Revenue blog · Metrics

EBITDA per room: the metric your hotel valuation hinges on

By Arshad Kacchi, founder of RevPerfect · 8 June 2026 · 10 min read

The first time a property I was running came up for sale, the acquirer's analyst asked for the trailing-twelve-month accounts, ran a single division on a calculator, and wrote one figure down next to the word EBITDA. Then she circled a second number: EBITDA per room. Everything else on the deck — the rate story, the comp-set story, the channel story — was background music. The price the building eventually sold at was that circled number multiplied by the multiple, and the multiple was set by how confidently you could defend the circled number.

What EBITDA per room actually means in 2026

EBITDA is earnings before interest, tax, depreciation and amortisation. EBITDA per room is that figure divided by the physical key count of the hotel — not by available room-nights, which is the denominator for the operating metrics. It is the cash profit one unit of inventory generates over a period, almost always reported on a trailing-twelve-month basis when a property is being valued.

The reason the metric exists, separately to GOPPAR and to NOI, is that hotel transactions need a number that scales cleanly between a 60-key boutique and a 400-key full-service property. A 60-key boutique running A$28,000 of EBITDA per key is comparable, immediately, to a 220-key resort running A$24,000. Two different revenue bases, two different segment mixes, one shared yardstick.

EBITDA per room is also the cleanest bridge between the operating dashboard and the valuation slide. A revenue manager runs ADR, RevPAR and GOPPAR. The acquirer runs price per key and multiple of EBITDA. EBITDA per room is the metric where the two conversations meet. The shorthand I use with teams: GOPPAR tells you how well the building operates; EBITDA per room tells you what the building is worth. They are the two halves of the same arithmetic.

The EBITDA per room formula (with a worked example)

One formula, two inputs. The complexity is entirely in calculating EBITDA itself.

Working from the top: total operating revenue minus cost of sales, payroll and operating expense produces gross operating profit. From GOP, subtract management fees, property tax, building insurance and any other non-financing fixed charges, and you arrive at EBITDA. Interest, tax, depreciation and amortisation sit below that line — deliberately. Removing them gives a number independent of how the building was financed or how aggressively the owner depreciated the fit-out, which is precisely why buyers prefer it.

The property converts about 44 cents of every revenue dollar into GOP and about 38 cents into EBITDA — a workable margin for a full-service property of this scale. EBITDA per key sits at A$41,667. At a typical 11x multiple, that implies a per-key valuation around A$458,000; at 9x it implies about A$375,000. The multiple does most of the work in headline price, but the EBITDA per key is the lever the operator actually controls. The companion read on the operating layer above this number — and the channel-mix trap that quietly compresses it — sits at ADR vs RevPAR vs GOPPAR.

Where EBITDA per room breaks down (and the trap behind it)

EBITDA per room is a clean number on a clean P&L. The problem is that most hotel P&Ls are not clean, and the dirty edges are exactly where the metric loses the buyer's trust.

First failure: non-recurring items left inside EBITDA. One-off insurance recoveries, asset disposal gains, opening-year ramp losses, a single large legal settlement — all find their way into the operating accounts and quietly inflate or deflate a single year. A trailing-twelve-month EBITDA per room of A$42,000 is meaningless if A$6,000 of it was a one-off that will not recur. Buyers normalise in diligence. Sellers who normalise before the data room opens enter the conversation with a defended number rather than a contested one.

Second failure: management fees missing. Owner-operated independents sometimes report EBITDA without subtracting a market-rate management fee, because the owner is also the operator. A professional buyer will impute one — usually 2 to 4 percent of total revenue as a base, plus an incentive layer. If the property quotes EBITDA per key without that imputed fee, the buyer silently revises the number down before the offer is written.

Third failure: capex confusion. EBITDA excludes depreciation by design, but a hotel with deferred maintenance is generating a flattering EBITDA the building cannot sustain. Buyers apply a furniture, fixtures and equipment reserve — commonly 3 to 5 percent of total revenue — and read the number post-reserve. A property running A$42,000 of EBITDA per key with a 4 percent FF&E reserve is really running about A$38,200 of sustainable EBITDA per key.

EBITDA per room is the metric buyers underwrite against. It is also the metric most easily distorted by the way the P&L is presented. Normalising it before the data room opens is the cheapest valuation work an operator will ever do.

The structural trap underneath all three is the same: EBITDA per room is presented as a fact, but it is a constructed number. Every adjustment compounds. A property with an unimputed management fee, an unrecognised FF&E reserve and one large non-recurring credit can quote A$48,000 of EBITDA per key while the defensible figure is A$36,000. The 25 percent gap is the difference between an offer at multiple-of-defensible and an offer at multiple-of-disputed.

Why owners read EBITDA per room (and most operators don't)

The gap between the operator dashboard and the valuation slide exists for three structural reasons.

