Stop chasing the 138% figure: it is AirDNA's 2025 national average STR premium, meaning short-term rental revenue ran about 138% above long-term rent on the same kind of unit, down from a 214% peak in 2021, according to Host Camp's summary of AirDNA data. Start instead with the guest and the experience you can deliver, then pick the geography and underwrite it to profit, because 138% describes gross revenue, not what you keep.
That number circulates in every STR Facebook group right now, and it's real. I don't dispute it. I just don't underwrite to it.
I've been in real estate for more than 18 years and focused on short-term rentals since 2016, and I've trained more than 10,000 operators. The ones who make this business work long-term don't start with "where can I get 138%?" They start with a different question: who am I building this for? The gap between those two starting points is the gap between building a business and building yourself another job.
What the 138% STR Premium Measures (and What It Leaves Out)
Read the figure precisely. A 138% premium means that if a unit rents for $2,000 a month on a 12-month lease, the average STR version of that unit grossed about $4,760 a month ($2,000 × 2.38). That is revenue. It comes before platform fees, cleaning, utilities, supplies, software, repairs, and your own time. A long-term tenant pays most of those costs; a short-term guest does not.
The figure is also a national average. It tells you nothing about your specific costs, your competitive position, or whether the market you're eyeing has room for one more operator. Host Camp's same summary lists individual markets where the premium runs well above the average and, by implication, many where it runs well below. An average blends both.
Then there's the trend. The premium fell from 214% in 2021 to 138% in 2025. Same national dataset, a little over a third less spread between STR revenue and rent. The operators who got in early didn't need to be good; they needed to be present. That window is closed.
Every operator hits a price ceiling set by what guests will pay. What you keep under that ceiling depends on how tight your operation is and whether the unit was built for a specific guest. It does not come from finding the highest headline number on a map.
The Market-First Trap
Here's how most new operators approach the market decision. They pull up a data tool, sort cities or zip codes by revenue or premium, pick the top of the list, and ask "how do I get in?" That's starting with the product instead of the person. In almost any other business category, that's considered backwards.
When you start market-first, you end up with a property optimized for a statistic instead of a human being. When the statistic changes (rates compress, a city tightens its ordinance, twenty new listings open within a mile), you have nothing underneath you. The guest was an afterthought, so there is no loyal guest to fall back on.
The table below shows the order in which each operator makes five core decisions. The questions overlap; the sequence changes what you end up owning.
Decision order | Market-first operator | Experience-first operator |
|---|
1st decision | Which city shows the highest premium? | Who is the guest, and why are they traveling? |
2nd decision | What property can I get into there? | What experience does that guest pay a premium for? |
3rd decision | How do I furnish it to match the comps? | Which locations are anchored to that experience? |
4th decision | Who books it? (answered by the listing) | Do the comps and local rules support it? |
5th decision | Does it still cash flow after costs? | Does it clear rent plus operating costs with margin? |
Notice where the guest shows up. The market-first operator meets the guest at step four, after the lease is signed. The experience-first operator defines the guest before looking at a single listing.
The Experience-First Framework (The Billy Bob Principle)
Picture a Billy Bob's Alligator and Airboat tour somewhere in the Florida Everglades. Billy Bob isn't sitting around asking which market has the best RevPAR. He's asking: who is the person who wants this exact experience, and how do I make it extraordinary for them?
The lodging around that experience follows. The geography was decided by the experience. A family driving in for a morning airboat ride needs somewhere to sleep the night before, somewhere that sleeps six, has parking for a truck, and doesn't mind muddy boots at the door. An operator who starts with that family builds a very different unit than one who starts with a city-level revenue average.
Field of Dreams logic applies: build it for the right person, and they come. This is how I've evaluated new STR markets for years. I don't start with the market. I start with the experience and the person who wants it. Location comes third.
What changes when you work this way:
The listing is easier to write. You know who you're talking to, so the photos, the amenities, and the house rules all point at one guest.
