavier than most groups assume, and they all share one property: no record, real cost.
You cannot compute profitability without a fixed/variable split
The single highest-return component of a budget is the split between fixed and variable cost. Without it you do not know what the next room night contributes to profit — which means your pricing decisions have no financial basis.
Variable cost per room night for the hypothetical property:
| Variable item | TRY / room night |
|---|---|
| Housekeeping supplies and laundry | 180 |
| Food and beverage cost | 900 |
| Variable energy and water | 150 |
| Channel and agency commission (12% of ADR) | 600 |
| Guest supplies and amenities | 70 |
| Total variable cost | 1,900 |
Annual fixed cost base:
| Fixed item | Annual (TRY) |
|---|---|
| Core year-round team (40 people) | 30,000,000 |
| Seasonal payroll floor | 40,000,000 |
| Building, insurance, property tax, depreciation | 24,000,000 |
| Fixed energy, maintenance, grounds | 13,000,000 |
| Brand and operator fixed fees | 10,000,000 |
| Fixed sales and marketing budget | 8,000,000 |
| IT, licences, subscriptions, advisory | 5,000,000 |
| Total fixed cost | 130,000,000 |
Three numbers fall out of that, and none of them can be computed without a budget:
- Contribution per room night: (5,000 + 2,500) − 1,900 = TRY 5,600.
- Break-even occupancy: 130,000,000 / 5,600 = 23,214 room nights → 23,214 / 36,800 = 63.1%.
- Operating result: 29,440 × 5,600 = 164,864,000 − 130,000,000 = TRY 34,864,000, or 15.8% of revenue.
Knowing break-even occupancy changes every in-season rate decision you make. Above 63%, each additional room night carries TRY 5,600 of contribution, so even a room sold at TRY 3,000 contributes. Below 63%, the same discount deepens the loss. One decision, two opposite signals — and the only thing that tells them apart is the cost structure in the budget.
Sensitivity reads off the same figures: five points of occupancy is 36,800 × 5% = 1,840 room nights × 5,600 = TRY 10,304,000 of contribution, or 29.6% of the hypothetical operating result. A five-point occupancy move carries almost a third of the profit. Management that does not know this discusses the occupancy target as a "somewhat flexible" one.
The same logic runs on the HR side: tying an HR score to financial impact is only possible where the budget has already separated the cost structure.
The materiality test: which variance actually changes a decision
A budget produces a long variance list, and investigating all of it is impossible. Accounting has a ready answer: materiality. Information is material if omitting or misstating it would change a user's decision. The IFRS practice statement Making Materiality Judgements (2017) sets that test out as a four-step process. The statement is non-mandatory — a framework for judgement, not a requirement.
Translated to budget variance it is one line: which decision would knowing this variance change? If none, report it and stop.
Set the threshold in the hypothetical at 5% of budgeted operating result: 34,864,000 × 5% = 1,743,200, rounded to TRY 1,750,000.
| Line | Comparison base (TRY) | Actual (TRY) | Variance (TRY) | Threshold test |
|---|---|---|---|---|
| Rooms revenue | 147,200,000 | 141,680,000 | −5,520,000 | Material |
| Channel and agency commission (flexed) | 17,001,600 | 19,100,000 | +2,098,400 | Material |
| Food and beverage cost (flexed) | 25,502,400 | 27,100,000 | +1,597,600 | Below threshold — investigated anyway |
| Fixed energy, maintenance, grounds | 13,000,000 | 13,420,000 | +420,000 | Immaterial |
| Housekeeping supplies and laundry (flexed) | 5,100,480 | 5,310,000 | +209,520 | Immaterial |
| IT, licences, subscriptions | 5,000,000 | 5,240,000 | +240,000 | Immaterial |
The third row misses the threshold and still goes on the list, because materiality is not measured by amount alone. Its driver is a unit cost, and a unit cost recurs every month. A one-off TRY 1.6 million can be waved through; TRY 56.4 per room night, every month, cannot. We draw the same distinction on the indicator side: the financial materiality test decides how much voting power each line carries.
A hypothetical: two properties, one budgeted, one not
Put the four costs in one table. Two hypothetical properties — same capacity, same revenue, same operation. The only difference: Property A budgets, Property B does not. Every figure was calculated earlier in this piece, and all of it is hypothetical.
| Cost heading | Property A — budgeted (TRY) | Property B — unbudgeted (TRY) | Gap (TRY) |
|---|---|---|---|
| Delayed-decision impact | 1,420,700 | 3,752,100 | 2,331,400 |
| Uncorrected unit-cost variance | 798,800 | 1,597,600 | 798,800 |
| Unplanned financing cost | 0 | 633,333 | 633,333 |
| Negotiating leverage lost in purchasing | 0 | 2,713,280 | 2,713,280 |
| Total | 2,219,500 | 8,696,313 | 6,476,813 |
The logic of the second row: both properties suffer the same unit-cost decay, but Property A sees it at the July close and corrects it across the remaining half of the season, so only half of it lands.
