Budget Calendar: 7 Steps to Build Next Year's Budget Mid-Season
By late August a hotel is running two jobs at once. One is closing the season. The other is the job nobody has put on a calendar: building next year's budget.
The second one gets pushed. Operations are at full stretch, everyone is on the floor, and the sentence you hear is "let's sit down once the season is over." The season ends, November arrives, the data goes cold, the memory disperses, and the budget gets manufactured in a single January meeting. What comes out of that meeting is not a budget. It is last year's numbers with a percentage on top.
A budget calendar exists to stop exactly that slide. It is not a reminder about when the budget gets written; it is a document that commits which decision is due by which date, and who owns it. This piece builds that calendar in seven steps — from fixing the data cut-off to locking the approved version. Every step has one deliverable, one owner and one date.
Why building the budget in January is already late
The problem with a January budget is not that it is late. It is that most of it is redundant by the time it lands.
The point of an annual budget timeline is to make the decision before the money moves. But for most of the lines you are budgeting, the decision has already been made. So the useful question when you lay out the hotel budget process is this: in which month does this decision actually get taken?
| Decision | When it is really made | Does a January budget reach it |
|---|---|---|
| Early-booking rates and allotment contracts | Autumn | No — the contract is already signed |
| Key-position headcount plan and hiring launch | Autumn to winter | No — the season team is already designed |
| Refurbishment and capex items | At season close | No — the window for the work has passed |
| Annual supply and framework agreements | Year-end | Partly — you have lost your negotiating position |
| Marketing and channel spend | Winter | Yes, but late |
The first three rows are the ones that matter. If those decisions get taken before the budget is approved, the budget stops being the reasoning behind the decision and becomes a record of it. That is precisely where the financial case for a budget evaporates: the number exists, but the trade-off behind the number was never written down.
The second cost is quieter. A budget approved in January leaves the first quarter with nothing to compare against. What exactly are you reading January and February actuals against?
The advantage of budgeting mid-season: the data is still fresh
Building the budget while the season runs buys you three concrete things.
First, the memory is live. Ask a department head in September why housekeeping overtime spiked in July and you get an answer. Ask in February and you get a guess. Stripping one-off items out of last year's actuals only works while the people who lived through them are still standing there.
Second, you can test the assumption. You still have weeks of season left. The occupancy or rate assumption you just wrote for next year can be checked against this year's remaining booking curve. In the middle of winter that option does not exist.
Third, the headcount decision is still open. Where the season team ran thin and where it ran heavy is visible in August. That observation is the single most valuable input to next year's payroll budget, and it evaporates within three months.
If you want to follow the research side of this, one of the institutional homes for hospitality research is the Center for Hospitality Research at Cornell's Nolan School; its publication list regularly covers seasonal operations and revenue management.
The advantage comes with a bill: budget setup collides with your heaviest operating weeks. That is why you cannot place the calendar arbitrarily — the rule for that is in section ten.
Step 1 — Fix the data cut-off date
The first decision in a budget process is not a number. It is a date. Which day's actuals form the base?
Fix it or the whole process runs on different data. Finance works off the 31 August cut, front office opens last night's report, F&B remembers last week's covers. Three tables, three versions of the truth.
There are two defensible approaches. Pick one and write it down.
| Approach | When to use it | Where it is weak |
|---|---|---|
| Rolling twelve months back from the cut | Year-round city hotels | Hides shifts in when the season starts |
| Year-to-date at the cut, plus last year's remaining months | Seasonal properties | Splices two different years — label it explicitly |
The rule: data arriving after the cut-off does not change the budget. It goes into the revision round in Step 6. Without that rule the process never closes, because a new number lands every week and every new number reopens the table.
Set the cut-off against when data is genuinely available, not when you would like it to be. If the accounting close finishes on the fifth of the month, putting the cut on the third is lying to yourself. Measurement-calendar discipline does the same work on the budget side that it does on the scorecard side: a line whose period has not closed is not missing data, it is data that is not yet due.
Deliverable: one date, written down, circulated to every department head.
Step 2 — Clean last year's actuals: strip the one-offs
Raw actuals cannot be a budget base. They contain things that will not happen again.
Do the stripping with a list, not by eye. The recurring categories:
| Item type | Example | What to do |
|---|---|---|
| One-off capex or refurbishment | Floor renovation, kitchen rebuild | Strip it, move it to the capital budget |
| Non-recurring income | Insurance claim settlement, grant | Strip it — do not budget what will not repeat |
| Lost capacity | Floor closed for works | Strip it, rebuild occupancy on corrected capacity |
| One-time group or event | Congress, a single long-stay account | Strip it, then ask sales whether it can repeat |
| One-time payroll cost | Severance, settlement, litigation | Strip it, hold it under provisions |
| Price or currency shock | Sharp move in the energy tariff | Do not strip it — write it as an assumption |
That last row is the one people get wrong most often. A one-off is something with no probability of recurring — not something that surprised you. If the energy tariff moved unexpectedly, that is not a one-off item; it is an assumption that still applies next year.
