Hotel Budgeting Basics: Occupancy, ADR, RevPAR, Payroll, and Expenses

A budget is a set of assumptions with numbers attached. That sounds obvious, but it has a practical consequence that most budget files ignore: if the assumptions are not visible, the budget cannot be defended, adjusted or learned from. It can only be copied.

This article works through how the assumptions are built, how they are spread across the year, how payroll and expenses follow from them, and how to avoid the most common failure in budgeting, which is precision that nobody has earned. Figures throughout are invented for a fictional property, the Harbor View Hotel.

Start with assumptions, not numbers

The common way to begin a budget is to open last year's file and adjust it. It is fast, and it works for one cycle. After three or four cycles, nobody can explain why a particular line is what it is, because the reasoning that produced it was never written down, only its result.

The alternative takes longer once and saves time afterwards: build an assumptions sheet first, and calculate the departmental schedules from it.

At minimum, the assumptions sheet holds:

• Rooms available, and any planned out-of-order periods for renovation.
• Occupancy by month.
• ADR by month, ideally split by segment.
• Food and beverage capture rates and average spend.
• Spa and retail participation rates and average value.
• Cost per occupied room for variable expenses.
• Labor standards, such as hours per occupied room or covers per labor hour.
• Wage rate assumptions, including the timing of any increases.
• Food and beverage cost percentages.
• Known contractual changes: leases, insurance, licenses, management fees, utilities contracts.

Everything else in the budget should calculate from these. When the general manager asks in November why food and beverage payroll is up four percent, the answer is a cell reference rather than an effort of memory.

Occupancy

Occupancy is the assumption everything else leans on, so it is worth building rather than estimating.

Useful inputs, roughly in order of reliability:

• Actual occupancy for the last two or three years, by month.
• Group business already on the books for the budget year, which is a fact rather than a forecast.
• Known changes to inventory, such as rooms out for renovation.
• Known changes to demand: a new competitor opening, a large corporate account gained or lost, a convention center calendar, a change in flight capacity.
• Market outlook, treated as directional rather than precise.

Build occupancy month by month rather than setting an annual figure and dividing it. Hotels do not trade evenly, and an annual average phased flat produces a budget that is wrong in every single month while being right for the year, which is the least useful arrangement possible.

Where group business is already contracted, separate it from transient. The contracted portion is the firmest number in the budget and should be visible as such.

ADR and RevPAR

ADR should be built by segment where the data supports it. Transient, group, corporate negotiated and contract business behave differently, price differently and move independently. A single blended rate assumption hides the effect of mix, which is often where the real story of a year sits.

Mix matters because it moves the blended rate without any individual segment rate changing. A year that shifts five percentage points of occupancy from group to transient can show a rate increase that reflects nothing but composition.

RevPAR should be calculated, never typed. It is occupancy multiplied by ADR, and if it is entered as its own assumption it will eventually contradict the two figures it is supposed to summarize. This is one of the more common errors in a budget file and one of the easiest to avoid.

Monthly phasing and seasonality

Phasing is where a budget becomes usable for management during the year. A budget that is right for the year and wrong every month produces twelve months of variance explanations about nothing.

Things to phase deliberately rather than evenly:

• Seasonal demand patterns, from the property's own history.
• Day-of-week composition. Months contain different numbers of weekend nights, which matters at any property whose weekday and weekend business differ.
• Holidays and local events, which move between months from year to year.
• Contracted group business, placed in the months it is actually booked.
• Renovation or closure periods.
• Cost timing: annual insurance premiums, license renewals, marketing campaigns, preventative maintenance programs.
• Wage increases, placed in the month they take effect rather than spread.
• Utilities, which follow both weather and occupancy.

An easy correctness check: sum the phased months and confirm they reconcile to the annual figures. Phasing errors are common and, once the budget is loaded, awkward to unwind.

Getting departmental input

Department heads know things the finance team cannot know from data: which contracts are being renegotiated, which equipment is failing, which supplier is putting prices up, which events are already pencilled in.

The input works better when the request is structured:

• Send each department a schedule that is already populated with volume assumptions, so they are working from the same occupancy figures.
• Ask for specific things, not for a budget: known contract changes, planned headcount changes, equipment nearing replacement, price increases already notified.
• Ask for justification on any line that moves materially from the current year.
• Give a deadline that leaves time for a conversation about the answers.

