Opens in a new tab

Blog / Hospitality & Travel

Hospitality & Travel

What is RevPAR? The RevPAR, ADR and occupancy formulas, with worked examples for hotels and serviced apartments

Nov 19, 2026·Updated ·14 min read

A small hotel owner reviews the weekly room revenue report and RevPAR figures at the reception desk

RevPAR (Revenue Per Available Room) is the room revenue a property earns for every room it could have sold in a period. You calculate it by dividing total room revenue by the total number of available rooms, or by multiplying the average daily rate (ADR) by the occupancy rate. It is the one figure that answers both questions an operator has to answer every week: are we selling the rooms, and are we selling them at the right price. A property can be full and still post a weak RevPAR if rates were cut too far, and it can hold high rates and still post a weak RevPAR if too many rooms sat empty.

This article is part of InCard’s hospitality operations glossary. It explains the formulas behind all three metrics, works through two complete numerical examples for a small hotel and a serviced apartment block, and then covers the reading mistakes operating teams make most often. A weekly RevPAR tracking table you can start using today sits near the end. Figures are in Vietnamese dong, with rough US dollar equivalents at about 26,000 dong to the dollar.

Key points

  • Occupancy measures selling power: rooms sold divided by rooms available.
  • ADR measures pricing power: room revenue divided by rooms sold.
  • RevPAR equals ADR multiplied by occupancy, so it punishes both errors at once, discounting to fill rooms and holding rates so hard that rooms stay empty.
  • RevPAR only means something against your own property in a comparable period, or against a market benchmark in the same segment and area.
  • RevPAR ignores costs and non-room revenue. It is a revenue metric, not a profit metric.

The three base formulas: occupancy, ADR and RevPAR

All three metrics run on two inputs: the number of rooms you could have sold in the period, and the net room revenue for that period. Getting those two inputs right is the hard part. The formulas themselves are one division each.

Occupancy rate is the share of available rooms that were actually sold:

Occupancy = Rooms sold / Rooms available × 100%

Rooms available means the property’s room count multiplied by the days in the period. A 30-room hotel in a 31-day month has 930 available room-nights. Rooms under renovation or blocked from sale long-term should come out of the denominator, and whichever rule you choose has to stay identical from one period to the next.

ADR (Average Daily Rate) is the average rate actually collected per room sold:

ADR = Room revenue / Rooms sold

Room revenue here is room money only. Food and beverage, laundry, transfers and other add-ons do not belong in the numerator. Put them in and the result is no longer ADR.

RevPAR connects the two:

RevPAR = Room revenue / Rooms available = ADR × Occupancy

How the three metrics relate:

Room revenue ──÷── Rooms sold ─────────► ADR (are we pricing right?)

│

└───────÷── Rooms available ────► RevPAR (what does each available room earn?)

Rooms sold ────÷── Rooms available ────► Occupancy (how much did we sell?)

How the mechanism works: all three metrics share the same two inputs, room revenue and room counts. The difference is the denominator each one divides by.

The point worth remembering: RevPAR’s denominator is every available room, including the empty ones. That is why one empty night drags RevPAR down immediately, while the ADR for that night still looks healthy, because ADR only ever looks at rooms that sold.

Example 1: a 30-room hotel calculates RevPAR for one month

Take a 30-room hotel in Da Nang in October, a 31-day month. The figures below are an illustrative example built to show the calculation, not data from a real property.

Item Value
Rooms at the property 30 rooms
Days in the period 31 days
Rooms available (30 × 31) 930 room-nights
Rooms sold 651 room-nights
Average rate collected (ADR) 850,000 VND (about US$33)
Room revenue (651 × 850,000) 553,350,000 VND (about US$21,300)

Inputs for example 1: one 31-day month at a 30-room hotel, at exactly 70 percent occupancy.

Working through each metric:

  • Occupancy = 651 / 930 = 70%.
  • ADR = 553,350,000 / 651 = 850,000 VND (matches the collected average rate).
  • RevPAR from revenue = 553,350,000 / 930 = 595,000 VND (about US$23).
  • RevPAR from the product = 850,000 × 70% = 595,000 VND.

