How to Estimate STR Revenue Before You Buy

Most short-term rental deals go wrong at the underwriting stage, not the operating stage. A buyer finds a projection that says $78,000, buys on it, and then spends two years discovering what was left out. The fix is not a better data source — it is building the estimate yourself, in three scenarios, with the expense side written before the revenue side.

What you'll learn

The four numbers that actually drive a revenue estimate
How to build a comp set that means something
Why seasonality and booking pace break annual averages
What third-party estimate tools are good for, and what they are not
How to build conservative, base, and upside cases you can defend

Start with four numbers, not one

Every credible estimate reduces to four inputs.

ADR (average daily rate) is revenue divided by nights booked — your price per sold night.

Occupancy is nights booked divided by nights available. Note the denominator: if you block dates for personal use, your occupancy math changes and so does your revenue.

RevPAR is ADR multiplied by occupancy. This is the number to underwrite on, because it punishes both empty nights and underpricing. A property at $400 ADR and 35% occupancy ($140 RevPAR) loses to one at $210 and 75% ($158 RevPAR), even though the first sounds more impressive.

Length of stay determines how many turnovers you pay for. Two 7-night bookings and fourteen 1-night bookings can produce identical revenue and wildly different profit.

Annual gross revenue is roughly RevPAR × 365. Everything else is refining those inputs.

Build a comp set you would actually lose a guest to

A comp set is not "listings in the same zip code." It is the handful of properties a guest choosing your place would click on instead. Match on:

Bedroom count and true sleeping capacity (beds, not couch capacity)
Amenity tier — hot tub, pool, view, walkability
Property type and vibe (a cabin and a downtown condo do not compete)
Distance to whatever generates the demand: beach, park entrance, convention center, hospital, ski lift

Six to twelve genuine comps beats fifty loose ones. Pull their calendars over several weeks and note which dates go from open to booked. That is real occupancy data, observed rather than modeled.

Seasonality: the reason annual averages lie

Many markets earn most of their revenue in a short window. A property that grosses $70,000 might do $42,000 of it between June and September. That matters in three ways:

Cash flow. Your mortgage does not take the winter off.
Debt service coverage. Lenders and your own nerves care about the worst quarter, not the average one.
Sensitivity. If peak season is most of the year's revenue, a rainy summer or a new competing supply wave hits harder than an annual model suggests.

Model month by month. Twelve rows in a spreadsheet is enough, and it will change how the deal looks.

Booking pace tells you what a snapshot cannot

Booking pace is how quickly future nights fill. In some markets, summer is 70% booked by March; in others, half of all reservations arrive within 14 days of check-in. Watch a few comps' calendars for 30 days and record how far out reservations land.

Pace matters after you buy — it is how you know whether a soft-looking month is a pricing problem or a normal short-lead market — but it also tells you how much working capital you need in year one, when you have no history and no reviews.

Write the expense side first

This is the step most spreadsheets skip. Typical recurring costs on a single short-term rental:

Expense	Typical shape	Notes
Cleaning	Per turnover	Usually passed to guests, but not always fully
Consumables and supplies	Per turnover + monthly	Paper goods, coffee, soap, batteries
Utilities	Monthly	Higher than a long-term rental; guests do not conserve
Internet, streaming, smart locks	Monthly	Small individually, real in aggregate
Insurance	Annual	STR-specific policy, not a standard homeowner's policy
Permits and licenses	Annual	Jurisdiction-dependent
Lodging tax	Per booking	Sometimes collected by the platform, sometimes your job
Platform commission	Per booking	Varies by channel and fee structure
Management or co-hosting	Monthly or %	Flat fee or commission
Repairs and maintenance	Reserve monthly	Higher wear than long-term tenancy
Furnishing replacement	Reserve monthly	Linens, mattresses, and sofas are consumables here
Vacancy and ramp-up	Year one	New listings take months to build review velocity

Reserve for furnishing replacement even in year one. Short-term rental furniture lives roughly a third as long as your own.

Three scenarios, one page

Do not produce a number. Produce a range with the assumptions visible.

Conservative: occupancy 10–15 points below your comp observation, ADR at the low end, a full ramp-up period in year one, and every expense at its high estimate. If the deal still works here, it is a real deal.
Base: your honest read of the comps with a normal ramp.
Upside: strong execution, professional photography, active pricing management, a good review run. Treat this as the ceiling, not the plan.

Underwrite on conservative. Celebrate base. Never buy on upside.

Key takeaway: if the deal only works in the upside case, you are not buying a rental — you are buying a bet on perfect execution in a market you do not operate in yet.

Where third-party estimate tools fit

AirDNA, Rabbu, and STR Insights (STRIQ) all publish market-level and property-level estimates derived from scraped or partner calendar data. They are genuinely useful for three things: identifying seasonality shape, sizing a market's supply, and sanity-checking your own comp work.

They are not guarantees, and the providers say so themselves — each publishes methodology notes explaining that estimates are modeled from observed availability rather than reported revenue. Two common distortions to watch: blocked owner dates can read as booked, and a market's top decile of professionally run listings can pull an average well above what a new listing will do in year one.

Use them as one input among several. When a tool's number and your comp observation disagree by more than 20%, the disagreement is the finding — go figure out which assumption is wrong.

Talk to someone who operates in that market

The most underrated diligence step costs nothing: call a local co-host, manager, or cleaner. Ask what a realistic first-year occupancy looks like, when the shoulder seasons actually start, what breaks in these houses, and what the permit office has been doing lately. Local operators know things no dataset captures, including new supply that has not been listed yet.

Want a second set of eyes? Email go@vacohost.com with the address and your assumptions, and we will build a conservative / base / upside revenue projection and tell you where we think you are wrong.

FAQ

Q What occupancy should I assume for a brand-new listing? A: Assume you will underperform comps for the first three to six months while you build reviews and booking history. Many operators model year one at 10 to 15 occupancy points below an established comparable, then step up.

Q Is AirDNA accurate? A: It is a model built on observed availability, so it is directionally useful and property-specific numbers can be well off. Treat it as one input to check against your own comp calendar observation, and read the provider's methodology notes.

Q Should I use gross revenue or net when comparing properties? A: Net. Two properties with identical gross can differ by thousands after cleaning frequency, lodging tax treatment, utilities, and management structure. Gross revenue is a headline, not a result.

Sources and further reading

AirDNA: market data and methodology — provider documentation for modeled STR estimates.
Airbnb Help Center: pricing your listing — official guidance on rate settings and discounts.
Rabbu STR revenue calculator — a second modeled estimate for cross-checking.