Comment Prévoir les Coûts d'Exploitation Hôtelière pour une Due Diligence d'Acquisition : Un Cadre de Modélisation Étape par Étape

Hotel financial modelling dashboard showing cost drivers, occupancy scenarios, and supplier contract timelines

Forecasting hotel running costs is the cornerstone of rigorous acquisition due diligence — a disciplined, assumption-driven process that separates viable investments from hidden liabilities. Unlike static cost summaries, a robust forecast models how expenses evolve under realistic operational scenarios: shifting occupancy, wage adjustments, supplier renegotiations, and FF&E lifecycle events. This guide walks buyers, investors, and current owners through building a dynamic, transparent cost model grounded in hospitality-specific drivers — not generic spreadsheets. We focus exclusively on the *modelling framework*: how to structure assumptions, layer sensitivity tests, validate inputs against property-level data, and interpret outcomes for negotiation, financing, and post-acquisition planning. No jurisdictional generalisations are made; where tax, labour law, or regulatory factors apply, their scope is explicitly named. The result is a repeatable, audit-ready methodology aligned with global best practices in hospitality investment analysis.

Key Takeaways

A credible hotel running cost forecast starts with granular, property-level expense categorisation — never aggregated 'opex' line items — and maps each to its primary driver (e.g., occupancy, room count, FTE headcount).
Sensitivity analysis must test three non-negotiable levers: occupancy variance (±15% from baseline), annual wage inflation (3–5% range), and supplier contract rollover timing (e.g., energy, linen, PMS).
FF&E reserve modelling requires dual tracking: a capitalised replacement schedule (by asset category and useful life) *and* an annual accrual rate calibrated to actual depreciation patterns, not industry averages alone.
In the UK, business rates, VAT treatment of service charges, and National Living Wage escalations must be hard-coded into the model — but these rules do not apply universally and are labelled as UK-specific throughout.
A forecast loses credibility if it omits documented assumptions about management structure: whether the buyer will self-manage, appoint a third-party operator, or retain existing staff — each carries distinct cost implications.
The final model output should include at least two scenario views: a base case (conservative occupancy, confirmed contracts), and a stress case (80% occupancy, full wage uplift, no supplier renewals before Year 2).

Step 1: Deconstruct Historical Costs Into Driver-Linked Line Items

Accurate forecasting begins not with projections—but with forensic dissection of what the property has *actually* spent. Generic benchmarks (e.g., 'F&B labour is 30% of revenue') mislead when applied without context. A 12-room boutique B&B with no restaurant operates on entirely different cost dynamics than a 180-room conference hotel with three food outlets and banquet service. The goal here is to convert every line item in audited P&Ls or management accounts into a driver-linked cost, where expense magnitude is explicitly tied to a measurable, property-specific operational metric.

Start by isolating the core drivers unique to the asset: total rooms, average room size (m²), number of food covers served weekly, public area square metres, number of full-time equivalent (FTE) staff, and guest-night volume. Then map each historical cost category against its most causally relevant driver. For example:

Laundry cost per occupied room-night: Extract 12 months of linen laundry invoices and divide total spend by total occupied rooms. If the property recorded £142,000 in laundry costs across 38,500 occupied rooms, the driver is £3.69 per occupied room—*not* a percentage of room revenue.
Electricity per square metre: Pull utility bills and floor plans. A 2,400 m² hotel spending £68,000 annually on electricity yields £28.33/m²/year—far more actionable than ‘utilities are 4.2% of total revenue’.
Front desk labour hours per 100 rooms: Review rosters and payroll summaries. If two full-time front desk agents work 38 hours/week across 140 rooms, that’s 3,952 annual hours ÷ 1.4 = 2,823 hours per 100 rooms.

