On a hotel or ryokan P&L, “distribution commission” has become a variable cost of a scale second only to payroll and cost of goods. Yet in day-to-day decision-making, this expense is almost always discussed as a rate: is it 8%, or is it 15%? Rates are convenient for comparison, but they are not the unit that operators actually work in. How many yen actually leave the building each time a room is sold — and how many rooms must be sold through the property’s own site each month to earn those yen back? Only once those two numbers are known does channel design become a discussion about figures rather than impressions.

This article multiplies the publicly documented range of distribution commission rates by MetroEngines Research’s estimated transacted ADR (the confirmed trailing 12 months) to derive the commission in yen per room-night, broken out by property type and by prefecture. It then builds up the fixed and variable costs required to shift weight toward direct booking, using published pricing where available, and calculates the monthly volume of direct bookings at which the shift breaks even, by property size. Note that this is a different cut from our earlier work on price gaps between two listing channels and on the share of total rooms placed on OTAs and similar channels. The former looked at listed price differences, the latter at inventory allocation (measured in what share of total rooms is actually placed on OTAs and similar channels); what this article addresses is the commission in actual yen, determined by rate × ADR, and the break-even point that follows from it.

Rates are undisclosed by design — use only the bands visible in published material

One thing has to be established first. The distribution commission rates charged by accommodation booking sites are not published as official rates for any individual company. Rates are set in individual contracts with each property and vary with size, listing tier, and participation in promotional programs. There is therefore no primary source that would allow anyone to state categorically that “Company A charges X%.” This article does not fill that gap with guesswork; it uses only the bands that can be confirmed from public and industry-association material.

The most reliable evidence available is the “FY2025 Survey of Business Conditions” published by the Japan Ryokan & Hotel Association (日本旅館協会) on 5 December 2025, based on FY2024 financial statements with 151 valid responses. The survey reports that operating expense ratio — the combined total of sales promotion, advertising and distribution commission — reached 12.9% of total revenue (up 1.3 points year on year and up 1.3 points versus FY2019), and states explicitly that the driver is that “promotional costs paid to OTAs (online travel agents) have grown larger. Properties participate in various campaigns to attract guests, and the cost burden of doing so has increased.” By size, the figures were 12.7% for large properties, 13.6% for mid-size and 11.4% for small.

The same survey also shows the booking-channel mix. The OTA share averages 46.9% (up 2.0 points year on year and up 16.0 points versus FY2019), with 11.5% booked through properties’ own websites, 11.5% direct but not via the property website, and 30.1% through traditional travel agencies. Among mid-size properties, the OTA share passed the halfway mark at 53.3%.

These two figures are mutually consistent. If the effective commission burden is put at 12%, then 46.9% via OTAs × 12% = 5.6% of revenue as distribution commission. Add commission on the 30.1% booked through travel agencies, plus advertising and other promotional spend on the property’s own site, and a 12.9% operating expense ratio is reached without strain. In other words, the “10–15% effective burden” band also holds when worked backwards from the industry association’s financial statistics. That band is the anchor for everything below. How costs that include distribution commission ultimately flow through to the bottom line is a question of GOP structure, and is beyond the scope of this article.

Source: Japan Ryokan & Hotel Association, “FY2025 Survey of Business Conditions” (published 5 December 2025, 151 valid responses) / compiled by the HotelBank Editorial Team

Applying the rate to national ADR — ¥877 to ¥2,193 per room-night

Now the actual amounts. Aggregating MetroEngines Research’s estimated transacted ADR over the confirmed 12 months from August 2025 to July 2026, the national level across all property types was approximately ¥11,000 (weighted average of the 47 prefectures by number of properties covered, N=16,821 properties). Applying commission rates to that rate gives ¥877 per room-night at 8%, ¥1,316 at 12% and ¥2,193 at 20%.

The national average is a blunt instrument, however. Because ADR levels differ substantially by property type, the same rate produces a burden per room-night that varies by nearly a factor of two.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team (confirmed 12 months, August 2025 to July 2026, estimated transacted ADR)

Source: MetroEngines Research; compiled by the HotelBank Editorial Team

There is roughly a two-fold gap between a business hotel’s ¥1,064 per room-night (at a 12% rate) and a resort hotel’s ¥2,023. Expressed as a rate, the two are identical at 12%. This is precisely where talking in rates breaks down. A business hotel’s lower nightly rate does mean a smaller commission per room-night, but for the same room count the annual volume of room-nights sold is unchanged. Assume 100 rooms at 75% occupancy (an assumption on a full-year basis, a level comparable to the 75.3% recorded for business hotels in the Japan Tourism Agency’s Overnight Travel Statistics Survey, 2025 annual figures) and the property sells 27,375 room-nights a year — which at a 12% rate puts annual distribution commission at roughly ¥29.1mn for a business hotel, ¥50.0mn for a ryokan and ¥55.4mn for a resort hotel.

