HRMIM
Departmental Budget Ownership: Who Commits to What?

Departmental Budget Ownership: Who Commits to What?

Timuçin Diş · Strategy Consultant· 21 October 2026· 12 min read

A budget line with no signature is a forecast

Mid-October. The revenue and expense schedules are built, department inputs are in, the file is about to go out for sign-off. The right question at this point is not whether the budget is accurate. It is: whose name sits against each line?

In most properties the answer is vague. Energy is booked to engineering, but purchasing negotiates the tariff. Laundry cost shows up in housekeeping, while F&B generates most of the volume. Payroll lives in the HR file, but the department head makes the shift call. The number is on the page; the accountability is in the air.

You pay for that in-year. In the variance meeting everyone traces their variance back to somebody else's decision, the meeting turns into a round of explanations, and no corrective decision comes out of it. It cannot — the person who would make that decision was never named.

This piece closes that gap. What ownership actually means, which lines belong to a department and which have to stay central, how to build the sign-off and revision flow, and how all of it collapses into a single matrix. The goal is something you can put on the table on Monday morning.

What ownership means: the difference between a forecast and a commitment

Ownership is not a job title. It is three things sitting on the same line. A line missing any of them is a forecast, not a commitment.

  1. A name. One person who will defend the number. Not a department — "F&B" does not sign anything, the F&B manager does.
  2. A date. The day the number locks. An unlocked number drifts quietly through the year, and by December the variance is unmeasurable.
  3. A revision rule. The condition under which the number may move — written before that condition occurs, not after.

The practical distinction is a sentence. A forecast says, "this line will probably land somewhere around here." A commitment says, "I will hold this line at this level on these assumptions, and I will flag it if an assumption breaks." If nobody can say the second sentence, the line has no owner.

One caution: ownership is not control. Full control over a line is rare. Ownership means the person with the most influence over that line has taken on the job of defending the number. How you measure that influence is the next section.

The controllability test: what does the department actually influence

Before you book a line to a department, ask three questions in order. Anything short of three yeses and the line is not theirs.

  1. Does the department set the volume? Is the call on how many units get consumed theirs to make?
  2. Does the department set the unit price? Supplier, tariff or contract choice — is that inside their authority?
  3. Can they change the outcome in-year by changing behaviour? When a variance appears, do they have real room to manoeuvre?

Three yeses → full ownership. Two → ownership stays with the department, but revision authority is shared with the centre. One or none → the line stays central and appears in the department's report as information only.

The test earns its keep on the expense side, where the fixed/variable split and the ownership split get confused with each other. A line being variable does not automatically make it the department's: energy consumption moves with occupancy, but the price per kilowatt-hour is a central contract decision. The correct build books the volume to the department and the price to the centre.

The common mistake is distributing ownership along the organisation chart. The chart tells you who reports to whom. Budget ownership tells you who can move which number. Those are different questions.

The lines that should stay central

Pushing certain lines down to a department produces noise, not accountability. The department head ends up explaining a variance they cannot influence, loses credibility doing it, and spends attention they needed for the lines they genuinely own.

Keep these central:

  • Depreciation and rent. Downstream of an investment or lease decision; operations cannot touch them in-year.
  • Financing costs and FX effects. Treasury's call.
  • Insurance and taxes. Contract and statute driven.
  • Mandatory brand and operator costs. They come out of the management agreement; the property does not negotiate them.
  • Group IT and central service allocations. Allocated, not generated.

Some of this separation already exists in the chart of accounts. The 12th Revised Edition of USALI, the lodging industry's shared account structure, takes effect on 1 January 2026, and that edition adds a dedicated schedule for annual mandatory brand and operator costs. The same edition replaces the old Utilities schedule with an Energy, Water and Waste schedule and introduces a separate Payroll FTE schedule. That gives you a ready-made dividing line: a cost already separated at schedule level is ready to be separated at ownership level too.

The rule: a central line can appear in a department's report, but it does not enter the department's target. Visibility and responsibility are not the same thing.

When two departments share one line

Not every line lands on one department. Hospitality has a standard set of shared ones: laundry (housekeeping consumes it, F&B drives volume), energy (engineering manages it, everyone burns it), banquet labour (paid out of the F&B roster, delivering business sales booked), cleaning supplies (bought centrally, consumed everywhere).

There is one rule here: a line has one owner and may have several contributors. Two owners means no owner.

Three steps to apply it:

  1. The owner is not whoever generates the volume — it is whoever can change the behaviour most. For laundry that is usually housekeeping: they set the linen standard, the rewash rate and the stock turn.
  2. The contributor is obliged to reopen the number once a volume threshold is crossed. The threshold gets written while the budget is being built.
  3. The split ratio is written before the year starts. An allocation key argued about mid-year detaches the variance conversation from the cost conversation entirely.

A hypothetical: a property books 70% of laundry cost to housekeeping and 30% to F&B, and writes into the rule that the ratio is recalculated if banquet linen volume runs 20% above budget. The percentages in this hypothetical are not a reference point — they are numbers you derive from your own actuals. What matters is not the ratio itself but that the ratio and the threshold exist in writing before the year opens.

