Teikoku Databank’s “Survey of Corporate Trends on Labor Shortages (April 2026)” put the share of ryokan and hotels reporting a shortage of non-regular staff at 38.5%. That is down more than 10 points in each of the past two years, from 63.8% in April 2024 and 51.8% in April 2025, and less than half the record 81.1% of January 2023. With staffing starting to come back, the next question is how many rooms that staff can put back on sale. In this article we measured, property by property across 11,152 properties nationwide (1,156,623 rooms), the share of total rooms actually offered on booking sites — the listed inventory ratio. The headline: the median listed inventory ratio is 51.8%, and it differs by property type — ryokan 62.5%, resort 57.1%, business 47.9% and city 35.8%. If a business hotel restored its listing to the median for its type, it could put roughly 24 more rooms on sale per property; assuming one in four of those added rooms sells, that is about ¥19M a year in additional revenue.
Metric Definitions Used in This Article
- Listed inventory ratio (daily): for a given check-in date, the maximum number of remaining rooms observed on the major booking sites (OTAs, etc.) between 45 days before and the day before the stay, divided by the property’s total rooms. Capped at 100%. Observations more than 45 days before the stay date are not used.
- Listed inventory ratio (ceiling): the maximum of the daily ratios for six dates around the second Wednesday of each month from May to October 2026 (5/13, 6/10, 7/8, 8/5, 9/9, 10/14). The weaker the demand and the more rooms left unsold on a date, the closer the value gets to the upper limit of what the property puts on sale, so this is the article’s primary metric. 10/14 is an interim figure based on observations up to the survey date (September 27).
- Scope: business hotels, city hotels, resort hotels, ryokan and capsule hotels with 10 or more total rooms and at least 5 observations between 45 days before and the day before the stay on all six dates. 11,152 properties / 1,156,623 rooms (based on listings on the major booking sites).
- ADR (average daily rate): an estimated settled rate (tax-excluded equivalent), calculated by applying a property-type correction factor to the lowest-priced plan each property publishes on OTAs (per room, two guests per room, tax included). Checked against property-level results disclosed by listed hotel REITs (91 properties, most recent 3 months), the median error is about 7%. These are estimates and differ from each property’s actual settled prices and accounting figures. Area-level ADR is the median of the properties covered (the level of a typical property in the area).
- Data sources: MetroEngines Research (listed inventory, ADR, room supply); Teikoku Databank (labor shortage share)
- — The share of ryokan and hotels short of non-regular staff was 38.5% (April 2026). Down 42.6 points in three years from 81.1% in January 2023, it is back in the 30% range for the first time in four years and two months, since February 2022.
- — Rooms offered on booking sites are 51.8% of total rooms (median). Ryokan 62.5%, resort 57.1%, business 47.9%, city 35.8%. Properties with 10–29 rooms list 68.4% and those with 200+ rooms 33.2% — the larger the property, the lower the ratio.
- — From May to October the daily ratio moves within a 25.9%–38.5% range. This swing mainly reflects the strength of demand (rooms already sold 45 days out), and cannot be explained by a staffing recovery alone.
- — Of the 33,419-room increase in listed remaining rooms from May to September, 6,212 rooms (about 19%) came from properties opened in 2026. The remaining ~80% arose at existing properties.
- — Restoring to the property-type median is worth about ¥19M a year for one business hotel (about 24 added rooms, 25% sell-through on added rooms, estimated settled ADR about ¥10,650, net of distribution commission and housekeeping labor cost). A sensitivity table with alternative assumptions follows later.
Ryokan and hotel labor shortage share: 81.1% → 38.5% in three years
Teikoku Databank runs its “Survey of Corporate Trends on Labor Shortages” every January, April, July and October. In the April 2026 survey (fielded April 16–30, 10,538 valid responses), the share of ryokan and hotels reporting a shortage of non-regular staff was 38.5%, down 13.3 points from 51.8% a year earlier. In its commentary by industry, the company groups ryokan and hotels with restaurants as showing “improvement for three consecutive years,” citing “productivity gains from the spread of DX and spot work, as well as guest numbers settling down due to the impact of high prices and a decline in Chinese tourists.”
