There is one question that always comes up when a new limited-service hotel project reaches the point of fixing the standard guest-room specification. Should the bathroom be a three-in-one unit bath (tub, washbasin and toilet integrated into a single module), or should the bath and toilet be separated? The former keeps the per-room floor area down and yields more rooms out of the same gross floor area. The latter may allow a higher rate, but it requires paying for several extra square metres per room plus the equipment cost differential up front.

This article narrows in on that single decision, connecting the demand side (how often it actually shows up in guest reviews, and how much it moves the rate) and the supply side (how many square metres and how much money it adds) into one payback calculation. Cross-referencing 4,515 limited-service hotels, 2.3 million guest reviews and listed prices by room type, what moves the rate turns out not to be the bathroom configuration itself, but the floor area that configuration demands.

First, the measurable volume of demand. Running a semantic analysis over guest reviews and extracting remarks touching on “separated bath and toilet / presence of an independent washbasin,” 50,367 reviews qualified across all formats over the most recent 24 months (September 2024 to September 2026). Of these, 45,816 (91.0%) were positive mentions, 1,334 (2.6%) were complaints or requests, and the remaining 3,217 were neutral.

On the positive side, the phrasing is dominated by appraisal of the specification’s mere existence — “the bath and toilet were separate,” “separate bath and toilet, nice and roomy.” On the complaint side, what stands out is the request form: “it would be nice if the bath and toilet were separate,” “please make it separated.” In other words, this tag carries two meanings at once: on the positive side it signals properties that have the specification, and on the negative side it marks unmet demand at properties that do not.

Narrowing to limited-service hotels (business-hotel format, 4,515 properties with 30 or more reviews in the last 24 months), positive mentions numbered 13,935 against 2,307,516 target reviews — a mention rate of 0.60%. As a number it is small. But because reviews only surface “things memorable enough to be worth writing down,” this 0.60% should be read as an indicator of intensity of interest, not of how widespread the specification is.

Month by month, the mention rate was 0.544% in the first 12 months (September 2024 – August 2025) against 0.629% in the second 12 months (September 2025 – August 2026) — an increase of 15.7%. Because the tagging rate for guest-room tags holds steady at 61–64% each month, this gap reflects not a shift in analytical coverage but a genuine rise in how often guests mention bathroom separation. A rising inbound share and growth in long-stay and solo female travel are plausible drivers, but review data alone cannot isolate the cause.

The geographic skew is unmistakable. Mention rates are highest in regional urban markets — Chiba 1.57%, Nagasaki 1.40%, Shiga 1.25%, Hiroshima 1.23% — while major metropolitan centres sit low: Tokyo 0.32%, Kyoto 0.45%, Osaka 0.55%. The natural reading is that this reflects differences in how rooms are built, not differences in demand. In central locations where land is expensive, compact rooms with three-in-one unit baths are the standard, so “being separate” is unlikely to register as an unusual experience. Conversely, in regional cities with room to build, and at properties with Japanese-style or family rooms, separated layouts appear more often — and end up in the reviews.

Lining up the six major metropolitan markets shows that the volume of expressed demand and the way supply responds do not necessarily move together.

Expressed demand (mention rate) and the supply response (share of properties selling such rooms) do not rank in the same order. Osaka sits below average on mentions at 0.55%, yet its share of properties selling such rooms is 9.1% (24 of 263) — the highest among the major markets. Kyoto likewise pairs a 0.45% mention rate with an 8.3% selling share. Tokyo, by contrast, is lowest on both counts among the major markets: 0.32% mentions and a 5.4% selling share. A coherent reading is that in central-Tokyo limited-service hotels the opportunity cost of guest-room floor area is at its highest, so area-hungry separated layouts are the least likely to be chosen.

Next, the state of supply. Counting the properties that state at least one of “separate bath and toilet,” “separated,” “with a washing area” or “independent washbasin” in the names of the room types they sell on booking sites, 222 of the 4,515 limited-service hotels (4.9%) do so. The room name is the only specification display a guest sees before booking, so whether it appears there also reflects the operator’s own judgement as to whether the specification can be presented as a justification for the rate.

In practice the wording tends to pair the configuration with the area — “Twin / Non-smoking / 26m², separate bath and toilet,” “Deluxe Twin [Separated Bath and Toilet Type],” “Separate-Bath Single,” “Japanese-style room, 12 tatami (49m², separate bath and toilet).” In other words, operators present the bathroom configuration not as a standalone selling point but bundled with space. That is consistent with the price analysis that follows.

Cutting by opening year brings out a clear generational gap. Among properties opened in 1989 or earlier, only 3.6% (34 of 932) sell such rooms; among those opened 2016–2019 the figure is 10.2% (58 of 567) — 2.8 times higher. The share of properties receiving positive mentions also doubles, from 22.6% to 45.3%. Properties opened from 2020 onward number only 38 and cannot bear interpretation of the ratio, so they are shown as a reference value only.

