Air conditioning ROI for short-term rentals: the one-summer payback

A 55m² apartment two blocks from a beach in southern Spain ran at 74% occupancy and a 4.9 rating from October to May. Then came its first July. Three guests in a row left the same sentence in the review — some version of "impossible to sleep, the flat was 31°C at midnight." The rating slid to 4.6, the listing dropped out of the top of search, and the following August's bookings came in 20% under the previous month's rate. A pair of window units that cost €430 in June would have prevented every bit of it.

This is the math on air conditioning for a short-term rental: what it actually costs to install, what it costs to run, and — the part most hosts get backwards — why in a hot market AC is not a premium amenity you weigh for ROI but a cost of being in the market at all. It's the mirror image of the hot-tub math: a hot tub pays back in cold climates and never in hot ones; AC pays back in hot climates and rarely in mild ones.

AC is table stakes, not a premium

Start with the thing that separates AC from every other amenity upgrade: it almost never raises your nightly rate, and that's fine, because that's not its job.

A hot tub works by pulling a smaller, higher-paying pool of guests toward your listing — it adds a premium. AC works the other way. In a market where July highs sit above 30°C, a majority of summer searchers apply Airbnb's "Air conditioning" amenity filter before they see a single listing. Estimates from the markets I've audited put it at 55–75% of peak-season searchers in southern Spain, southern France, inland Italy, Greece, and the US South. If your listing doesn't have the amenity checked, you are simply invisible to two-thirds of the people shopping your dates in the months you make most of your money.

So the mechanism isn't premium-capture. It's loss-avoidance. The AC listing and the non-AC listing next door might charge the same $140 a night in July — but the non-AC one is shown to a third of the audience, books later and thinner, and then absorbs the heat reviews that the AC one never gets. The gap doesn't show up as a higher rate. It shows up as occupancy and rating, which is where the real money is anyway.

That reframing matters because it changes the question. For a hot tub you ask "will the premium repay the install?" For AC in a hot market you ask "what does not having it cost me every July?" — and that number is almost always larger than the hardware.

The install paths, honestly priced

There are four ways to cool a rental, and they span two orders of magnitude in cost. Size the unit first: budget roughly 20 BTU per square foot (about 600 BTU per m²), so a 25m² bedroom needs a ~9,000 BTU unit. Oversizing is a real mistake — an oversized unit short-cycles, never runs long enough to pull humidity out, and leaves the room cold and clammy.

The window unit is where most hosts should start, and not just on price. It goes in the same afternoon you buy it, it needs no electrician, and if a compressor dies in year four you swap the whole box for $250 instead of calling a service company. Its two costs are noise (more on that below) and looks — a window unit photographs as exactly what it is, which in a design-forward listing is a small conversion tax.

The mini-split is the grown-up answer. It's near-silent, roughly twice as efficient, cools evenly, and — the line that flips the whole ROI — it runs in reverse as a heat pump in winter. That means the mini-split isn't only a summer purchase; it shaves your winter heating bill too, which is what lets it out-earn a $400 window unit over a five-year horizon despite costing ten times as much up front.

What it actually costs to run

The running cost is the line hosts either ignore or wildly overestimate. Here's the real math for a window unit.

A 12,000 BTU window unit at an EER of 10 draws about 1.2 kW while the compressor runs. Run it 8 hours across a hot night and that's 9.6 kWh. At $0.20/kWh — a middling rate; EU hosts often pay more, US Sun Belt hosts often less — that's $1.92 a night. It doesn't run every night of the year; it runs the cooling season, call it 60 occupied peak nights for a busy listing. So one unit costs roughly $115 a season, and a one-bed running two units lands at $200–$260.

An inverter mini-split at a SEER of 20 does the same cooling for close to half the electricity because it modulates instead of slamming on and off, so the same coverage costs $70–$140 a season. Over a hot-market year where the thing also heats in winter, the mini-split's operating advantage compounds.

Two things quietly inflate the electricity line, and both are guest behavior. A guest who runs the AC at full blast with the balcony door open will double the bill and ice up the coils; a guest who sets it to 16°C to cool the room "faster" (it doesn't — the unit runs flat out to an unreachable target) does the same. A one-line house note and a thermostat cap or smart plug handle most of it, and the utility-cost breakdown has the full open-window math.

The break-even on a real listing

Take that southern-Spain one-bed. The peak window (June–September) is roughly 120 nights and carries the year's highest rates.

Without AC, in a hot market: you're shown to a third of peak searchers, so you discount to compete and still run thin — call it 60% peak occupancy at a defensive $125. That's 120 × 0.60 × $125 = $9,000 of peak revenue. Then the tax you can't see on the spreadsheet: the heat reviews. Two or three "too hot to sleep" reviews in July drop the rating 0.2–0.3 stars, and a rating dip suppresses bookings for months afterward — the damage bleeds into the shoulder season, plus $40–$150 per mid-stay heat complaint in partial refunds.

