Hotel Software in Japan 2026: 104 Companies Earn Almost Half the Revenue, 60,000 Small Properties Have the Empty Rooms

Japan looks like the obvious place to sell hotel software. Foreign visitors spend more on accommodation than on anything else — 36.6% of their travel spending in 2025, JPY 34,578 hundred million (Japan Tourism Agency, 2025) — the country is full of independent inns, and every founder has heard that Japanese hotels can't find staff. The trap is treating "Japanese hotels" as one customer. Government statistics split the market into two buyers who want different products, pay in different ways and are reached through different doors. Pick one before you write the first line of code.

What the data shows

The money is concentrated. Accommodation companies in Japan (industry 75 in the Japanese classification) had sales of JPY 7,709,750 million in 2024, and ryokan and hotels (751) account for JPY 7,348,884 million of it (e-Stat Economic Structure Survey, 2024). The same survey splits companies by annual sales:

Annual sales of the company	Companies	Sales, JPY million
JPY 10 billion and over	104	3,604,803
JPY 1–10 billion	838	2,265,466
JPY 0.3–1 billion	1,846	953,688
JPY 100–300 million	3,306	575,417
JPY 30–100 million	4,315	251,840
JPY 10–30 million	2,659	51,153
JPY 3–10 million	1,044	6,776
Under JPY 3 million	482	605

By our calculation from that table, the 104 largest companies earn 46.8% of the industry's sales, and the 942 companies above JPY 1 billion earn 76.1%. At the other end, 8,500 companies with sales under JPY 100 million — 58.2% of the 14,594 companies in the table — earn 4.0% of the sales.

The properties are small. Of 74,125 accommodation establishments counted at the end of 2025, 60,128 employ 0–9 people — 81.1% by our calculation (Japan Tourism Agency, Overnight Travel Statistics Survey, 2025). By type, there were 20,250 ryokan, 23,110 simple lodgings such as guesthouses and hostels, 8,970 business hotels, 3,170 resort hotels and 1,400 city hotels at the end of 2024.

The small ones are half empty. In the latest annual figures, establishments with 0–9 people employed filled 35.6% of their rooms. In 2025, establishments with 10–29 people filled 70.4%, those with 30–99 filled 71.5%, and those with 100 or more filled 70.8% (Japan Tourism Agency, Overnight Travel Statistics Survey). Bed occupancy shows the same gap: 19.9% for the smallest class against 52.6% for the largest. The split by type is just as sharp: in August 2026, business hotels sold 75.7% of their rooms and city hotels 72.3%, while ryokan sold 47.4%, against 64.4% across all types.

Demand is cooling, not collapsing. Japan received 38,962,366 tourist arrivals in 2025, up 15.9% on the year (JNTO, 2025). In 2026 the curve has turned: arrivals in August were 3,098,900, down 9.6% year on year, and visitors from China in the second quarter fell 52.5% (JNTO, 2026; Japan Tourism Agency, 2026). Regional Japan still matters: in July 2026, 58.4% of all guest nights were outside the three big metropolitan areas.

The small operator's problem is people. In a Japan Finance Corporation survey of small hotels and ryokan, 59.2% reported a shortage of employees in 2025. Of those that were short, 24.8% shortened opening hours, 45.0% used short-term spot workers, and only 19.3% responded with mechanisation or information technology. Payroll took JPY 1,537,941 million in 2024, or 19.9% of the industry's sales by our calculation (e-Stat, 2024). The pressure is sharpest on the floor: in July 2026 there were 2.30 job openings per applicant for customer service and waiting staff (MHLW, 2026).

The barrier to new tools is know-how, not price. In a March 2026 survey of the same segment, 33.9% of small hotels and ryokan already used generative AI in some operations and 19.9% were considering it; 26.3% had no plans. Asked about barriers, 62.7% named lack of know-how, 48.0% doubted the accuracy of the output and 47.1% lacked staff able to use it. Only 16.7% named the cost of setting it up and running it (Japan Finance Corporation, 2026).

What this means for your product

Choose your buyer: the 942 or the 60,000. The companies above JPY 1 billion are where three quarters of the money sits. Selling to them looks like enterprise sales: long cycles, integration with existing systems, procurement. The small inns are the opposite: many logos, little money each, and owners who say they lack the know-how to adopt new tools. The 4,315 companies in the JPY 30–100 million class average about JPY 58 million in sales by our calculation, so a hypothetical JPY 10,000 monthly subscription would be about 0.2% of a typical company's revenue in that band. Price is not the objection there; adoption effort is.

Sell filled rooms, not software. A small inn that sells 35.6% of its rooms while larger hotels sell around 70% has an obvious, measurable gap. A product that moves that number — distribution to online channels, direct booking, guest messaging in other languages — has a sales pitch that needs no translation. Ryokan at 47.4% against business hotels at 75.7% in August 2026 tells the same story by type.

Replace a shift, don't add a screen. With 59.2% short of staff and only 19.3% of those turning to mechanisation or information technology, the opening is in tasks that take hours: check-in, guest questions, cleaning schedules, rostering spot workers. The test for every feature: does it remove work for the owner and the one or two people they employ, or does it add a dashboard someone has to learn?

Ship the setup, not the tool. When 62.7% say their barrier is know-how and 16.7% say it is cost, a cheaper plan won't win. Onboarding done for the customer, a person who sets it up on site or by phone, local-language support and templates that work on day one are the product. Budget for that in your CAC: in this segment, channel partners such as regional tourism bodies, booking-system resellers and lenders to small inns may be cheaper than direct sales.

Plan for a softer 2026. A product sold on "ride the inbound boom" is exposed when arrivals fall 9.6% in a month and Chinese visitors drop by half. A product sold on saving staff hours keeps its case when demand dips.

What the data does not tell you

• What these properties pay for software today. No official source publishes software spend by Japanese hotels. Our room-rate data for small ryokan comes from an industry association survey and is not strong enough to quote.
• Companies are not properties. The sales table counts companies; occupancy is measured by establishment. One large company can run hundreds of hotels, so the 104 companies and the 1,088 establishments with 100+ people are different units.
• The surveys of small inns are lender surveys. The Japan Finance Corporation surveys its own borrowers among small hotels and ryokan, not all inns.
• Wage and job-opening data cover hotels and restaurants together, or occupations across all industries. They describe the labour market the inn competes in, not the inn alone.
• Year labels. The occupancy figures for the smallest establishments (35.6% of rooms, 19.9% of beds) come from an annual workbook whose edition may be a year older than the 2025 figures for larger establishments. The gap is large enough that the conclusion holds, but check the edition date before you put it in a deck.

The Japanese market is real, but it is two markets. Decide whether you are building for a few hundred operators who already own systems or for tens of thousands of inns that need someone to do the setup for them — and build the sales motion for that choice before the product.