This generational gap points in the same direction as the expansion of guest-room area itself. The 2016–2019 cohort has a median estimated ADR of ¥11,200, far above the ¥7,100 of the pre-1989 cohort. Bathroom separation has not been a standalone fashion; it has been implemented as part of a design philosophy of widening rooms to capture higher rates.

Here is the core of this article. When measuring a rate effect, comparing different properties mixes in differences of location, building age and brand. So the comparison was made within the same property, between room types with the same meal conditions and the same maximum occupancy. The sample is the 141 properties selling such rooms for which a comparable pairing of room types could be established. Median listed prices (double occupancy, tax-included, excluding sold-out dates) for check-in dates between June and August 2026 were computed per room type, and the difference between separated-bathroom rooms and other rooms was taken within each property.

The last two rows of the table are the important ones. Separated-bathroom rooms are listed at a median of roughly ¥1,500 more than other rooms. But the area gap is +4.5m², and once the comparison is restricted to pairs of near-identical area (within ±2m²), the price gap has a median of -¥960 — it disappears. That raises the suspicion that the price gap reflects not the bathroom configuration but simply the fact that those rooms are larger.

To separate the two, a regression of “listed price = floor area + bathroom-configuration dummy” was run on the 2,394 room types carrying an area notation in the room name, holding property, meal conditions and maximum occupancy fixed (a fixed-effects specification centred within property and condition; 2,303 valid observations across 98 properties). The results are as follows.

In the scatter plot, the separated-bathroom room types (blue) cluster toward the larger-area side, but within any given area band their price range overlaps that of other rooms (grey). The bathroom configuration does not stand on its own as an explanatory variable for price. A specification that guests clearly endorse shows up in price not as a configuration but as floor area.

This rewrites the premise of the investment decision. Not “separating the bath and toilet raises the rate,” but “the floor area added in order to separate them raises the rate.” Put the other way: where separation is achieved purely through clever layout without adding area, a price premium should not be expected. The payback calculation should be built on the marginal rate per square metre.

Converting to an estimated settled ADR basis: the median ratio of estimated settled ADR to listed price (all-plan average) measured across 4,462 properties is 0.540, so the marginal rate per square metre is approximately ¥570 per m² (estimated settled ADR basis, tax-excluded equivalent). For comparison, the average rate per square metre is the limited-service median estimated ADR of ¥7,800 divided by the median room area of 22.0m², or ¥354/m². That the marginal rate (about ¥570) exceeds the average rate (about ¥350) is why the payback calculation below works at all.

So how many square metres does separating the bathroom actually add to a room? Build it up from the dimensions of the three-in-one unit. The size designations standard in hotel guest rooms are “1216” (1.2m × 1.6m = 1.92m²) and “1418” (1.4m × 1.8m = 2.52m²), which assume tub, washbasin and toilet packed into a single box. Separating them requires, on top of the bathroom itself (1.92m² for a 1216 equivalent), an independent toilet (1.0–1.3m²), an independent washbasin (0.8–1.2m²), and the door swings and circulation each of those needs — which in practice lands in a range of +2 to 5m².

The measured data agree with that range. Comparing room types with an area notation in the room name, separated-bathroom rooms have a median area of 27.0m² (365 room types) against 22.0m² for other rooms at the same properties (1,897 room types). The median of the property-matched differences is +5.0m² (84 properties). That figure, however, captures more than the geometric difference of the bathroom: it also reflects the fact that separated layouts tend to be implemented in room types — twins, deluxes — that were already larger. The scenarios therefore span from +2.0m², close to the geometric floor, to +5.0m², the measured median.

Unit cost comes from the Ministry of Land, Infrastructure, Transport and Tourism’s “Building Dynamics Statistics Survey,” lodging-use buildings. The 2024 national figures are 1,756,286m² of floor area against ¥1.0384 trillion in planned construction cost — ¥591,279 per m² (¥1.955 million per tsubo). That is up 41% in two years from ¥419,164 in 2022, so additional floor area now costs more than it used to. To this is added an equipment differential from a three-in-one unit to a separated specification of ¥400,000 per room (¥200,000 to ¥1,000,000 tested in the sensitivity analysis below).

Payback is set on a GOP (gross operating profit) basis. The ADR uplift generates almost no additional variable cost, so the contribution ratio is high, but it is set conservatively at 70% to allow for booking-channel commissions and increased housekeeping supplies. The annual revenue gain is “ADR uplift × 365 × occupancy.” Occupancy is not a measured value but a full-year business-plan assumption, set at three levels: 70%, 80% and 85%.

Payback shortens as the additional area grows because the fixed ¥400,000 equipment differential is spread thinner across more square metres. Since the capex for area and the marginal rate per square metre move in proportion, adding more area does not sharply improve the return on capex: the gap between Options B and C is only 0.8 percentage points.

What matters most is the marginal rate per square metre itself. This article’s measured value is ¥1,049/m² on a listed-price basis (¥566 converted to estimated settled ADR), but it is estimated from within-property variation across the 98 properties selling such rooms, and will run higher or lower depending on an individual project’s location and guest mix. Consider the sensitivity.