Both routes land on the same 595,000 dong. If your two routes ever disagree, one of the inputs is being counted differently somewhere, usually in what counts as an available room.

What the figure means in operating terms: on average, every room in the hotel, empty ones included, produced 595,000 dong per night. That is the number to set against October last year, and against the festival week and the storm week inside the same month, instead of looking at 70 percent occupancy alone and deciding whether the month went well.

Example 2: where a 20-unit serviced apartment block calculates RevPAR differently

Now take a 20-unit serviced apartment block renting by the night and by the month, over a 30-day month. This is also a built illustrative example.

Item Value
Units for rent 20 units
Days in the period 30 days
Units available (20 × 30) 600 unit-nights
Unit-nights sold 480 unit-nights
Average rate per unit-night sold 1,150,000 VND (about US$44)
Stay revenue (480 × 1,150,000) 552,000,000 VND (about US$21,200)

Inputs for example 2: one 30-day month at a 20-unit serviced apartment block, mixing nightly guests and monthly tenants converted back into unit-nights.

  • Occupancy = 480 / 600 = 80%.
  • RevPAR = 552,000,000 / 600 = 920,000 VND (about US$35), and the cross-check agrees: 1,150,000 × 80% = 920,000 VND.

Two things change when the formula moves to serviced apartments, a segment where the operating detail differs enough that we keep a separate guide on serviced apartment management. First, a monthly tenant pays one lump sum for the whole period, so before anything can be calculated the contract has to be converted back into the actual nights stayed. Second, apartment blocks usually also track revenue per unit per month, because monthly contracts make nightly ADR look lower than the business really is. Even so, RevPAR remains the fastest comparison between two blocks, or between one block and its own past seasons.

One counting trap is specific to this segment: the gap nights between a monthly tenant leaving and the next one arriving. Those nights are available inventory, so they stay in the denominator and pull RevPAR down, even though no sales effort could realistically have filled a two-night gap between contracts. Blocks that want a cleaner read keep a second line in the weekly sheet, RevPAR excluding contract gaps, and watch both. The headline figure stays the strict one; the second line just stops the team chasing a gap that was structural rather than commercial.

Where RevPAR, ADR and occupancy differ

Metric Question it answers Formula Blind spot
Occupancy How much did we sell? Rooms sold / Rooms available Says nothing about the price achieved; a full house bought with discounts still looks good
ADR Are we pricing right? Room revenue / Rooms sold Says nothing about how many rooms sat empty
RevPAR What does each available room earn? ADR × Occupancy Says nothing about the cost of earning that revenue

This table is why a weekly report built on a single metric always ends in an argument between the sales side and the operations side. One worked contrast makes it concrete. Take the same 30-room hotel from example 1 and run the month two different ways.

  • Approach A, discount to fill: 85 percent occupancy, ADR 700,000 VND, RevPAR = 700,000 × 85% = 595,000 VND.
  • Approach B, hold the rate: 70 percent occupancy, ADR 850,000 VND, RevPAR = 850,000 × 70% = 595,000 VND.

RevPAR comes out identical to the dong, yet approach A hosts far more guests for the same revenue: more housekeeping turns, more OTA commission if the extra guests arrived through paid channels, more wear on the rooms. RevPAR cannot judge between the two approaches, because it contains no costs. Judging that gap is the job of GOPPAR, covered in the limits section below.

How to read RevPAR properly: against yourself first, against the market second

There is no universal RevPAR threshold that separates good properties from bad ones. A RevPAR of 595,000 dong is respectable for a three-star hotel in a second-tier city in the rainy season and weak for a beach resort in a holiday week. A sound reading has two layers of comparison.

The first layer is your own property in a comparable period: the same month last year, the same holiday last year, or a rolling average of the last four weeks. This layer strips out seasonality, which distorts every month-on-month comparison in hospitality.