Crucially, validate driver logic using *cross-source triangulation*. Does the reported linen usage align with room count × average linen set per room × turnover frequency? Does the reported food covers match POS system totals, not just kitchen labour hours? Discrepancies signal data gaps—not modelling errors—and must be flagged before proceeding. Avoid aggregating costs prematurely: split ‘maintenance’ into ‘preventative HVAC servicing’, ‘reactive plumbing call-outs’, and ‘carpet cleaning’—each with its own driver (e.g., HVAC cost per tonnage of cooling capacity; carpet cleaning per m² cleaned quarterly). This granularity exposes hidden leverage points: a high cost per occupied room on laundry may reflect inefficient outsourcing, while low electricity/m² could indicate outdated lighting still on legacy tariffs. Never assume consistency across periods—adjust for one-off items (e.g., storm damage repairs, temporary staffing during renovation) by annotating them as non-recurring in your source log. This step transforms static financials into a living cost anatomy—ready for dynamic recalibration in Step 2.

Step 2: Build the Core Modelling Framework in Excel or Google Sheets

A robust forecast model is modular, transparent, and auditable—not a single spreadsheet crammed with nested formulas. Build it across six clearly labelled tabs: Drivers, Base Cost Inputs, Contract Rollover Calendar, Wage Escalation Rules, FF&E Reserve Schedule, and Scenario Outputs. Each tab serves a distinct function, and all inter-tab references use structured, traceable functions—not hardcoded numbers.

The Drivers tab holds your validated metrics from Step 1: e.g., ‘Laundry cost per occupied room-night = £3.69’, ‘Electricity cost per m² = £28.33’, ‘Front desk FTE hours per 100 rooms = 2,823’. These are *constants*, not assumptions—derived directly from historical data.

The Base Cost Inputs tab houses all forward-looking variables: projected occupancy, ADR, food covers, wage rates, utility tariff bands, and contract start/end dates. Use XLOOKUP (not VLOOKUP) to pull the correct utility rate based on contract end date—e.g., `=XLOOKUP(TODAY(),ContractStart,RateBand,0,-1)` ensures automatic rate switching at renewal. For wages, separate hourly rates by role (receptionist, sous chef, maintenance tech) and link to the Wage Escalation Rules tab, which defines annual uplift % by job band and effective date.

The Contract Rollover Calendar tab lists every supplier agreement—energy, telecoms, linen hire, security, waste collection—with start date, end date, notice period, and auto-renewal status. This powers conditional cost resets in Base Cost Inputs.

Common pitfalls to avoid: double-counting insurance (e.g., including both property insurance *and* business interruption cover under ‘insurance’ without clarifying overlap), misallocating shared overhead (e.g., assigning 100% of head office IT support to one property instead of apportioning by room count or revenue share), and hardcoding tax percentages without jurisdiction toggles (addressed in Step 3). All formulas must be visible, documented in-cell comments, and tested with zero-input scenarios (e.g., 0% occupancy should yield near-zero variable costs but retain fixed elements like building insurance and core security). Use data validation dropdowns, not free-text cells, for contract statuses (‘Renewing’, ‘Open Negotiation’, ‘Terminating’) to prevent input errors. Finally, protect input cells—but leave formula cells unlocked for due diligence reviewers to audit logic. This architecture ensures the model responds accurately to changes in occupancy, wage rules, or supplier terms—without requiring rewrites.

Step 3: Embed Jurisdiction-Specific Rules Without Overgeneralising

Global buyers cannot apply a single tax or compliance framework across acquisitions in Lisbon, Brisbane, or Nashville. Yet inserting country-specific rules shouldn’t fracture the model’s usability. The solution is explicit jurisdictional labelling and toggle-based activation—not universal defaults. Create a dedicated Jurisdiction Settings tab with a dropdown selector (e.g., ‘UK’, ‘USA – State: California’, ‘Australia – NSW’, ‘Canada – Ontario’) and build all legally contingent calculations behind IF statements tied to that selection.