Revenue ownership is not expense ownership

Expense ownership has one dimension: who triggers the spend. Revenue ownership has two: who sets the price and who brings the volume. Those almost never sit with the same person.

On the rooms side, pricing authority sits with revenue management, part of the volume comes from sales, and marketing feeds the direct channel. Try to make all three the owner of one line and accountability dissolves. The build that works:

  • Price owner: defends the achieved average rate.
  • Volume owner: defends room nights, or covers.
  • Total line owner: one person defending the product of the two, usually the general manager.

The same separation holds in F&B: cover count is an operations outcome, average spend is a menu and pricing outcome.

Once revenue lines are split across departments, a weighting question follows — how much does each department's variance move the group result. That is where the budget hands over to performance evaluation, and departmental weights deserve their own decision rather than a default. In the budget itself you only need one thing: ownership clean enough that the weighting can be calculated later.

The sign-off flow: who prepares, who reviews, who locks

Ownership without a sign-off flow is just a spreadsheet. The flow has four roles, and they must not collapse into one person.

RoleWhoWhat they doOutput
PreparerDepartment headBuilds the assumptions and the numberDepartment budget form + assumption note
ReviewerFinanceConsistency, account mapping, double-count checkQuestion list and revision request
ApproverGeneral managerSettles the cross-department balance and the totalApproval decision
LockerFinance director or the centreCloses the version, takes over revision authorityLocked budget version

Three rules hold the table up.

The preparer does not approve. A manager who approves their own number sets their own margin of comfort.

The reviewer asks questions, it does not change numbers. If finance unilaterally corrects a department's figure, ownership moves to finance — and the department will not defend that figure in June.

A lock is a date, a person and a version number. "Approved in October" is not a lock. You cannot measure variance against an unlocked budget, because the comparison base keeps moving.

The revision rule: under what condition does the budget change in-year

Write the revision rule mid-year and you will have written it to legitimise a variance. It gets written before the year opens, and it is limited to three conditions:

  1. Scope change. A unit opens or closes, a floor goes into refurbishment, the season length changes.
  2. A unit price outside the department's control moves past a threshold. The threshold is written as a percentage, in advance.
  3. Contract change. Brand, operator, lease or collective labour agreement.

There is exactly one reason that does not belong on that list: "we are not going to hit the target." That is a performance conversation, not a revision trigger.

The most useful distinction here is between the budget and the forecast. The budget stays locked; the forecast moves. When your expectation changes in-year you update the forecast, not the budget. That way December answers two separate questions: what did we commit to, and what did we learn after committing. A property that rewrites its budget every quarter loses the first question permanently.

Why the number starts to bend once the budget is wired to a bonus

This is where the whole ownership structure is most fragile. If a department head's bonus depends on the budget that same head prepared, they are now doing two jobs: finding the most realistic number, and finding a number that is defensible for them. When those conflict, the second one wins.

Michael Jensen's 2003 paper on the budgeting process describes the mechanism directly: target-based pay pushes people to game the system both when the budget is formulated and when its realisation is managed — numbers get talked down, results get shifted between periods. Jensen's remedy is not abandoning budgets. It is removing the kinks from the pay-for-performance profile — the thresholds where the payout jumps just above or just below target.

Three moves in practice:

  1. Continuity instead of cliffs. Payout should rise smoothly with performance rather than stepping up at 100% of target. Wherever there is a step, the number bends toward it.
  2. Separate the lock date from the bonus conversation. Lock the budget; discuss bonus thresholds later, in a different meeting. Held together, the budget becomes a negotiating document.
  3. Reward the assumption, not just the result. If the manager who is first to say their assumption broke gets penalised for saying it, they will not say it again.

Participative budgeting: the benefit and the limit

Having the number built by the person closest to the line is a reasonable default. The executive housekeeper knows the rewash rate better than the finance director; engineering has a better read on which equipment fails this year. Participative budgeting has its own literature (Shields & Shields, 1998, Accounting, Organizations and Society 23(1):49-76).

The limit: participation is not a veto. The department participates in the assumptions, not in the target level. The split works like this:

  • Where the department has a say: consumption assumptions, roster structure, in-season distribution, how a productivity target will actually be achieved.
  • Where it does not: the group's total profitability target, investment decisions, central contracts, allocation keys.

The decision criterion is simple. If the argument is about how a number is built, the department has a say. If the argument is about whether the number has to be that high, the decision sits with the centre — and saying so plainly beats dressing the process up as consultative. A process where participation is theatre costs more trust than no participation at all.

The ownership matrix: department × line × approval

Everything above fits in one table. That table belongs on page one of the budget file, not in an appendix.