Second rule: every strip carries a reason line. "Unusual" is not a reason. Which event, which date, which amount. An unexplained strip is the easiest way there is to flatter a budget.
Third rule: keep both columns — gross actuals and normalised actuals. The first time you argue about a variance next year, you will need both.
Step 3 — Build the revenue side by segment
The revenue side closes before the expense side. The reason is mechanical: most variable costs move with revenue volume, not the other way round.
Do not build rooms revenue on one occupancy and one average rate. Split the segments — OTA and travel agent, corporate and contracted, direct, group and MICE. Their pricing behaviour, cancellation behaviour and booking windows are all different. Buried inside a single multiplication, the budget loses its power to explain what happened.
Write three assumptions per segment: room nights, achieved average rate, and channel or commission cost. Achieved average, never rack rate.
Build non-rooms revenue with the same discipline. Tie F&B, spa, meetings and other revenue to guest count, not to occupancy. When occupancy holds but guests per room shift, non-rooms revenue moves — and you only see it if you picked the right driver.
Deliverable: monthly, rooms revenue by segment, non-rooms revenue tied to its driver.
Step 4 — Build the expense side on a fixed/variable split
Do not draw the expense skeleton from scratch. Lodging already has a chart of accounts for this: USALI, the uniform system of accounts for the lodging industry. The 12th Revised Edition was published by HFTP, the revision was overseen by the Global Finance Committee jointly sponsored by AHLA and HFTP, and it took effect on 1 January 2026.
That date bears directly on your setup, because the departmental schedules changed. The new edition replaces the Utilities Schedule with an Energy, Water and Waste (EWW) Schedule, adds a section for all-inclusive hotels, introduces two new schedules — Payroll FTE, and Annual Mandatory Brand and Operator Costs — and redefines the cost categories for loyalty programmes and executive lounges.
The practical consequence: if your prior-period actuals sit on the older schedule structure, building the new budget on the new one will not reconcile by itself. Decide once how the old energy and water lines map onto the new EWW lines, write the mapping down, and stop relitigating it. The Payroll FTE schedule also lets you build payroll on headcount rather than on a lump sum — in a seasonal property that means a monthly FTE curve, held separately from your wage-inflation assumption.
Classify every expense line and name its driver:
| Class | Definition | How to budget it |
|---|---|---|
| Fixed | Independent of volume, decision-driven | Annual amount / 12, with the decision date attached |
| Variable | Moves directly with volume | Unit cost × driver |
| Step (semi-variable) | Jumps at defined thresholds | Threshold table — which occupancy opens which shift |
That third row is the crowded one in hotels. Housekeeping shifts, the number of outlets you open, pool lifeguards — none of it is linear, all of it is stepped. Budget them as variable and you will cut too deep in low season and overshoot in peak.
The rule: no variable cost enters the budget without a written driver. The driver is room nights, guest count, covers, or FTE. Nothing else.
Step 5 — Collect department inputs and resolve the collisions
Sending departments an empty template is the classic failure in this step. Every department that receives a blank sheet invents its own volume assumption, and four different occupancy expectations come back to you.
Send the template with the assumptions already filled in: monthly room nights, guest count, opening and closing dates, average length of stay. The department does not change those; it states how much resource it needs under them. If it disagrees with an assumption, it writes the objection on a separate line — and that objection is an input to Step 6.
Once inputs land, collisions appear. Work them in order:
| Collision | How to resolve it | Who decides |
|---|---|---|
| F&B guest count does not match front office occupancy | Single source: guest count from the revenue budget | Finance |
| Housekeeping staffed to peak occupancy | Convert to the threshold table, spread across monthly FTE | Operations |
| The same refurbishment sits in both opex and capex | Delete it in one place — do not shrink both | Finance |
| Training budget appears in both HR and the department | HR owns it, the department consumes it — one line | HR |
| Two departments each requested the same equipment | Shared line, single owning department | General Manager |
Double-counting that survives this step goes to sign-off inflated, and the cut then gets made blind in the approval meeting. A blind cut trims the right line and the wrong line at the same rate.
Step 6 — The revision round: which assumption moves the budget
The revision round is not a "let's look at it again" meeting. It has a defined input: the assumption register.
Pull every assumption the budget rests on into one table: the assumption, its value, its owner, its source, its sensitivity.
