What two Airbnbs near Tokyo actually earned in their first six months

Same host, same listing style, same six months. The Tokyo house brought in more money. The beach house kept more of it.

By the host · Published 2026-10-09 · Figures from my own books, March–August 2026

Both houses are rented, not owned, and registered under Japan's private lodging law (the minpaku law), which caps each house at 180 nights a year. Every figure below is what actually landed in my bank account after platform fees, set against the bills I actually paid. Nothing is projected except where I say so.

Tokyo, residential ward, 2 min from a subway station. Rent is ¥187,000 a month. Setup — everything spent before the first guest — cost ¥1,444,321.

Rent is 92% of everything this house costs to run. Utilities barely matter, with one exception: electricity went from ¥3,660 in March to ¥18,408 in August, because guests run the air conditioning all day in a Tokyo summer. Two of the six months lost money. In May and August, a handful of empty weeks were enough to put the month in the red, because the rent does not care how many guests came.

Shonan coast, Kanagawa, about an hour from Tokyo. Rent is ¥89,330 a month. Setup cost ¥1,235,919.

The beach house made ¥344,513 more than the city house on ¥263,055 less revenue. The reason is the rent line: at under ¥100,000 a month in total costs, a single good weekend covers the month.

Two honest caveats. First, ¥111,043 of the July–August payouts came from two stays that guests cancelled after my cancellation deadline — Airbnb still paid me. Without that, the six-month profit would be about ¥439,189. Second, the July and August water bills had not arrived when I wrote this (it runs about ¥1,600 a month).

1. Pick the house by rent, not by location

The Tokyo house had the better occupancy in its best month (71%) and the better address. It still lost to a beach town. At the ¥18,755 it earned per booked night, the city house needs about 11 booked nights a month just to cover rent and bills.

2. The Airbnb fee is 15.5%, not 3%

On the host-only fee, which both my listings are on, Airbnb keeps about 15.5% of the nightly rate and cleaning fee. On a real 7-night booking I had: ¥146,310 booked, ¥22,768 fee, ¥123,632 paid out. Model your numbers with 15.5%.

3. Beach demand is last-minute

At the end of July the beach house had about ¥129,000 booked for August. It finished August at ¥305,998. Seven bookings came in during those last weeks. An empty-looking calendar three weeks out did not mean an empty month.

4. More channels did not mean more bookings

The city house is also on Booking.com. In the 90 days to late August it was shown in search 57,295 times and its page was opened 12,332 times. It got zero bookings. As of late August, Rakuten's vacation rental site had brought in two bookings for the beach house and none for the city house. Airbnb did almost all of the work.

5. The 180-night cap changes what "occupancy" means

Under the minpaku law the year starts on April 1 and each house can sell at most 180 nights. A 50% annual occupancy is not a bad result here; it is the legal ceiling. What matters is how much each of those 180 nights earns compared with twelve months of rent.

Rent is the only number that decides whether a rented minpaku works. Everything else is rounding.

• Payout is money received from Airbnb, Rakuten and Booking.com after their fees, split across the calendar months the guest actually stayed. Test bookings by friends and family are excluded.
• Costs are rent, electricity, gas, water and home Wi-Fi only. Cleaning, linen and consumables are not included, so budget for them separately.
• Setup cost is everything spent before the first guest. Payback months = setup cost ÷ average monthly profit so far.
• Dollar figures use ¥158 = $1 as a rough guide.

Next: the software I use (and skip) to run both houses.