The hypothetical gap is TRY 6,476,813 — 18.6% of budgeted operating result, or 6,476,813 / 29,440 = TRY 220 per room night. To restate it: that number is not an industry finding, it is the arithmetic consequence of the assumptions above.
Note that Property A's column is not zero. A budget does not prevent the variance. It shortens detection lag — and in the hypothetical, that alone is worth 6.5 million.
Why the USALI account structure makes budgeting easier
The most tiring part of building a budget is the chart of accounts: which cost belongs to which department, which item to which row. On the lodging side that work is already done. The 12th Revised Edition of USALI was published by HFTP, the revision was overseen by the Global Finance Committee jointly sponsored by AHLA and HFTP, and it takes effect on 1 January 2026 — which is the budget year you are building this month.
Changes in this edition that bear directly on budgeting:
- The Utilities Schedule is replaced by an Energy, Water and Waste (EWW) Schedule.
- A new section covers all-inclusive hotels.
- A new Payroll FTE Schedule.
- A new Annual Mandatory Brand and Operator Costs Schedule.
- Refined categories for loyalty programme and executive lounge costs.
The all-inclusive section is a direct practical gain for resort operators: the perennial argument over where the food cost embedded in a board rate belongs now has a standard answer. The brand and operator costs schedule moves one row of our fixed-cost table above into a schedule of its own, which makes the fixed/variable split legible at the chart-of-accounts level.
The real benefit is that budget and actual share one chart of accounts. Where budget and actuals sit in separate structures, calculating variance means re-mapping them every period — and that re-mapping is the answer to why the variance report always arrives late.
The fair criticisms of budgeting — and why they are not an argument for none
The sharpest criticism of budgeting comes out of the finance literature and deserves to be taken seriously. Michael Jensen's Paying People to Lie: the Truth about the Budgeting Process (European Financial Management, 2003, 9(3):379-406) argues that target-based pay pushes people to game the system: superiors and subordinates lie in formulating budgets, and they game the realisation of them.
The hospitality equivalents are familiar. Targets get set low so the bonus is safe. Revenue gets pushed into next period because this period's target is already cleared. Maintenance slips to October so the in-season budget stays clean.
Jensen's remedy is not to abandon the budget. The paper argues for removing the kinks from the pay-performance profile: structures where the reward jumps just above and just below target are what manufacture the gaming. Removing the budget does not solve the problem, it only removes the measurement.
Two practical sentences follow. A budget is a measurement document, and it can be held separately from the bonus contract. Where it is held separately, gaming the budget pays less, because there is no bonus curve left to game.
The real return on a budget is decision quality, not savings
Budgets are usually defended on savings, and that defence is wrong. A budget lowers no cost by itself; no spreadsheet cancels an invoice.
What a budget produces is decision quality, and it shows up in four concrete forms. Detection time shortens. Variance separates from volume effect. Financing needs get negotiated out of season. Volume commitment converts into price. In the hypothetical, those four together came to roughly a fifth of the operating result — without cutting a single cost line.
The step after that is holding the same structure across a portfolio. Holding budget for every property in one structure is what makes comparison between properties possible in the first place; budgets kept in separate files produce separate charts of accounts, and comparison ends there.
Conclusion
What managing without a budget costs comes down to three sentences.
First, that cost is invisible but incurred. No statement carries a "no budget" row; the cost distributes itself into the late decision, the uncorrected unit cost, the unplanned financing and the volume commitment nobody could make.
Second, a budget's function is to record decisions, not to predict. A budget that forecasts accurately is not a good budget; a budget that puts on paper which decision you made and why is. Variance is measured against that decision, not against a prediction.
Third, the criticisms of budgeting are sound but aimed wrongly. Target-linked bonuses manufacture gaming, and what manufactures it is the reward curve, not the measurement.
The statement that lands on your desk at the end of the season is being determined this month. August is not too late to decide what will sit opposite it. January is.
Rebuilding this arithmetic with your own property's figures is usually a day's work; the hard part is not the arithmetic, it is deciding which row is a decision and which is a forecast. At HRMIM we work on the measurement and financial reporting side of hospitality, and we enjoy talking to teams opening this argument for the first time — if you would like help pinning down where to start, get in touch.