Training. Revenue management curricula spend hundreds of hours on RevPAR optimisation and a handful on the lines below GOP. Most accreditation programmes treat property tax, insurance and management fees as background — costs managed by someone else. The result is a workforce who can name the seven drivers of RevPAR and not one of the four lines between GOP and EBITDA.

Data access. RevPAR lives in the PMS. GOPPAR lives in the property accounting system. EBITDA per room lives in the entity accounts, often kept by the owner's accountant. Three files, three teams. Closing the loop takes a deliberate handshake — see the companion piece on owner-grade reporting in hotel revenue management strategies for 2026.

Comfort. EBITDA per room moves on a slower cycle and pulls in lines the revenue manager does not control — utilities, repairs, leasehold renewals, insurance. Speaking the EBITDA language pulls procurement, facilities and finance into the room, and the conversation forces collaboration outside the operator lane. Public-market hotel investors have made the gap explicit for years; World Travel & Tourism Council reporting tracks contribution and earnings, not topline revenue, for the same reason.

What to do about it — a five-step playbook

Closing the gap between the operating dashboard and the EBITDA-per-room number is a reporting layout shift more than a strategy shift. The numbers already exist. They are sitting in three different files.

1. Build the bridge once. A single page that shows total revenue → GOP → EBITDA → EBITDA per room, with the four lines between GOP and EBITDA itemised. Update it monthly with rolling twelve-month numbers. Once it exists, the conversation with the owner shifts from defending RevPAR to discussing EBITDA per key, which is the conversation the owner wanted in the first place.
2. Impute the management fee even if you do not pay one. For owner-operated properties, assume a 3 percent base plus a 6 to 8 percent incentive on EBITDA, and report EBITDA per room both with and without the imputation. This pre-empts the silent buyer adjustment and demonstrates financial literacy in any conversation.
3. Hold an FF&E reserve in the model. 4 percent of total revenue is the conventional figure. Report sustainable EBITDA per room — the number after the reserve — alongside the headline. The first time a broker sees both numbers on the same page, the trust premium is measurable.
4. Normalise non-recurring items in a footnote. Every adjustment over A$25,000 lives in a footnoted line. Subsidies, recoveries, opening-year ramp, one-off legal — all visible, all defended. Buyers normalise either way; the question is whether they trust the operator who did it for them.
5. Forecast EBITDA per room alongside RevPAR. Most demand models stop at occupancy × ADR = RevPAR. Adding a coarse cost stack produces a forecasted GOP and EBITDA per room. The forecasting discipline behind this lives in hotel demand forecasting for revenue managers and pairs with the channel-cost lens in OTA commission rates in 2026. A team that forecasts profit per key has a different relationship with ownership inside two quarters.

A real scenario — anonymised, but it lands every time

At a 90-key urban property I worked with, the team had spent eighteen months pushing RevPAR. The headline was clean: RevPAR up 11.4 percent year-on-year, GOPPAR up 6.1 percent. The owner had agreed to a sale process and the broker was preparing the information memorandum. The in-house accounts presented EBITDA per key at A$32,400.

When we rebuilt the trailing-twelve-month accounts for the data room, three things surfaced. The property had received a A$140,000 one-off insurance recovery. No management fee had been imputed because the owner self-managed. An FF&E reserve had never been booked. Walking the adjustments: minus A$1,556 per key for the non-recurring recovery, minus A$3,800 per key for the imputed fee, minus A$3,444 per key for the FF&E reserve. The defensible figure landed at A$23,600 per key.

The 27 percent revision sounds painful — and it was, briefly — but the alternative was discovering it mid-diligence after the marketing campaign had set expectations against the higher number. The transaction closed at a 10.5x multiple on the defended A$23,600 figure: a per-key price of A$248,000. Had the broker marketed at the unadjusted A$32,400, the likely outcome was a buyer offer at 8.5x on a renegotiated EBITDA per key of A$22,000 — a per-key price of A$187,000. The pre-marketing normalisation was worth roughly A$5.5 million on a 90-key property. The companion read on the operating layer that feeds these EBITDA inputs sits at how to calculate RevPAR.

FAQ

The honest summary

EBITDA per room is the number every professional hotel buyer underwrites against. Not RevPAR, not GOPPAR, not the comp-set indices. The operator who can produce a clean, defensible EBITDA per key — with the management fee imputed, the FF&E reserve held, the non-recurring items footnoted — is the operator the next owner wants. The metric is simple arithmetic. The discipline behind it is what separates a building that sells at its multiple from one that sells at a renegotiation.

At RevPerfect we built the dashboard with the EBITDA bridge already on the page — ADR, RevPAR, NRevPAR, GOPPAR and EBITDA per room running in one continuous line so the owner conversation never starts in the wrong place. Book a 20-minute walkthrough, or try RevPerfect free →.