Looking back, in the January 2023 survey the non-regular staff shortage share for ryokan and hotels was 81.1%, the highest of any industry, and the regular-staff share was 77.8%, close to 80%. Since then it has fallen to 63.8% in April 2024, 51.8% in April 2025 and 38.5% in April 2026 — more than 10 points a year on an April-to-April basis. In the same April 2026 survey, the non-regular staff shortage share across all industries was 28.3%, so ryokan and hotels are still 10.2 points above the all-industry average. That can also be read as room for further improvement.
Note that Teikoku Databank’s list of drivers includes “guest numbers settling down.” In other words, the easing of the labor shortage mixes two effects: staff coming back, and demand growth pausing. The listed inventory ratio used in this article is also affected by demand, so without separating the two, one cannot read it as “staff came back, so more rooms went on sale.” With that caveat in place, the sections below quantify the rooms currently on sale and the upside from putting more back.
Source: Compiled by the HotelBank Editorial Team from Teikoku Databank, “Survey of Corporate Trends on Labor Shortages” (January 2023, April 2026). April 2024 and April 2025 are the values shown in the industry table of the April 2026 survey.
Rooms on booking sites are 51.8% of total rooms — a 27-point gap by property type
Of the properties tracked by MetroEngines Research, we calculated the listed inventory ratio (ceiling) for each of the 11,152 properties whose inventory was observed on all six check-in dates from May to October 2026. The overall median is 51.8%; the bottom 25% of properties are at 28.0% or below and the top 25% at 77.8% or above. 14.0% of properties put nearly all of their rooms (90% or more) on booking sites, while 4.7% list less than 10%.
Our earlier article “Median 39.2% of Rooms Online — Allocation Across 12,941 Hotels” (August 2026) put the median at 39.2% from the same room-level inventory observations. The difference comes from the design of the tally. That article took the maximum within 90 days of the stay for two check-in dates (September 19 and October 6, 2026) and excluded observations exceeding total rooms. This article takes the value on the highest of six check-in dates (the ceiling) and caps observations above total rooms at 100%. On a single-date basis, this article’s median is also 25.9%–38.5%, close to that article’s 39.2%. The more check-in dates observed, the higher the ceiling comes out, so the gap between the two figures is best read as a difference in how far each captured the upper limit of rooms put on sale.
By property type, in descending order: ryokan 62.5%, resort 57.1%, business 47.9%, city 35.8% and capsule 30.4%. The gap between ryokan and city hotels reaches 26.7 points. City hotels carry a large share of group blocks tied to banquets and weddings, corporate contracts, and direct website and phone bookings, so the rooms they put on booking sites tend to be only part of the total. Ryokan, by contrast, have few rooms per property and often sell that small inventory mainly through booking sites, which lifts the ratio. Capsule hotels split in two depending on how bed-level inventory is listed: 30.1% list under 10%, while 16.8% list 90% or more.
| Type | Properties | Total rooms | Median | Bottom 25% | Top 25% | Share at 30%+ | Share at 90%+ |
|---|---|---|---|---|---|---|---|
| Ryokan | 2,738 | 94,393 | 62.5% | 43.6% | 82.4% | 88.4% | 17.2% |
| Resort | 1,036 | 99,748 | 57.1% | 33.9% | 84.6% | 79.5% | 19.5% |
| Business | 6,259 | 759,864 | 47.9% | 24.4% | 75.4% | 68.0% | 12.9% |
| City | 1,006 | 187,013 | 35.8% | 18.9% | 59.9% | 57.6% | 6.0% |
| Capsule | 113 | 15,605 | 30.4% | 8.6% | 79.6% | 50.4% | 16.8% |
| All | 11,152 | 1,156,623 | 51.8% | 28.0% | 77.8% | 73.0% | 14.0% |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Listed inventory ratio (ceiling). 11,152 properties / 1,156,623 rooms (based on listings on the major booking sites).
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Share of properties within each type. Business 6,259, city 1,006, resort 1,036, ryokan 2,738 properties.
The more rooms a property has, the smaller the share it puts on booking sites
Much of the gap between property types can be explained by property size. Taking the median listed inventory ratio (ceiling) by room-count band, it falls step by step as size increases: 68.4% for 10–29 rooms, 60.5% for 30–49, 48.9% for 50–99, 42.6% for 100–199 and 33.2% for 200 or more. Properties with 200+ rooms put less than half the share on booking sites that properties with 10–29 rooms do.