For example:

UK Business Rates Calculation: In England and Wales, business rates are assessed by the Valuation Office Agency (VOA) and billed by local councils. The model must pull the property’s Rateable Value (RV) from the sale pack, then apply the current poundage (e.g., £0.497 per £1 RV), *plus* any small business relief (SBR) eligibility (e.g., 75% relief if RV < £12,000). This logic lives only behind `IF(Jurisdiction="UK", [formula], 0)`.

US State-Level Workers’ Compensation Premiums: Premiums vary by state, industry classification, and claims history. A hotel in Florida pays ~$1.42 per $100 of payroll for housekeeping staff, while in Washington State it’s $2.89. The model references a lookup table keyed to State + NAICS code (721110 for hotels), activated only when ‘USA’ is selected.

Australia Fair Work Commission Award Updates: Minimum wage increases for hospitality roles (e.g., Cook Level 3, Housekeeper Grade 2) are mandated biannually. The model pulls award-defined base rates *only* when ‘Australia’ is selected—and flags upcoming review dates from the FWC website (e.g., first Sunday in July) via a conditional alert.

Never embed jurisdictional logic inside cost-line formulas. Instead, build standalone calculation modules—e.g., a ‘Payroll Tax Calculator’ tab that activates only for Australia (where payroll tax applies above $1.3M annual wages) or for UK employers liable for Apprenticeship Levy (0.5% on payrolls > £3M). Also include compliance guardrails: if ‘UK’ is selected, the model validates that Employers’ Liability insurance meets the statutory minimum of £5M and flags missing Certificate of Insurance upload. These toggles ensure the same model supports due diligence in 12 countries—without exposing users to misapplied legal assumptions. Full transparency comes from documenting *which* jurisdictional sources were used (e.g., ‘UK business rates poundage sourced from GOV.UK guidance HS281’) in the Assumptions Appendix—not buried in cell notes.

Step 4: Run Targeted Sensitivity Tests That Reflect Real Operational Risk

Generic sensitivity analysis—such as ±10% on all costs—fails to expose true acquisition risk. Real-world volatility is lumpy, asymmetric, and driver-specific. Focus testing on three high-impact, non-linear scenarios that directly affect breakeven viability: occupancy shock, wage inflation tiering, and supplier contract exposure.

Occupancy shock testing must account for seasonality—not just annual averages. Model three distinct demand profiles: *shoulder season* (e.g., March–April, 62% occupancy), *peak season* (July–August, 94% occupancy), and *off-season* (November–January, 41% occupancy). Then simulate a 20% drop *within each segment*: off-season falls to 33%, peak drops to 75%. Variable costs respond differently—laundry and housekeeping scale sharply with occupied rooms, while credit card fees compress as ADR rises in peak season. Output the resulting breakeven occupancy threshold for each scenario: the minimum occupancy needed to cover all cash operating costs (excluding debt service). A property requiring 68% occupancy to break even in off-season—but only 49% in peak—reveals structural vulnerability to prolonged low-demand periods.

Wage inflation tiering recognises that frontline roles (housekeepers, porters) face higher escalation pressure than managers. Apply separate annual uplifts: 6.2% for hourly staff (aligned with national minimum wage trends), 3.8% for salaried supervisors, and 2.5% for GM-level roles. Test compounding over five years—then overlay a ‘strike risk’ scenario where frontline wages jump 11% in Year 3 due to collective bargaining. Track how this shifts the cash flow inflection point: the month when cumulative net operating income turns negative post-acquisition.

Supplier contract exposure tests the financial impact of non-renewal. Identify contracts expiring in Years 1–3 (energy, telecoms, linen, security). Simulate a 12-month gap where the property reverts to default tariffs or spot-market pricing—e.g., UK energy default tariff 28% above negotiated rate; US telecoms spot pricing 40% higher. Quantify the annual cost delta and flag which suppliers represent >15% of total opex. If linen hire accounts for £210,000/year and renewal fails, the model shows an immediate £59,000 annual shortfall—triggering renegotiation urgency. These tests don’t just produce numbers—they reveal *where* and *when* the investment becomes operationally fragile. Document each test’s assumptions clearly: ‘Off-season occupancy shock assumes no marketing response or rate adjustment’—so reviewers understand boundary conditions.