By prefecture — a spread of up to ¥1,213 per room-night

Even at an identical rate, the commission per room-night changes with the ADR level of the area the property sits in. Applying a 12% rate to the estimated transacted ADR of each of the 47 prefectures, commission per room-night ranges from ¥1,988 in Kyoto to ¥775 in Tokushima — a spread of ¥1,213.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team (N=16,821 properties, confirmed 12 months, August 2025 to July 2026)

At the top of the ranking sit Kyoto (¥1,988), Oita (¥1,789), Tokyo (¥1,756), Kanagawa (¥1,704) and Shizuoka (¥1,627). What stands out here is that prefectures with hot-spring destinations — Oita, Shizuoka, Nagano — fall into the same band as the major metropolitan areas. A high share of higher-rate ryokan is pulling the whole prefecture up.

To see this properly, the same prefectures need to be split by property type. Placing business hotels and ryokan side by side on commission per room-night makes the structural differences between prefectures explicit.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team (converted at a 12% rate, based on estimated transacted ADR by property type)

In Kyoto, commission per room-night is ¥2,939 for ryokan against ¥1,686 for business hotels, a gap of ¥1,253. Kanagawa (ryokan ¥2,571, business ¥1,111) and Hyogo (ryokan ¥2,680, business ¥918) show the same pattern, with ryokan well ahead. Tokyo runs the other way — ryokan ¥1,319 against business hotels ¥1,713 — and Okinawa is close to level at ryokan ¥985 and business ¥1,061. “Ryokan carry the higher commission per room-night” is therefore not a nationwide rule but a structure that shows up sharply in hot-spring and leisure-destination prefectures.

The cost structure of direct booking — split into fixed and variable

This is the heart of the matter. Knowing the commission in yen does not by itself tell you how much is “saved by moving to direct,” because direct booking carries costs of its own. The key is to build those costs up split between variable costs (incurred only when a room sells) and fixed costs (incurred whether it sells or not).

■ Assumptions

Items with confirmable published pricing are shown separately from those that could not be confirmed and are set as assumptions in this article.

Source: published pricing and IR disclosures; compiled by the HotelBank Editorial Team (simplified calculation including assumptions)

The remaining assumptions are as follows. (1) Occupancy is set at 75% as the base case, with 65% and 85% also tested in the sensitivity analysis. (In the Japan Tourism Agency’s Overnight Travel Statistics Survey, 2025 annual figures — preliminary, published 27 February 2026 — occupancy was 75.3% for business hotels and 74.2% for city hotels, so for these two types the 75% base case is close to actual. Ryokan at 38.4% and resort hotels at 56.9% diverge from it, so a restatement at actual occupancy is provided later.) (2) The transacted rate on direct bookings is assumed identical to that via OTAs and similar channels, with no price differential. (3) Loyalty-program point funding and the property’s own coupons are not included in variable costs. (4) Payroll (handling direct reservations and enquiries) is assumed absorbed within existing staffing, with no incremental cost booked. (5) Channel manager costs are not booked, on the assumption the system is already in place (the same Japan Ryokan & Hotel Association survey puts adoption at 97.3%).

On these assumptions, the incremental margin from moving one room-night to direct booking is “ADR × (distribution commission rate − 4.84%).” At a 12% rate that is ¥635 per room-night for a business hotel (ADR ¥8,900) and ¥1,090 for a ryokan (ADR ¥15,200).

How many rooms a month is break-even — modelled at 30, 100 and 200 rooms

Dividing the ¥180,000 of fixed costs by the incremental margin gives the monthly volume of direct bookings at which the shift breaks even. At a 12% rate, that is 284 room-nights per month for a business hotel, 182 for a city hotel, 165 for a ryokan and 149 for a resort hotel.

Converting that monthly volume against room count gives the required direct-booking share. At 75% occupancy (an assumption, full-year basis), a 30-room property sells 675 room-nights a month, a 100-room property 2,250 and a 200-room property 4,500.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team (simplified calculation assuming 75% occupancy on a full-year basis, ¥180,000/month fixed costs and 4.84% variable costs)

Source: MetroEngines Research; compiled by the HotelBank Editorial Team

The result is unambiguous. At 200 rooms, any property type recovers its fixed costs at a direct-booking share of 3–7%. Given that the Japan Ryokan & Hotel Association survey puts the average share booked through properties’ own websites at 11.5%, many properties are already past that level. At 100 rooms the requirement is 6.6–12.6%, which is likewise within realistic range.

At 30 rooms, by contrast, 22.1–42.0% is required. The 42.0% for business hotels is a high bar against the 11.5% national average for own-website bookings. This is not an argument that direct booking does not suit small properties; it shows that how the fixed costs are set moves the break-even point substantially. Drop the ¥100,000 advertising line, or hold maintenance costs down by working with an existing website, and compressing fixed costs to ¥80,000 brings the required direct-booking share for a 30-room business hotel down to 18.7%. The smaller the property, the more the design of the fixed cost base itself determines the break-even point.