The second layer is the market in your segment. In its third-quarter 2026 reporting, Savills put Ho Chi Minh City’s hotel market at nearly 17,000 rooms, with average rates around 2.4 million dong per night (about US$92) and occupancy of 70 percent during what is the tourism low season. The city’s serviced apartment segment ran at about 85 percent occupancy across more than 7,000 units in 146 projects.

Multiplying the two published Savills figures gives an implied average RevPAR of about 1.68 million dong per room per night (about US$65) for the city’s three-to-five-star hotel stock. That 1.68 million is our arithmetic on Savills’ numbers (2.4 million × 70 percent), not a figure Savills publishes directly, and it should be used as a rough market marker only.

Mauro Gasparotti, Senior Director and Head of Hospitality Advisory for Southeast Asia at Savills, described the quarter as one where operating performance eased slightly with the tourism season while occupancy and average rates still ran ahead of the same period a year earlier (as reported by Znews on the Savills third-quarter 2026 report). The detail carries the real lesson: a month where your RevPAR falls against the month before is not automatically a bad month, if the whole market fell with the season and your property held its position against the benchmark.

Larger hotels formalise that second layer as a RevPAR index: your RevPAR divided by the market or competitor-set RevPAR, multiplied by 100. An index above 100 means you took more than your fair share of the market’s room revenue; below 100 means the market earned more per room than you did. Small properties rarely have a true competitor-set feed, but the habit behind the index is available to everyone: never read a RevPAR movement until you know whether the market moved with you. A 5 percent fall in a month when the market fell 12 percent is a gain in position wearing a disguise.

Six mistakes teams make when calculating and reading RevPAR

  1. Counting complimentary rooms, staff rooms and unpaid guest rooms as rooms sold. These rooms leave revenue unchanged while inflating ADR’s denominator, dragging both ADR and RevPAR artificially down.
  2. Changing how available rooms are counted between periods. This month renovation rooms come out of the denominator, next month they go back in, and the team concludes RevPAR rose because operations improved, when only the counting changed.
  3. Folding food, spa and transfer revenue into room revenue. ADR and RevPAR both inflate and stop being comparable to any benchmark. Total revenue per available room is a different metric, called TRevPAR.
  4. Praising the sales team on occupancy alone. A full house bought by deep discounting is exactly the situation RevPAR exists to expose, as the two approaches above show.
  5. Comparing your RevPAR with a property in a different segment or a different region. A beach resort and a city business hotel run opposite seasons; a direct comparison only produces wrong pricing decisions.
  6. Reading the monthly average and ignoring the spread inside the month. A month with a healthy average RevPAR but all revenue packed into eight weekend nights is a completely different staffing and rostering problem from a month with even demand.

How operators use the three metrics, week by week

At properties that run this properly, the three metrics get one fixed review a week, usually on Monday morning for the week just closed, on a rhythm that never changes:

  • Compare this week’s RevPAR with the same week last year and with the trailing four-week average. Any gap beyond about 10 percent in either direction needs a named cause: a festival, weather, a competitor opening, or a sales channel that suddenly stopped.
  • Split weekday and weekend RevPAR. The two groups carry different guests, different rates and different reasons for empty rooms.
  • Calculate net RevPAR by channel: take each channel’s ADR and subtract that channel’s commission. Suppose an OTA charges 18 percent commission on the room rate; a room sold through it at 850,000 dong keeps only about 697,000 dong before it enters RevPAR (an illustrative calculation using an assumed commission rate). The direct channel usually posts a higher net RevPAR at the same headline ADR, which is why properties invest in direct bookings instead of depending entirely on OTAs, and why the after-hours check-in journey matters: it is where direct guests often arrive.
  • Count booking enquiries missed outside staffed hours. Every room enquiry that lands at 23:00 with nobody answering is a room-night at risk, and it shows up in next week’s occupancy. This is where small properties lose revenue most quietly, because no report records conversations that never happened. Properties that run AI self check-in with an AI receptionist covering overnight messages usually find this leak the first time they reconcile overnight conversations against actual bookings.
  • Watch booking pace for the next four weeks, not just results for the last one. Count how many room-nights for each future week were added since the last review. A week whose pace suddenly slows is giving you time to react with a rate or channel move; the same week discovered after it closes is just a bad result to explain. RevPAR tells you what already happened, and pace is the closest weekly proxy for what is about to.