Your pricing floor rises. A guest who finds exactly what they came for compares you to fewer alternatives.
Your reviews get specific. "Perfect for our airboat trip" helps the next family decide. "Nice place" does not.
Your repeat bookings climb. People who travel for the same reason every year come back to the place that got it right.
The Experience-First Market Selection Process
Use this sequence when you're choosing your next market. Do the steps in order; skipping ahead to the numbers is how operators slide back into market-first habits.
Define the guest. Write one paragraph describing a single guest: who they travel with, why they're traveling, what they spend money on, and what ruins a trip for them. The SBA's market research guidance is a useful template for this: start with who your customer is.
Name the experience. List the specific thing that guest travels for: a tournament, a lake, a hospital rotation, a wedding venue, a trail system, an airboat tour. If you can't name it in five words, you don't have a niche yet.
Map the geography to the experience. Find the places anchored to that experience, then narrow to the drive time or walk time your guest will accept. Only now do you open a data tool, and you open it to study specific streets and zip codes, not to rank cities.
Check comps and rules. Pull 10 to 20 comparable listings that already serve your guest and note their nightly rates, minimum stays, and amenities. Then read the local short-term rental ordinance and permit requirements; Airbnb's responsible hosting page for the United States is a starting point, but the city code is what binds you.
Underwrite to profit, not premium. Get the long-term rent for the unit (HUD's Fair Market Rents give you a public benchmark), estimate conservative STR revenue from the comps, and subtract every operating cost. If the unit doesn't clear with margin at a conservative revenue estimate, walk away, no matter what the premium says.
What 138% Looks Like After Expenses
Here is the worked math, using the $2,000 rent and 138% premium from earlier plus two comparison cases. All costs are illustrative for a mid-size arbitrage unit and will differ in your market. The platform fee uses Airbnb's single-fee structure, where the fee comes out of the host payout and most hosts pay 15.5%, per Airbnb's service fee page. Guest-paid cleaning fees and occupancy taxes are left out of revenue to keep the comparison clean, but cleaning labor stays in costs as a conservative assumption; if you charge guests a cleaning fee, it offsets part of that line.
Monthly line item (illustrative) | 100% premium | 138% premium | 200% premium |
|---|
Gross STR revenue | $4,000 | $4,760 | $6,000 |
Rent | $2,000 | $2,000 | $2,000 |
Airbnb single host fee (15.5%) | $620 | $738 | $930 |
Cleaning labor (10 turns × $95) | $950 | $950 | $950 |
Utilities and internet | $350 | $350 | $350 |
Supplies and consumables | $150 | $150 | $150 |
Software and pricing tools | $60 | $60 | $60 |
Repairs and replacement reserve | $150 | $150 | $150 |
Total costs | $4,280 | $4,398 | $4,590 |
Net before your time and taxes | -$280 | $362 | $1,410 |
Check the rows: in the 138% column, $2,000 + $738 + $950 + $350 + $150 + $60 + $150 = $4,398, and $4,760 − $4,398 = $362. The fee is 15.5% of gross revenue, so $4,760 × 0.155 = $737.80, rounded to $738.
Three things jump out.
First, the national average premium barely clears in this example. $362 a month is about $4,300 a year before your own hours, and one broken appliance or a slow month wipes it out.
Second, you can compute the break-even. Non-rent, non-fee costs here total $1,660 ($950 + $350 + $150 + $60 + $150). Break-even revenue R satisfies R − 0.155R = $2,000 + $1,660, so R = $3,660 ÷ 0.845 = about $4,331. That's roughly 2.17 times rent, or a premium of about 117%. Any unit that can't reliably beat that line loses money every month in this cost structure.
Third, the gap between the 138% column and the 200% column is $1,048 a month, and it comes entirely from revenue. Costs rose only $192, all of it platform fee. That is where the experience-first approach earns its keep: a unit built for a specific guest has a better shot at pricing above the market average, because that guest is comparing it to fewer alternatives.