This is also the flip side of larger properties having more diverse sales channels. If rooms are allocated to group and travel-agency blocks, corporate contracts, member inventory on the hotel’s own website or long-stay contracts, the allocation to booking sites naturally shrinks. A low ratio therefore does not in itself mean a small sales shelf. What matters as staffing returns is how many rooms were being held off sale because they could not be sold through other channels either and could not be covered by housekeeping or front-of-house staff. That portion cannot be observed from outside, so this article treats the gap to the property-type median as a rough guide to the number of rooms that could potentially be put back.
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Median listed inventory ratio (ceiling). 10–29 rooms 2,837; 30–49 rooms 1,555; 50–99 rooms 2,353; 100–199 rooms 2,883; 200+ rooms 1,524 properties.
Six urban prefectures and Okinawa — city hotels in Tokyo and Osaka sit in the 20% range
By area, we set the six urban prefectures (Tokyo, Osaka, Kyoto, Aichi, Fukuoka and Hokkaido) alongside Okinawa, where resort demand weighs heavily. The overall median is highest in Kyoto and Okinawa at 50.0%, with Osaka (44.0%), Hokkaido (44.5%) and Tokyo (44.8%) on the low side. All seven are below the national 51.8%, but that is because the national population includes many small regional ryokan with high ratios, not because urban levels are especially low.
Comparing like-for-like property types sharpens the differences. City hotels stay in the 20% range in Tokyo (26.5%) and Osaka (26.3%), versus 40.4% in Fukuoka and 34.1% in Kyoto. Among business hotels, Okinawa (55.2%) and Kyoto (51.2%) are high, while Fukuoka (40.4%) and Hokkaido (41.8%) come out lower. Ryokan in Fukuoka are high at 74.3%. The structure of large urban city hotels — a heavy weighting of corporate and group business — shows up directly in the numbers. For extra revenue via booking sites, one option is to route returning weekday corporate allocations or unused weekend group blocks to booking sites.
| Prefecture | Properties | Total rooms | All | Business | City | Resort | Ryokan |
|---|---|---|---|---|---|---|---|
| Tokyo | 981 | 162,734 | 44.8% | 47.1% (815) | 26.5% (98) | — | — |
| Osaka | 559 | 99,344 | 44.0% | 48.5% (442) | 26.3% (90) | — | — |
| Kyoto | 376 | 42,862 | 50.0% | 51.2% (261) | 34.1% (48) | — | 61.5% (51) |
| Aichi | 371 | 49,518 | 46.5% | 44.9% (281) | — | — | 56.2% (53) |
| Fukuoka | 385 | 51,304 | 46.7% | 40.4% (295) | 40.4% (34) | — | 74.3% (38) |
| Hokkaido | 725 | 80,593 | 44.5% | 41.8% (358) | 31.2% (71) | 49.0% (98) | 53.8% (194) |
| Okinawa | 370 | 42,623 | 50.0% | 55.2% (166) | — | 46.7% (184) | — |
| Japan | 11,152 | 1,156,623 | 51.8% | 47.9% | 35.8% | 57.1% | 62.5% |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Median listed inventory ratio (ceiling); figures in parentheses are property counts. Cells with fewer than 30 properties are shown as “—”.
May to October — swings in the daily ratio mainly reflect demand
Listed inventory observation began in March 2026, so this metric cannot be compared with the same period last year. Instead, we tracked the same 11,152 properties from May to October on check-in dates falling on the same weekday (Wednesday). The median daily listed inventory ratio moved as follows: 32.8% on May 13, 37.8% on June 10, 38.5% on July 8, 30.4% on August 5, 33.3% on September 9 and 25.9% on October 14.
This movement is more naturally explained by seasonal demand than by a staffing recovery. Because the daily ratio is the maximum number of remaining rooms seen between 45 days out and the day before, the stronger the demand — i.e. the more rooms already sold 45 days out — the lower the value. It is high in the rainy-season months of June and July, and low in August (summer holidays) and October (the start of the autumn travel season). October 14 is an interim figure with 17 days still to go before the stay date as of the survey, so the ratio may fall further. The same pattern appears by day of week. Comparing Wednesday September 9 with Saturday September 12, the ratio drops sharply at the weekend for business (29.6% → 22.2%), city (22.8% → 16.2%) and resort (34.5% → 28.8%), while ryokan barely move (34.7% → 33.3%). Weekend demand for ryokan is stronger than on weekdays, but because they list a small number of rooms mainly via booking sites in the first place, the gap is likely harder to see.