Budget lineOwnerContributorApproverIn-year revision
Rooms revenueRevenue managementSales, marketingGeneral managerOn scope change
Group and conference revenueSalesRevenue managementGeneral managerOn scope change
F&B revenueF&B managerSalesGeneral managerOn scope change
Cost of goods sold (F&B)F&B managerPurchasingFinanceAutomatic with volume
Payroll — housekeepingExecutive housekeeperHRGeneral managerIf season length changes
Payroll — F&BF&B managerHRGeneral managerIf season length changes
Energy, water and wasteEngineeringPurchasing (shared)FinanceIf tariff threshold is crossed
LaundryExecutive housekeeperF&B (shared)FinanceIf volume threshold is crossed
Repairs and renewalEngineeringCentral engineeringCentreFixed for the year
Mandatory brand and operator costsCentral finance—CentreOn contract change
Depreciation and rentCentral finance—CentreFixed for the year

The table shows three things at a glance: which lines have a single owner, which are shared, and which are not the property's decision at all. The two bold cells are deliberately shared — making a shared line visible instead of hiding it is what prevents the argument in June.

Keep the matrix at line level, not account-code level. Drop to account codes and the table runs to a hundred rows and nobody reads it. The lodging schedule structure helps you here: the schedule headings already give you a readable breakdown.

The same logic applies one layer up. Budget ownership shares its skeleton with the distribution of performance targets, and in a multi-property group the ownership layer in a goal tree should carry the same names as the budget matrix. If they differ, one of them is wrong.

Checklist before ownership goes live

Run eight items before you circulate the matrix. Any "no" and ownership is not ready to go live.

  1. Does every revenue and expense line have a single owner by name? A department label is not enough.
  2. Are the split ratio and the reopening threshold written down for every shared line?
  3. Have central lines been kept out of departmental targets?
  4. Are the preparer and the approver different people?
  5. Is there a lock date and a version number?
  6. Are the revision conditions down to three, with "we are missing target" struck off the list?
  7. Have bonus thresholds been moved to a meeting separate from the budget lock?
  8. Can every owner write the assumption behind their line in one sentence?

Item eight is the best test in the list. If a manager cannot state the assumption behind their line in a single sentence, they will not be able to defend it either — and when the variance shows up, neither the explanation nor the correction will arrive.

In a multi-property group the next job is holding the same ownership structure at every property and keeping the matrices comparable across them. If role names drift from property to property, there is no such thing as a single group-level variance meeting.

Conclusion

Budget ownership looks like an accounting topic and behaves like a decision topic. Three things to keep.

One: ownership is a name, a date and a revision rule on the same line. Missing any of the three, the line is a forecast.

Two: the controllability test derives ownership from real influence, not from the organisation chart. Book the volume to the department and the price to the centre; they do not have to go to the same place.

Three: bonus thresholds are where ownership breaks most easily. Separate the budget lock from the bonus conversation, and take the steps out of the payout curve.

What you get once this is built is not a more accurate budget — it is a faster variance meeting. When the number holds you know who held it, and when it does not you know who is bringing a correction to the table. That is the real test of next year: how many minutes the first variance meeting in February takes to reach a decision.


Working with multi-property hotel groups at HRMIM, the pattern we see most often is a well-built budget with the ownership never written down anywhere. If you want a second pair of eyes while you build your own ownership matrix, or you are trying to align role names across a group, get in touch — talking it through on your existing budget file is always faster than filling in a template from scratch.

Get New Posts by Email

We will email you when a new blog post is published.

Loading the form…

Related Posts

Budget Calendar: 7 Steps to Build Next Year's Budget Mid-Season

Budget Calendar: 7 Steps to Build Next Year's Budget Mid-Season

Next year's budget does not get built in January. It gets built while the season is still running — August to October is both your busiest operating stretch and the moment your data is freshest, and the calendar has to hold both. This piece lays out a seven-step budget calendar: the data cut-off date, collecting department inputs, building the revenue and expense sides separately, a revision round, sign-off and lock. Each step names who delivers what, and which decision is due by when.

26 August 2026 · 12 min read

BDGT - Budget Planning: What Hotel Budgeting Software Builds, and What It Doesn't

BDGT - Budget Planning: What Hotel Budgeting Software Builds, and What It Doesn't

The first seven posts in this series described a budget in two halves: building it and tracking it. BDGT - Budget Planning does the first half: it builds one property's annual budget, produces a G.O.P-to-N.O.P Summary, compares it with the previous budget and writes a numbered version with Approve And Lock. It does not track: no budget-versus-actual, variance, consolidation across properties or approval workflow, and this post says so.

18 November 2026 · 13 min read

Data Residency: Your Backups and Logs Live Somewhere Too

Data Residency: Your Backups and Logs Live Somewhere Too

When a vendor says "your data is in Türkiye", they are usually describing one tier: the application server. A whistleblowing record also lands in backups, a disaster-recovery replica, logs, an error tracker, a notification email and a support engineer's screen. This piece walks the three steps of KVKK article 9, explains where explicit consent sits after 1 September 2024, and puts eight tiers into one table with the evidence to demand at each.

28 October 2026 · 11 min read

Let's work with HRMIM

Explore our platform free of charge.