Step 5: Validate, Document, and Present the Model for Due Diligence Review

A model is only as credible as its verification trail. Lenders, solicitors, and internal investment committees will scrutinise assumptions—not just outputs. Follow this six-point validation checklist before finalising your model:

Source traceability: Every cost line in the Base Cost Inputs tab must reference a specific document—e.g., ‘Electricity rate: Page 7, Utility Contract Ref #LH-2241, dated 12/03/2023’ (note: do *not* include the year in live documentation—use ‘Contract dated [DD/MM]’ or ‘Contract ref #LH-2241’). No ‘estimated’ or ‘assumed’ labels without sourcing.
FF&E reserve alignment: Cross-check accrual rates against the physical asset condition report. If carpets are rated ‘poor’ and due replacement in 18 months, the model must accrue at least £X/m² to cover full replacement—not the standard 4% of gross revenue.
Management fee fidelity: Verify fee structure (e.g., ‘2% base fee + 5% incentive fee on NOI above target’) matches the signed management agreement—not a generic industry template.
Tax treatment consistency: Confirm VAT/GST treatment (e.g., ‘UK hotel accommodation is standard-rated at 20%’; ‘Australian GST applies to room nights but not to long-term lease components’) aligns with jurisdictional settings and is applied *only* to taxable revenue lines.
Labour cost completeness: Ensure all statutory employer costs are included—not just gross wages. In the UK, this means adding 15.05% Employer NICs; in Australia, 11% Superannuation Guarantee *plus* payroll tax if applicable.
Non-recurring exclusion: Audit every historical cost for one-offs: litigation settlements, insurance claim reimbursements, or pandemic-related grants. These must be excluded from baseline forecasts and listed separately in the appendix.

All assumptions belong in a dedicated Assumptions Appendix—a standalone PDF or tabbed sheet *outside* the live model. Never bury critical logic in cell comments or hidden rows. The appendix should list: (a) source documents reviewed, (b) driver derivation methodology (e.g., ‘Laundry cost per room derived from 12 months of Hertz Laundry invoices, verified against linen inventory logs’), (c) jurisdictional rules applied, and (d) sensitivity test parameters. For presentation, export Scenario Outputs as clean, printer-friendly dashboards: one-page summary showing base case, worst-case occupancy shock, and supplier gap impact—highlighting breakeven thresholds, cash runway months, and key risk triggers. This level of rigour transforms the model from a forecasting tool into a due diligence artefact—trusted by banks, legal teams, and board members alike. It also forms the foundation for post-acquisition performance tracking: compare actuals against the validated baseline, not against generic benchmarks.

What are the most commonly overlooked variable costs when forecasting hotel operating expenses?

Seasonal maintenance contracts, credit card processing fees, and outsourced labour costs (e.g., seasonal staffing agency premiums) are frequently underestimated

How do you adjust utility cost projections for hotels with inconsistent historical data?

Back-calculate using industry benchmarks (e.g., ENERGY STAR’s hotel metrics) for kWh per occupied room, then apply local utility rates. For water/sewer, use the

What’s the most effective way to model payroll costs for unionised vs non-unionised hotel staff?

Union contracts mandate fixed annual wage escalators (typically 2-4%), benefit contribution changes, and strict overtime rules—build these into cell formulas. F

How should renovation periods be factored into operational cost forecasts during ownership transitions?

Create a parallel ‘Phased Operations’ scenario modelling reduced revenue streams against ongoing fixed costs (insurance, security, base utilities). Include cont

What property-specific red flags indicate unreliable historical cost data?

Watch for round-number repeating expenses (suggesting estimates rather than actuals), sudden dips in routine maintenance costs pre-sale, or missing line items l

Related Resources

Browse hospitality properties for sale | List your property | Free valuation