In short, tracking the daily ratio alone cannot distinguish between staff coming back and the sales shelf widening, and demand simply being weak with many rooms left unsold. That is why this article uses the maximum of the six dates, the “ceiling,” as its main metric: the ceiling picks up a value close to the upper limit of rooms a property puts on sale. Note, however, that for properties with strong demand on all six dates, the ceiling also comes out low.
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Median daily listed inventory ratio (same 11,152 properties). 10/14 is an interim figure based on observations up to the survey date (September 27).
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Median daily listed inventory ratio. Properties observed on both dates (business 6,730, city 1,076, resort 1,329, ryokan 4,435).
The increase in rooms on sale comes more from existing properties than from new openings
As staffing returns, it matters to separate whether rooms increased because new hotels opened or because existing properties put more rooms on sale. First, comparing room supply within a 3 km radius of the central station in seven major cities (total rooms at properties open and not yet closed) at the end of 2025 and at the survey date: Naha (Kenchomae) grew by +853 rooms (+3.2%), Fukuoka (Hakata Station) by +1,180 (+2.7%) and Sapporo (Sapporo Station) by +440 (+1.2%), while Tokyo (Tokyo Station) rose only +372 (+0.5%) and Kyoto (Kyoto Station) +216 (+0.3%). Osaka (Osaka Station) is −2,732 rooms (−4.7%) after closures. Total room counts in city centers moved by only a few percent over these nine months.
| Area (3 km radius) | End-2024 | End-2025 | 2026 (survey date) | vs. end-2025 |
|---|---|---|---|---|
| Tokyo (Tokyo Station) | 77,044 | 77,961 | 78,333 | +372 (+0.5%) |
| Osaka (Osaka Station) | 55,958 | 58,628 | 55,896 | −2,732 (−4.7%) |
| Kyoto (Kyoto Station) | 63,239 | 64,087 | 64,303 | +216 (+0.3%) |
| Nagoya (Nagoya Station) | 35,310 | 35,191 | 35,422 | +231 (+0.7%) |
| Fukuoka (Hakata Station) | 42,881 | 43,259 | 44,439 | +1,180 (+2.7%) |
| Sapporo (Sapporo Station) | 36,053 | 37,350 | 37,790 | +440 (+1.2%) |
| Naha (Kenchomae) | 25,392 | 26,606 | 27,459 | +853 (+3.2%) |
Source: MetroEngines Research (based on OTA listing confirmation). Total rooms at properties open and not yet closed at each year-end. * Openings are identified from listings, so recent openings may increase as listings appear.
Next, we compared listed remaining rooms nationwide (the numerator of the daily listed inventory ratio) on May 13 and September 9. The total rose from 412,973 to 446,392 rooms, up 33,419 (+8.1%). Broken down: +17,935 rooms at the 12,285 existing properties observed on both dates; +6,212 from properties opened in 2026 (91 properties not observed in May but observed in September); +11,375 from existing properties not observed in May but observed in September (1,344 properties); and −2,103 from properties not observed in September (143 properties). New openings account for about 19% of the total increase; the remaining ~80% arose at existing properties.
However, the increase at existing properties mixes several factors. September 9 may have had weaker demand than May 13, leaving more rooms unsold 45 days out. The “existing properties observed in September but not in May” include sales resuming after renovations or temporary closures, changes in how room types are listed, and new listings on booking sites, and the cause cannot be identified property by property. By prefecture, Tokyo (+12,362 rooms at existing properties), Kyoto (+6,386) and Osaka (+4,999) increased, while Hokkaido, where September is peak season, fell by −8,687 — the seasonal difference in demand shows through directly.
| Component (national, 5/13 → 9/9) | Properties | Change in listed remaining rooms | Share |
|---|---|---|---|
| Existing properties (observed on both dates) | 12,285 | +17,935 rooms | 53.7% |
| Existing properties (observed in September only) | 1,344 | +11,375 rooms | 34.0% |
| Properties opened in 2026 (observed in September only) | 91 | +6,212 rooms | 18.6% |
| Observed in May only | 143 | −2,103 rooms | −6.3% |
| Total | — | +33,419 rooms | 100% |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team. Sum of the maximum remaining rooms observed between 45 days before and the day before the stay. Properties with 10+ total rooms and 5+ observations. 12,428 properties on 5/13; 13,720 properties (1,275,753 rooms) on 9/9.