Then review the returns against the volume assumptions. An expense that rises while its driver falls needs a reason, and asking for it during the budget process is considerably easier than asking about it in a variance meeting eight months later.

Modeling payroll

Payroll should be modelled from staffing and standards, not adjusted from last year's total. Two properties can spend the same amount for very different reasons, and only a model shows which one you are.

A workable structure has three layers.

Fixed positions

Management and salaried positions that exist regardless of volume. Budget these by position, with the annual salary, the effective date of any increase and any vacancy assumption stated explicitly.

Variable positions

Hourly roles that scale with volume: room attendants, servers, bell staff, spa therapists. Budget these using a productivity standard, such as hours per occupied room or covers per labor hour, applied to the phased volume assumptions. This is what makes the payroll budget move correctly when the volume assumption changes.

Related costs

Payroll taxes, benefits, insurance, holiday and vacation accrual, agency labor where it is expected. These are usually a percentage of the base, but the percentage should be stated as an assumption rather than embedded in a formula nobody has looked at for three years.

Once built, sense-check the result: labor as a percentage of departmental revenue, and cost per occupied room, against the current year. A model can be internally consistent and still produce an implausible number, and these two checks catch most of it.

Variable expenses

Variable expenses should be budgeted per unit of the driver, not as an annual total. Guest supplies, laundry, amenities and cleaning supplies are best expressed as a cost per occupied room. Food and beverage cost of sales is best expressed as a percentage of the relevant revenue. Credit card commissions and travel agency commissions follow revenue and channel mix.

The advantage of this approach appears during the year. When occupancy comes in above or below budget, the expense expectation moves with it automatically, and the variance that remains is the part that actually reflects performance rather than volume.

Fixed and semi-fixed expenses

Fixed costs are the easiest to budget accurately and the easiest to leave stale. Insurance, property taxes, licenses, subscriptions, contracts, leases and management fees should each be checked against the current agreement rather than uplifted by a percentage.

Semi-fixed costs deserve more thought. Maintenance, utilities and administrative expenses have a base level that exists regardless of volume plus a portion that moves with it. Budgeting them as entirely fixed understates cost in a strong year; budgeting them as fully variable overstates the savings available in a weak one.

Scenario planning

A single budget number implies a confidence nobody has. A small set of scenarios is more honest and more useful, provided the number of them stays small enough to be discussed.

Three is usually right:

• Base case. What you actually expect. This is the budget.
• Downside. A specific, plausible deterioration, such as occupancy three points lower with rate held.
• Upside. A specific, plausible improvement.

The value is in what the scenarios reveal about the cost structure. If a three-point occupancy fall wipes out most of the departmental profit, that is worth knowing in October rather than in June, and it is a conversation about flexibility rather than about forecasting.

Build scenarios by changing assumptions, not by changing outputs. If the model is built properly, changing the occupancy assumption should flow through to variable payroll, guest supplies, laundry and commissions without anyone touching those lines.

Avoiding false precision

The most common failure in budgeting is not inaccuracy. It is precision that has not been earned: a budget calculated to the dollar from assumptions that are educated guesses, presented with a confidence the underlying inputs cannot support.

Some habits that keep it honest:

Round the assumptions. Occupancy to a tenth of a percent is fine. ADR to the cent implies a rate-setting precision that does not exist in practice.

Do not build detail you cannot maintain. A budget modelled by day of week by segment is impressive until someone has to reforecast it in March. Detail that will not be updated is decoration.

Write the assumption down next to the number. A note reading "assumes the Riverside account renews at current volume" is the difference between a variance you can explain in ten seconds and one you investigate for an afternoon.

Say what you do not know. A budget presented with two or three explicit uncertainties, each sized, is more credible than one presented as settled. It also makes the reforecast conversation much easier, because the uncertainty was flagged in advance rather than discovered afterwards.

Keep last year's assumptions. Comparing what you assumed with what happened is the only reliable way to get better at this. Most budget files overwrite the assumptions each cycle and throw that away.

This article describes general practice. It is not accounting, tax, audit or legal advice, and it does not replace your own professional judgment or your company's procedures. Every example uses invented figures for a fictional property. Read the disclaimer.