Two cautions. Cleaning labor rises with turnovers, so a higher-revenue unit with shorter stays may carry more than 10 turns. And for tax purposes, the IRS rules on rental income, expenses, and personal use in Publication 527 affect what you keep; talk to a tax professional before you count the bottom row as yours.
Market Matured, Not Saturated
The market has matured, which is different from saturated.
Saturation means there's no room. Maturity means you have to be good to win. In a young market, guests book whatever is available and forgive rough edges. In a mature one, they compare you to a dozen alternatives, read the reviews, and book the one that fits their trip. The falling premium is what maturity looks like in the data.
In my experience, the operators struggling right now mostly started with geography. The operators doing well mostly started with the guest. You don't need to be in the highest-premium market. You need to be in the right market for the experience you're building, and those are rarely the same place.
The failure modes I see most often in a maturing market:
Copying the comps. Furnishing to match the top listing means you are, at best, tied with it.
Underwriting to the average. Using a city-wide or national premium instead of the comps your guest would compare you to.
Ignoring the rulebook. Signing a lease before reading the ordinance, then learning the zone caps permits.
Pricing without a guest in mind. Discounting to fill nights instead of building the unit that a specific guest pays full rate for.
Where the Framework Compounds: Seven Properties and Up
This framework gets more useful the more units you run.
When you're running one or two properties, it feels like real estate. In my view, somewhere around seven properties it starts to feel like a business, because the systems, the team, and the guest relationships compound in ways a single unit can't support. That threshold is my own observation from operating and coaching, not an industry statistic, but it has held up across a lot of portfolios.
A clear guest identity helps in three ways once you're past a handful of units:
Acquisition gets faster. You stop evaluating every deal from scratch. The guest profile screens out most properties before you run numbers.
Operations get repeatable. The same guest needs the same amenities, the same welcome guide, the same check-in flow. Your cleaners and your checklists stop changing with every address.
Concentration risk drops. If your guest identity is clear, you can serve that same person in more than one market. A regulation change in one city stops being an existential event for the whole portfolio.
If you want a second set of eyes on how this applies to your own portfolio, book a one-on-one diagnostic call. It's free, and if you want the plan written out afterward, the $997 action plan builds it with you.
Frequently Asked Questions
Should I still use the 138% premium data when evaluating a market?
Yes, but as a benchmark, not a goal. Use market data to stress-test your numbers after you've defined the guest and the experience. In the worked example above, a unit needed roughly a 117% premium just to break even, so the 138% average is close to the line, not a cushion.
Is the 138% STR premium the same as profit?
No. It measures how far STR gross revenue ran above long-term rent, averaged nationally for 2025. Platform fees, cleaning, utilities, supplies, software, and repairs all come out after that. In the illustrative table, $4,760 of monthly revenue at a 138% premium left $362 before your time and taxes.
What if the experience I want to build doesn't have obvious comp data?
That can be a sign you're onto something. New experiences have thin comp data because the market doesn't have a category for them yet. Underwrite conservatively using the closest comparable listings, and if the unit still clears at that conservative estimate, being first lets you set the rate.
Is rental arbitrage still viable as an entry strategy in 2026?
Yes. You don't need to own property to build an STR portfolio. What changed is the execution bar: the premium dropped from 214% in 2021 to 138% in 2025, so you need a clear guest identity and a differentiated unit from day one.
How many units do I need before the experience-first framework pays off?
The approach works at one unit. In my experience it compounds at around seven or more, when your guest profile starts driving acquisition decisions and your operations become repeatable across addresses. Below that, the payoff is sharper listings, higher rates, and more specific reviews.
Which Airbnb fee should I use when underwriting?
Use the structure that applies to your account. Under Airbnb's single-fee structure, the whole fee comes out of your payout and most hosts pay 15.5%; under the older split-fee structure, which Airbnb is phasing out, most hosts pay 3% and the guest pays the rest. The worked example uses 15.5% because Airbnb is transitioning all home hosts to the single fee.
Sources
Related Reading