Factors to check alongside this ratio
The listed inventory ratio can correlate with a staffing recovery, but it is not determined by staffing alone. When the ratio rises or falls, check at least the following factors as well.
| Factor | Effect on the ratio |
|---|---|
| Strength of demand | Rooms already sold 45 days out do not appear as remaining rooms. The ratio falls in peak seasons and at weekends (see the May–October trend and the Wednesday vs. Saturday comparison in this article). |
| Channel allocation | The more rooms go to group and travel-agency blocks, corporate contracts, or direct website and phone bookings, the smaller the allocation to booking sites. This is a large part of why the ratio is low at large properties and city hotels. |
| How inventory is released | If a property does not release its full inventory to booking sites at once but adds it gradually as the stay date approaches, the maximum seen between 45 days out and the day before comes out lower than total rooms. |
| How room types are listed | Merging or splitting room types, or listing only certain types on booking sites, changes the number of remaining rooms observed. |
| Renovation / temporary closure | During floor-by-floor renovations or equipment upgrades, the rooms available for sale temporarily fall. The ratio rises once the work ends. |
| Staffing (housekeeping, front of house) | On days when housekeeping cannot keep up, rooms are closed out even when they could sell. As staffing returns, these closed-out rooms go back on sale. |
Prepared by the HotelBank Editorial Team
Isolating the staffing effect alone requires lining up the same property, the same weekday and days with similar demand, and supplementing channel allocations and renovation schedules with information from the property itself. What external data alone can show is the scale of rooms on sale and where a property sits within its type; this article shows the upside within that scope.
Upside from restoring listings to the property-type median
For properties whose listed inventory ratio (ceiling) is below their type’s median, converting the gap to the median into rooms gives a median of 23.5 rooms per property for 3,128 business hotels (these properties have a median of 118 total rooms and a median ratio of 24.3%). It is 25.2 rooms for 503 city hotels (179 rooms, 18.9%), 18.2 rooms for 516 resorts (85.5 rooms, 33.8%) and 4.5 rooms for 1,359 ryokan (23 rooms, 43.5%).
These rooms do not translate directly into revenue. If a room is already being sold through another channel, moving it to booking sites adds nothing. The upside only arises when rooms that were not selling through any channel are put back. We therefore set the share of returned rooms that actually sell (the sell-through rate on added rooms) at three levels — 10%, 25% and 40% — and estimated the annual upside net of distribution commission and housekeeping labor cost.
■ Estimate assumptions
- Additional rooms listed on booking sites: 5, 10 and 20 rooms (the median gap to the business-hotel type median is 23.5 rooms)
- Sell-through rate on added rooms: 10%, 25% and 40% (the share of the 365 days a year on which a returned room sells; assumes no displacement of existing sales)
- Estimated settled ADR: business hotels ¥10,650 (simple average of the 12-month averages from September 2025 to August 2026 for seven prefectures — Tokyo, Osaka, Kyoto, Aichi, Fukuoka, Hokkaido and Okinawa; 193–924 properties per prefecture per month)
- Distribution commission: 10% of revenue (assumption)
- Housekeeping labor cost: ¥640 per room (the national level under the 25-minute scenario in our estimate of hotel housekeeping labor cost per room). Excludes linen, amenities and utilities
- Upside per room-night = ¥10,650 × (1 − 10%) − ¥640 = ¥8,945
| Additional rooms listed \ Sell-through rate | 10% | 25% | 40% |
|---|---|---|---|
| 5 rooms | ¥1.63M | ¥4.08M | ¥6.53M |
| 10 rooms | ¥3.26M | ¥8.16M | ¥13.06M |
| 20 rooms | ¥6.53M | ¥16.32M | ¥26.12M |
Source: MetroEngines Research; prepared by the HotelBank Editorial Team. Annual upside per business hotel (net of distribution commission and housekeeping labor cost). Estimates.