With AC: you're in the filtered pool, you hold rate, and the heat reviews stop. Call it 85% peak occupancy at $140. That's 120 × 0.85 × $140 = $14,280.

The peak-season difference alone is +$5,280, before counting the review drag it also removes. Now put the hardware against it:

• Two window units at $520 all-in. Payback: the first weekend of the first summer. Everything after is gravy. This is, without much competition, the highest-return check a host in a hot market ever writes.
• A two-head mini-split at $6,800. Payback against the peak lift alone is a little over one summer — roughly 15 months — and then it keeps earning quietly for a decade while also cutting winter heating by $200–$400 a year.

The window units win on payback speed; the mini-split wins on lifetime value, silence, and the winter heat-pump bonus. For a listing you plan to keep more than three or four years, the mini-split is usually right. For one you're testing, or a room where a window unit fits cleanly, start cheap.

The climate line: where AC stops paying

AC is not universal. The same decision that makes it mandatory in Seville makes it a waste in Cornwall. The dividing line is the July average high and whether the nights cool off.

Table stakes — install it, don't overthink the ROI. Markets with a July high above 30°C and warm nights: inland and southern Spain (Madrid, Seville, Valencia), Mediterranean and inland France, central and southern Italy, Greece, most of the US South and Southwest, the Gulf, Dubai, much of Southeast Asia. Here the question isn't whether AC pays — it's how fast you can get it in before the next heat wave.

Marginal — depends on the building. Markets with a July high of 26–30°C: Paris, Berlin, northern Italy, the Lisbon coast. In an old stone building with thick walls and cross-ventilation you may never need it. But two building features push a marginal market into "install it": a top-floor or west-facing unit that bakes in the afternoon sun, and a modern airtight flat with windows on only one side that can't cross-ventilate. A Berlin top-floor apartment can hit 32°C indoors in a heat wave even though the city's July average is mild — and heat waves are exactly when the reviews get written.

Skip it — a fan wins. Markets with a July high under 25°C or reliable night cooling: coastal northern Europe, the UK, the mountains, the Nordics. A $45 tower fan, blackout blinds that keep the afternoon sun out, and cross-ventilation do the whole job. Here AC is a cost with no booking behind it, and the fan is the actual ROI winner.

A 3-listing AC audit

Three real listings, to make the bands concrete:

Listing A — one-bed, Seville city centre. Baseline: 4.55 summer rating, thin July bookings, recurring heat complaints. Installed two 9,000 BTU window units for $520. The next summer the rating recovered to 4.87, peak occupancy rose 22 points, and the heat reviews vanished. Payback: the first summer, several times over.

Listing B — two-bed stone house, Provence. Baseline: thick walls kept it around 27°C, borderline. Installed a two-head mini-split for $6,800. The cooling lift was real but modest — the stone was already doing half the work. The bigger surprise was winter: the heat-pump mode cut the gas heating bill by about $300 a year. Payback: ~28 months, carried as much by heating as cooling.

Listing C — coastal cottage, Cornwall. Baseline: July indoor peak around 24°C with a steady sea breeze. Considered AC; installed a $45 fan and blackout blinds instead. Two summers, zero heat reviews. AC would never have come close to paying back — the money went into a better mattress instead.

The pattern is the same as the hot tub, inverted: climate dwarfs every other variable, and the cheap fix in the right climate beats the expensive fix in the wrong one every time.

The failure modes AC adds

Cooling isn't free of downside. Three new ways the listing can lose rating or money, all budgetable:

Condensate drips. Window and portable units produce water, and a bad install or a clogged drain line puts that water on the floor or through the ceiling of the flat below. Get the install angled correctly and clear the drain once a season. A ceiling stain in the downstairs neighbour's flat is a far more expensive problem than the AC ever solved.

Noise. A window unit runs 50–60 dB — dishwasher-loud, right next to the bed. Light sleepers review it. This is the single best argument for a mini-split in the bedroom specifically, where they run a near-silent 25–35 dB. If budget forces window units, put the noisy one in the living room and cool the bedroom with the quieter path.

Servicing. Filters clog within a season of guest use; a blocked filter drops output and overworks the compressor until it fails early. Clean or swap filters monthly in season and budget one $80–$150 service call a year. Miss it and a $250 window unit becomes an annual purchase.

One opinionated take

AC in a hot market is the one amenity you shouldn't run an ROI spreadsheet on, because the honest spreadsheet always says buy it — usually a pair of window units, usually today. The mistake isn't overspending on cooling; it's the opposite. It's treating AC as an optional upgrade, holding off through a mild spring, and discovering in the third July that the €430 you didn't spend on window units was the cheapest review insurance you ever declined. Price the mini-split if you're keeping the place; grab the window units if you're not sure. But in a market where the flat hits 31°C at midnight, the only wrong move is to spend another summer finding out what the heat reviews cost. If you want the peak-season rate lift, the review recovery, and the payback tracked against the install date automatically, that's the kind of thing RentTools is built to surface.