You can fill the sensitivity column without generating a single figure — high, medium, low is enough for the decision and far faster. The high-sensitivity assumptions are your watch list for the year after sign-off. If you do want numeric scenarios, build no more than three and label each one hypothetical in plain sight; a model output must never read as a commitment.
The round has two outputs:
- The assumptions that changed, and how that flows through revenue and expense.
- A watch list of assumptions that did not change but carry high sensitivity.
The rule: the revision round happens once, and its date is set in advance. If you are opening a second round, you have returned to Step 1 and effectively cancelled your own cut-off date.
Step 7 — Approve and lock: an unlocked budget is not a budget
A budget sign-off process is three things: a date, a named approver, a version number. All three in writing.
Locking means that version does not change after approval. If it gets revised in-year, you open a new version; the original is never overwritten. The comparison base is always the locked version. In properties that skip this, you hear one sentence in November: "we'd already updated that line." A variance measured against an updated budget is not a variance.
If budget targets are going to feed a scorecard, wire the connection here rather than afterwards. In a multi-property group the structure is the same one behind a KPI goal tree: which budget line rolls into which target, and which indicator measures it.
One more separation to make: the bonus conversation does not happen in the budget-lock session. Michael Jensen's Paying People to Lie (European Financial Management, 2003) argues that budget- and target-linked pay produces gaming in two places at once — in how the budget is formulated, and in how the actuals are managed. Jensen's remedy is not to abandon budgeting; it is to remove the kinks from the pay-performance profile. In practice: if you build a threshold that pays out at target and pays nothing one point below it, you have handed the person building the budget a reason to build it low.
Placing the calendar around your operating peak
The seven steps are fixed. The dates move by property. The placement rule is one sentence: in peak weeks you collect data, you do not make decisions.
| Property type | Peak | Data cut-off | Department inputs | Sign-off |
|---|---|---|---|---|
| Summer resort | July-August | End of August | September | End of October |
| Winter resort | December-March | End of April | May-June | August |
| Year-round city hotel | Spring and autumn | End of August | September | November |
| Multi-property group | Varies by property | Same date at every property | September | Consolidation + 3 weeks |
That last row matters. Even when peaks differ, the cut-off date has to be shared, or consolidation adds up two different time windows.
Build the calendar backwards. Put the sign-off date down first, then set the revision round three weeks before it, then set the department input deadline four weeks before that. The cut-off sits ahead of all of them. Plan forwards and the sign-off date drifts; plan backwards and it holds.
Three mistakes almost everyone makes during setup
One — adding a percentage to last year. Incremental budgeting is fast and contains no assumptions at all. It has a direction but no reasoning. When a line moves during the year, nobody can explain why, because no decision was ever made about it.
Two — one person building the whole thing. A budget built by finance alone is technically correct and operationally ownerless. If a department head does not recognise their own lines, you will hear "that's not my budget" all year — and they will be right.
Three — never putting the sign-off date on the calendar. The budget becomes "ready", sits in a folder, and nobody signs it. Then it gets verbally accepted in a January meeting and no one can say which version is live. The sign-off date is the first date you nail down, not the last.
Checklist before the budget goes for sign-off
Do not send it up unless all ten of these are yes.
- The data cut-off date is written down and circulated to every department
- One-off items are stripped, each with a reason line
- Gross and normalised actuals sit in separate columns
- Rooms revenue is built by segment on achieved average rate
- Non-rooms revenue is tied to guest count, not occupancy
- Every variable cost has a written driver
- Step costs have a threshold table
- Department inputs came back on a pre-filled assumption template, and double-counting is cleared
- The assumption register exists, with high-sensitivity items flagged
- The sign-off date, the approver and the version number are all fixed
A multi-property group adds one more line to that list: keeping one budget calendar across every property. When the calendar drifts property by property, consolidation jams in the same place every year — head office tries to prepare for sign-off on a date when some properties are still collecting inputs.
Conclusion
A budget calendar is not a reminder. It is an ordering of decisions. Six of the seven steps happen before anyone produces a number.
Hold the sequence: fix the cut-off, clean last year, build revenue by segment, build expense on a fixed/variable split, collect department inputs on a pre-filled template, revise the assumptions in one round, then approve and lock. Then run the checklist.
Place this calendar correctly once and January stops being the month you write the budget and becomes the month you start reading it. That is the real return: the year's first variance meeting opens with "did that assumption hold?" instead of "where did this number come from?".
If you are sitting down to build next year's budget while the season is still running, and the part you are stuck on is fitting the calendar around your operating peak, let's talk it through. The HRMIM team works on planning and measurement structure in hotel groups; to walk through your own property set with us, get in touch.