By property type, restoring the median gap to the type median and assuming a 25% sell-through rate on added rooms gives the following upside per property: about ¥19M a year for business hotels, about ¥37M for city hotels, about ¥23.5M for resorts and about ¥4.7M for ryokan. Ryokan have few rooms per property and high ratios to begin with, so the room-count headroom is small. Note too that ryokan need more cleaning and meal-service work per room than hotels, so the ¥640 housekeeping labor cost is close to a lower-bound assumption for them.
| Type | Properties below median | Gap to median | Estimated settled ADR | Added room-nights sold / yr | Upside / yr |
|---|---|---|---|---|---|
| Business | 3,128 | 23.5 rooms | ¥10,650 | 2,144 room-nights | ¥19.18M |
| City | 503 | 25.2 rooms | ¥18,420 | 2,300 room-nights | ¥36.65M |
| Resort | 516 | 18.2 rooms | ¥16,435 | 1,661 room-nights | ¥23.50M |
| Ryokan | 1,359 | 4.5 rooms | ¥13,301 | 411 room-nights | ¥4.65M |
Source: MetroEngines Research; prepared by the HotelBank Editorial Team. 25% sell-through on added rooms, net of 10% distribution commission and ¥640 housekeeping labor cost per room. Estimated settled ADR is the simple average of 12-month averages by type for seven prefectures (note that for resorts, the five prefectures other than Hokkaido and Okinawa have only 5–21 properties per month). Estimates.
This upside requires funding the labor. The FY2026 minimum wage revision pushes up labor cost per room (by about ¥9,600 per room per year on our estimate). Even so, in the estimate above about ¥8,900 per room-night remains after deducting housekeeping labor cost. As long as the returned rooms sell, the additional housekeeping labor cost is less than 10% of the upside.
The order to put rooms back — day of week, room type, price band
As staffing returns in stages, putting closed-out rooms back in order of how likely they are to sell places less strain on housekeeping than returning them all at once. The data here suggest the following order.
Day of week: start with weekends, when demand is strong but the ratio tends to fall. In the September Wednesday vs. Saturday comparison, the daily ratio fell 7.4 points at the weekend for business hotels and 6.6 points for city hotels. Weekends are already heavily sold 45 days out, so remaining rooms added to the shelf are more likely to sell. If the sell-through rate on added rooms rises from 25% to 40%, the annual upside on the same 10 rooms grows from ¥8.16M to ¥13.06M. On weekdays, prioritize whichever of Tuesday to Thursday — when corporate demand returns — has historically seen the most closeouts.
Room type: start with types that are quick to clean and hold their rate. Standard single and double rooms for one or two guests take less time to clean, so returning them adds little housekeeping load. Cutting cleaning time by just five minutes changes labor cost per room by about ¥150. Types that take longer to clean, such as Japanese-Western rooms and suites, are logically returned once staffing has stabilized.
Price band: start near the median price. Releasing returned rooms at the lowest price band tends simply to shift existing bookings to a cheaper price. Listing near the type’s estimated settled ADR and adjusting as they sell is more likely to leave a genuine increment.
Rooms that are easier to put back under housekeeping constraints
Even while housekeeping staff are limited, the rooms that are easier to put back have things in common. First, rooms for consecutive-night stays. Combined with plans that simplify cleaning during the stay (no cleaning, towel change only, etc.), rooms in a multi-night stay do not need a full clean every night, so the same staff can handle more rooms. Second, rooms for the night after an early check-out day. If guests leave early the previous day, cleaning can be spread across more of the day and handled without adding staff at peak hours. Third, returning rooms on the same cleaning route together. Returning rooms in units on the same floor and the same route keeps work time per room, including travel time, down.
By size, the listed inventory ratio (ceiling) for properties with 200+ rooms was 33.2%, less than half the 68.4% of properties with 10–29 rooms. Larger properties have more channels besides booking sites, but they also find it easier to return rooms floor by floor, and the effect of reorganizing housekeeping staffing shows up as a larger number of rooms. A period in which the labor shortage keeps easing is an opportunity to review channel allocation and housekeeping together.