For that weekly rhythm to work, the inputs have to be clean: rooms sold, rooms available and room revenue pulled from one booking record system, not patched together from notebooks and separate spreadsheets. Organising the operating data that way is part of the wider hospitality technology picture, and it is the precondition for trusting any metric in this article.

Weekly tracking table (ready to use)

Week Rooms available Rooms sold Occupancy Room revenue (VND) ADR (VND) RevPAR (VND) RevPAR same week last year Note on the gap
3 to 9 Nov 210 147 70% 124,950,000 850,000 595,000 560,000 Local festival week

The figures in this row are an illustrative example for a 30-room hotel. In real use, keep the columns and never change the counting rules between weeks.

When RevPAR is not enough, or should not be used

RevPAR is a revenue-per-room metric, and its limits sit inside its own definition.

  • RevPAR contains no costs. Two properties with identical RevPAR can be far apart on profit. When the question turns to profit, the matching metric is GOPPAR: gross operating profit divided by available rooms.
  • RevPAR ignores non-room revenue. For resorts and hotels with strong restaurants, spas or event space, TRevPAR (total revenue per available room) reflects the health of the business more honestly.
  • For homestays and small long-stay apartment rentals, the denominator is so small that one booking swings the weekly figure wildly. These properties should read RevPAR monthly or quarterly, alongside revenue per unit, instead of reacting to weekly movement.
  • During renovation, opening phases or partial closure, the available-room denominator moves constantly and every RevPAR comparison loses meaning until the property stabilises again.

Frequently asked questions

How is RevPAR different from ADR? ADR is calculated only on rooms that sold. RevPAR is calculated on every available room, including empty ones. ADR answers whether you priced right; RevPAR answers what each available room actually earned.

What RevPAR counts as good? There is no threshold that fits every property. A good RevPAR is higher than your own property’s same period last year and not behind your segment’s market in your area. For an outside marker, take the average rate and occupancy that research firms such as Savills publish for your market and derive RevPAR the way this article does.

Can a homestay with a handful of rooms use RevPAR? Yes, the formula does not change. With a small denominator, though, read it monthly or quarterly and next to revenue per room, because a single booking moves the weekly number sharply.

Do food and spa revenue count toward RevPAR? No. RevPAR uses room revenue only. If you want food, spa and other services included per available room, the metric you need is TRevPAR.

Should RevPAR be reviewed daily, weekly or monthly? Weekly is the rhythm that is fast enough to fix pricing and rosters and long enough that one unusual night does not distort the picture. Daily review makes sense in peak season and holidays; monthly and quarterly views are for trend and market comparison.

Should taxes and service charges be included in RevPAR? Calculate RevPAR from the same net room revenue you use for ADR, after discounts, and then never change the basis. Whether taxes sit inside or outside that figure matters far less than consistency, because RevPAR is read against your own history. It only starts to matter when you compare with a published benchmark, in which case match the benchmark’s basis before comparing, or the comparison is measuring tax treatment rather than performance.

More from the hospitality operations glossary

If your property still reads these metrics from spreadsheets patched together across sources, the sensible next step is a demo built on your own booking data, so you can watch the three metrics run themselves week by week. Book a demo and bring your room count and the number of selling days in your most recent month; the InCard team will build on exactly that data.

RevPAR improves when the operation behind the rate improves: the serviced apartment running end-to-end AI operations is one worked example of the booking, check-in and in-stay chain behind the number. More pieces like this sit in our hospitality and travel guides.


Tam Ho is the Founder of InCard, the agentic AI platform for small and mid-sized businesses in Vietnam. He works directly with operations teams to bring each chain of work onto the system, with a human keeping the checkpoint on the decisions that matter.

Get advice

Want to apply this to your business?

Leave your details and the InCard team will contact you within 24 hours to look at which job is worth automating first. No cost and no commitment.

Apply it to your business

Want to see this running on your own work?

Book a 30-minute demo — we run it on the channels and processes you already have.