Sensitivity analysis — flexing the commission rate and occupancy

Because the calculation rests on a number of assumptions, the main variables are flexed to check how robust the result is.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team (required direct-booking share = fixed costs ÷ margin per room-night ÷ monthly room-nights sold)

Restated at actual occupancy — the break-even rises for ryokan and resorts

Everything so far applies a uniform full-year occupancy of 75% to every property type. But occupancy in the Japan Tourism Agency’s Overnight Travel Statistics Survey, 2025 annual figures (preliminary), varies widely by type: 75.3% for business hotels and 74.2% for city hotels, against 56.9% for resort hotels and 38.4% for ryokan (61.8% overall). The break-even volume of monthly direct bookings does not depend on occupancy, but the denominator used to convert that volume into a “required direct-booking share” — monthly room-nights sold — is proportional to occupancy, so the lower a type’s occupancy, the higher its required direct-booking share. Restated at actual values, the picture is as follows.

Source: occupancy by property type from the Japan Tourism Agency’s Overnight Travel Statistics Survey, 2025 annual figures (preliminary, published 27 February 2026), and MetroEngines Research; compiled by the HotelBank Editorial Team (required direct-booking share = fixed costs ÷ margin per room-night ÷ monthly room-nights sold)

Business hotels and city hotels sit almost exactly on the 75% base case, so the figures given earlier can be used as they stand. The divergence is in ryokan and resort hotels: the required direct-booking share for a 30-room ryokan rises from 24.5% to 47.7%, and for 100 rooms from 7.3% to 14.3%. At 14.3%, a 100-room ryokan is above the 11.5% average share booked through properties’ own websites in the Japan Ryokan & Hotel Association survey, so the conclusion that “100 rooms is already past break-even” does not transfer to ryokan as it stands. Resort hotels come in at 29.1% for 30 rooms and 8.7% for 100, a smaller divergence from the base case than ryokan. For types with strong seasonality, applying the property’s own occupancy separately for peak and off-peak periods, rather than a full-year average, will track reality more closely.

Flexing the rate from 10% to 15% roughly halves the required direct-booking share. Moving occupancy from 65% to 85% produces a comparatively small change, which shows that the break-even point is driven principally by the commission rate and by fixed costs, with occupancy a second-order factor. For 100-room properties, the required share stays within a 5.2–17.5% range under every condition tested — the same band as the 11.5% national average for own-website bookings.

How to use the break-even point — designing for addition, not replacement

What the calculations point to can be summarised as follows.

First, commission per room-night varies by nearly a factor of four across property types and areas. Against ¥1,316 nationally across all types (at a 12% rate), the actual amount in the business-hotel band in Tokushima and the ryokan band in Kyoto are entirely different figures. Describing your own property in terms of the national average leads to misjudging which levers matter. The place to start is your own ADR × your own rate, to get the actual yen per room-night.

Second, at 100 rooms and above, the share required to cover the fixed costs of direct booking is already within reach. Against a national average of 11.5% booked through properties’ own websites, break-even was 6.6–12.6% at 100 rooms and 3.3–6.3% at 200. For many mid-size and larger properties, then, direct booking is not “an investment whose fixed costs are still to be recovered” but “an area where recovery is already complete and every additional booking drops straight through as margin.” That framing does, however, assume 75% full-year occupancy: for ryokan, whose actual occupancy is lower, even a 100-room property needs 14.3%, above the average own-website share (see the restatement at actual occupancy above). From that point on, each additional room-night generates ¥635 to ¥1,207 of margin depending on property type.

Third — and this deserves emphasis — none of this is an argument that one channel is better than another. Distribution through OTAs and similar channels carries clear value: reach into demand a property cannot access on its own, fill outside peak periods, and capture inbound demand. That the OTA share in the Japan Ryokan & Hotel Association survey has grown to 46.9%, up 16.0 points versus FY2019, is evidence that the value is real. It is equally telling that in the same survey 97.3% of respondents sell “through both their own website and OTAs,” while 0.0% sell “through their own website only.” The practical choice is a question of allocation, not replacement.

Fourth, the useful application of a break-even point is not “what percentage to move to direct” but “which demand to place on direct”. The largest margin per room-night comes from higher-rate bookings — by property type, resort hotels at ¥1,207 and ryokan at ¥1,090 lead. Multi-night stays, anniversaries, repeat guests, late incremental demand: work outward from the areas where a property’s own membership base and guest relationships do the work, and fixed-cost recovery accelerates. Conversely, new audiences that require awareness-building, and demand from distant or overseas markets, are where distribution earns its value. On how the entry point for those new audiences is itself changing, preparing your data for the era of asking an AI where to stay is a useful reference. Knowing the actual yen per room-night lets that allocation be decided in money rather than by feel. That is the point of this article.

Frequently asked questions

Related reading

• Hotel Acquisition Costs 7.1%: A 6.5% Gross Yield Becomes 4.65% Net
• Japan’s ¥2.83tn Hotel Room Revenue Pool: Top 5 Prefectures 48.0%
• Revenue Management vs Dynamic Pricing: 19,365 Hotels, Median 4 Moves
• Japan Lodging Tax 2026: All 62 Municipalities, Rates and ADR Burden
• Median 39.2% of Rooms Online — Allocation Across 12,941 Hotels
• Google UCP & ChatGPT: 3 Data Fixes Hotels Should Make Now in 2026
• Intra-Prefecture ADR Gaps 2026: Median 2.74x, Okinawa Widest at 5.86x