Frequently Asked Questions
Q. Does a low listed inventory ratio mean a property is closing out rooms?
A. Not necessarily. The ratio also comes out low when a property allocates rooms to channels other than booking sites, such as groups, corporate clients or its own website, or when it adds inventory to booking sites gradually as the stay date approaches. The ratio only indicates the scale of rooms offered on booking sites; it does not indicate the property’s overall sales or occupancy.
Q. Is the fall in the labor shortage share related to changes in the listed inventory ratio?
A. They may be correlated, but not one-to-one. The swings in the daily ratio from May to October 2026 are mainly explained by seasonal demand, and the effect of the staffing recovery alone cannot be isolated. Channel allocation, how room types are listed and renovations also affect the ratio.
Q. Is the listed inventory ratio higher than last year?
A. Listed inventory observation began in March 2026, so a comparison with the same period last year is not possible. This article tracks the same properties on check-in dates falling on the same weekday from May to October 2026.
Summary
The share of ryokan and hotels short of non-regular staff fell from 81.1% in January 2023 to 38.5% in April 2026. Across 11,152 properties nationwide, the median listed inventory ratio (ceiling) is 51.8%, varying widely by type from 62.5% for ryokan to 35.8% for city hotels, and lower the more rooms a property has. Of the 33,419-room increase in listed remaining rooms from May to September, new openings accounted for about 20%; the rest arose at existing properties. This change, however, mixes seasonal demand, channel allocation, how inventory is released and renovations, and cannot be explained by a staffing recovery alone.
With that said, if a business hotel restores its listing to the type median and one in four of the added rooms sells, there is about ¥19M a year of upside per property. By putting rooms back in step with returning staff — starting with high-demand days of the week, room types that are quick to clean and the type’s standard price band — a hotel can widen its sales shelf without overstretching housekeeping.
⚠ Note on future-date data: The listed inventory ratio for October 14 check-ins is an interim figure based on observations up to the survey date (September 27, 2026) and may change with sales progress and inventory additions before the stay date. Observations more than 45 days before the stay date may include states before listing allocations are loaded and may not reflect booking progress, so they are not used in this article.
Related Articles
References and Sources
■ Data sources
Room-level listed inventory observed by MetroEngines Research on the major booking sites (check-in dates May 13, June 10, July 8, August 5, September 9, September 12 and October 14, 2026; observed between 45 days before and the day before the stay); estimated settled ADR (four property types in Tokyo, Osaka, Kyoto, Aichi, Fukuoka, Hokkaido and Okinawa, September 2025 to August 2026); room supply within 3 km of the central station in seven major cities (at each year-end). Labor shortage shares are from Teikoku Databank’s “Survey of Corporate Trends on Labor Shortages.”
■ Estimate assumptions
Upside = additional rooms listed × 365 days × sell-through rate on added rooms × (estimated settled ADR × (1 − 10% distribution commission) − ¥640 housekeeping labor cost). Sell-through rate on added rooms at three levels: 10%, 25% and 40%. Estimated settled ADR is the simple average of 12-month averages by type for seven prefectures (business ¥10,650, city ¥18,420, resort ¥16,435, ryokan ¥13,301). Excludes linen, amenities and utilities, and any displacement of existing sales.
■ Limitations and caveats
The listed inventory ratio indicates the scale of rooms offered on booking sites; it does not directly indicate a property’s overall sales or the volume allocated to other channels. The ceiling depends on the number of check-in dates observed and on demand, so its level differs from the inventory allocation ratio in our earlier article (median 39.2%), which used different observation dates. Observation began in March 2026, so no comparison with the same period last year is possible, and the October 14 figure is interim. Estimated settled ADR is an estimate; for resorts, the five prefectures other than Hokkaido and Okinawa have only 5–21 properties per month. The relationship with the labor shortage share indicates only a possible correlation; causation has not been established.
- Teikoku Databank, “Survey of Corporate Trends on Labor Shortages (April 2026)” (May 19, 2026, in Japanese)
- Teikoku Databank, “Survey of Corporate Trends on Labor Shortages (July 2026)” (August 17, 2026, in Japanese)
- Teikoku Databank, “Labor-short companies above 50% for five straight months; ryokan and hotels at about 80%, a record high” (February 17, 2023, in Japanese)
- MetroEngines Research — listed inventory on the major booking sites (six check-in dates from May to October 2026, 11,152 properties), estimated settled ADR (seven prefectures, by property type), room supply in central areas of seven major cities