Moving the regression coefficient’s uncertainty and occupancy together makes the range of judgement explicit. At the measured ¥566/m², payback stays within 7.0 years even at 70% full-year occupancy; at the lower end of the confidence interval (¥304/m²) it stretches to 10.1 years even at 90% occupancy and 11.4 years at 80%. The estimation range for the marginal rate moves payback far more than a 5-point change in occupancy does. Verifying the relationship between price and area by room type around the planned site is therefore a higher priority than refining the occupancy assumption.

The break point is clear. Once the marginal rate per square metre falls below ¥400/m² on an estimated settled ADR basis, payback stretches past 8.6 years — beyond the tolerance of a typical limited-service investment horizon. Conversely, in a location that can command ¥700/m², payback is 4.9 years. That means the decision can flip between high-rate central and resort locations and lower-rate regional cities.

In a new-build project, the binding constraint is usually not capex but the ceiling on gross floor area. Where the floor-area ratio fixes the total, adding 3.5m² per room is the same thing as cutting rooms. Taking the 35.0m² room unit area for limited-service hotels shown in our earlier work as the starting point, the table below models the outcome per 1,000m² of gross floor area.

Under Option B the room count falls from 28.57 to 25.97 — a 9.1% reduction — but the ADR uplift more than offsets it, and annual room revenue per unit of gross floor area rises 14.0%. This works because of the relationship confirmed in the previous section: the marginal rate (¥566/m²) exceeds the average rate (¥354/m²). Since the marginal rate on added area is higher than the average rate of the existing stock, revenue increases by diverting fixed floor area into room size.

That relationship does, however, depend on the baseline ADR level. Using the 2016–2019 cohort’s median estimated ADR of ¥11,200 as the base, the same Option B revenue gain narrows to +7.0%. In projects already designed for high rates, the average rate approaches the marginal rate, so the benefit of adding area thins out. The conclusion is counter-intuitive: the lower the rate positioning of a project, the greater the relative benefit of investing in floor area.

For owners of existing properties, the case is harder than for new builds. With the structure already fixed, the only way to enlarge rooms is to cut their number. If the bathroom is separated purely through layout without adding area, this article’s regression gives little basis for a price premium (configuration dummy coefficient +¥54, 95% confidence interval -¥1,759 to +¥2,456). As our earlier report on renovation downtime showed, a renovation involving wet areas costs ¥3 million per room in construction and swells to ¥3.98 million in effective capex once the opportunity cost of the closure period is included. Spending an effective ¥4 million on work that cannot be expected to raise the rate is difficult to justify.

Going in the direction of creating area — cutting the room count to enlarge each room — does pencil out. The case modelled below consolidates three rooms of 20m² each into two rooms of 30m².

A payback of 8.3 to 11.1 years is clearly worse than the 6.1 years of new-build Option B. Investing in bathroom separation at an existing property is defensible only where several conditions coincide: (1) consolidation opens room to move the room type into a higher tier, (2) existing rooms are 20m² or smaller and the tightness itself shows up in guest appraisals, and (3) the unmet-demand signal discussed below appears at the property in question.

Extracting the properties where the request form “please make the bath and toilet separate” appears yields 253 limited-service properties and 336 mentions in total. The profile of those 253 properties is instructive.

The requests are not coming from small, ageing properties. They come from the urban mid-to-upper tier: around 172 rooms with an estimated ADR of ¥12,600. Guests paying that rate hold bathroom separation as an expectation, so a three-in-one unit turns the gap with that expectation into words. At a property with an estimated ADR of ¥7,300 and 95 rooms, by contrast, it is not expected in the first place and so never surfaces as a complaint.

The bathroom specification decision therefore means different things at different rate levels. In a project targeting above ¥12,000, a separated layout is less a means of raising the rate than a precondition for sustaining it. Around the ¥8,000 band, by contrast, separation does work as a differentiator — but in locations where the marginal rate per square metre is low, it will not reach payback.

Guests clearly value a separated bathroom. There were 50,000 mentions over the most recent 24 months, 91.0% of them positive, and the mention rate grew 15.7% year on year. Supply has responded: in the 2016–2019 opening cohort, 10.2% of properties sell such rooms, 2.8 times the 3.6% of the pre-1989 cohort.

Yet once the price data are carefully aligned within each property, what moves the rate is not the bathroom configuration but the area. Listed price rises +¥1,049 per m² (about ¥570 converted to estimated settled ADR), while the configuration dummy comes to +¥54 with a confidence interval straddling zero. This fact demands a change in how the investment case is phrased — from “separating the bath and toilet raises the rate” to “the floor area added in order to separate them raises the rate.”

As long as the marginal rate exceeds the average rate, diverting fixed gross floor area into room size increases revenue. An additional 3.5m² at ¥2.47 million of capex pays back in 6.1 years, and room revenue per 1,000m² of gross floor area rises 14.0% after absorbing a 9.1% cut in room count. The break point is a marginal rate of ¥400 per m². Checking whether your own site can command that level, from the relationship between price and area by room type nearby, is the practical first step in settling the specification.

Related reading

• 582 Hotels Opening 2026: 30-Room Luxury vs 200+ Scale Polarization