Estimated OCC (based on OTA-listed inventory) for Hiroshima Prefecture’s business hotels (N=93 properties, 11,025 rooms) averaged 88.2% in August 2026. Averaged by day of week, Thursday reached 93.6%, Tuesday 93.2% and Wednesday 92.5%, while Sunday held at 79.6% and Monday at 83.2%. July showed the same two-night trough at the start of the week: Sunday 72.8% and Monday 75.2% against Thursday 88.7% and Saturday 89.7%. On an average August Sunday, 2,241 rooms of OTA-listed inventory were still unsold when the stay date arrived—more than three times Thursday’s 707 rooms. This article sets out how to fill that trough through three routes—“extending Saturday guests by one night”, “multi-night stays checking in on Sunday” and “business travellers arriving the night before a Tuesday-or-later trip”—alongside a snapshot of October booking pace.
Scope: Hiroshima Prefecture business hotels, N=93 properties, 11,025 rooms (mode for July and August 2026; N=95 properties, 11,118 rooms in September). City hotels, shown for reference, are N=14 properties, 1,662 rooms. This article covers occupancy only (estimated OCC = an estimate based on OTA-listed inventory) and does not cover rate levels or year-on-year change. Definitions are at the end of the article. Data as of: September 30, 2026.
- — The trough is the first two nights of the week: in August 2026, estimated OCC (based on OTA-listed inventory, N=93 properties, 11,025 rooms) was 79.6% on Sunday and 83.2% on Monday, against 93.6% on Thursday and 93.2% on Tuesday. The gap to Thursday is 14.0pt.
- — The same shape for three months: July also showed Sunday 72.8% and Monday 75.2% against Thursday 88.7% (a 16.9pt gap). Excluding the holiday week, September Sundays stayed at 71.3–76.0%.
- — Unsold inventory concentrates on Sunday: on an average August Sunday, 2,241 rooms of OTA-listed inventory were left, more than three times Thursday’s 707. Sunday and Monday together came to 4,091 rooms, 1.8 times the Tuesday–Thursday three-night total.
- — The trough reappears in October: October 18 (Sun) and 19 (Mon) stood at 68.8% and 64.5% at 45 days out, and their pickup since then (+12.6pt, +13.3pt) is also smaller than Thursday’s (+19.9pt).
- — The lever is selling conditions, not price: set up Saturday extensions, multi-night stays restricted to Sunday check-in, and a design that does not apply multi-night conditions to Tuesday-or-later check-ins—and return to standard selling in holiday weeks.
Estimated OCC by Day of Week (August 2026) — Thursday 93.6% vs Sunday 79.6%, a 14pt Spread
The day-of-week pattern of Hiroshima Prefecture’s business hotels is “double-peaked”: high from Tuesday through Thursday, dipping on Friday, recovering on Saturday and lowest on Sunday and Monday. August’s day-of-week averages (N=93 properties, 11,025 rooms) were Thursday 93.6%, Tuesday 93.2%, Wednesday 92.5%, Saturday 90.0%, Friday 88.2%, Monday 83.2% and Sunday 79.6%, so the spread between the highest day (Thursday) and the lowest (Sunday) reaches 14.0pt. July (N=93 properties, 11,025 rooms, monthly average 83.9%) ran at a lower level than August but had the same shape—Saturday 89.7% and Thursday 88.7% against Sunday 72.8% and Monday 75.2%—and an even wider spread of 16.9pt. The same direction of day-of-week spread was already visible in a late-July snapshot, when Hiroshima showed a Sunday trough of 8.5pt alongside Aichi’s 7.1pt.
The trough becomes clearer once public-holiday placement is stripped out. Of August’s five Mondays, August 10 (the eve of Mountain Day) reached 91.7%, but the other four ranged from 69.6% to 87.7% and averaged 81.1%. Of July’s four Sundays, July 19—the middle day of a three-day weekend—reached 89.1%, but the other three were 59.3%, 68.1% and 74.9%, averaging only 67.4%. In September (N=95 properties, 11,118 rooms, monthly average 89.5%), the Sunday and Monday of the five-day holiday (September 19–23) hit 100.0%, while the three non-holiday Sundays were 71.3%, 73.5% and 76.0%. The picture of “Sundays in the 70% range in ordinary weeks” holds across all three months.
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Figure 1: Estimated OCC by day of week, Hiroshima Prefecture business hotels (based on OTA-listed inventory; each month’s daily values averaged by day of week). July and August N=93 properties, 11,025 rooms; September N=95 properties, 11,118 rooms. September’s Monday includes September 21 (100.0%), during the five-day holiday.
Translating the day-of-week levels into “volume of unsold inventory” makes the operational meaning clear. On August Thursdays, the share of properties with no OTA-listed inventory visible (sold-out property share, estimated) averaged 40.2%, and the rooms left unsold across the whole cohort averaged 707. Sunday showed a sold-out share of 19.9% and 2,241 unsold rooms; Monday 19.4% and 1,850 rooms. Sunday and Monday together average 4,091 rooms, which means 1.8 times as much inventory as the Tuesday–Thursday three-night total (2,276 rooms) is left over at the start of the week.
| Day | Jul est. OCC | Aug est. OCC | Sep est. OCC | Aug sold-out share | Aug avg unsold rooms |
|---|---|---|---|---|---|
| Mon | 75.2% | 83.2% | 92.5% | 19.4% | 1,850 rooms |
| Tue | 87.0% | 93.2% | 92.6% | 40.6% | 751 rooms |
| Wed | 87.4% | 92.5% | 90.8% | 40.2% | 818 rooms |
| Thu | 88.7% | 93.6% | 94.0% | 40.2% | 707 rooms |
| Fri | 84.0% | 88.2% | 86.2% | 22.7% | 1,297 rooms |
| Sat | 89.7% | 90.0% | 89.3% | 34.5% | 1,088 rooms |
| Sun | 72.8% | 79.6% | 80.2% | 19.9% | 2,241 rooms |
| Monthly avg | 83.9% | 88.2% | 89.5% | — | — |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Table 1: Day-of-week summary, Hiroshima Prefecture business hotels. Estimated OCC is based on OTA-listed inventory. Sold-out share is the share of properties with no listed inventory visible on OTAs and other channels (estimated). Unsold rooms is the average number of rooms left as OTA-listed inventory across the whole cohort. July and August N=93 properties, 11,025 rooms; September N=95 properties, 11,118 rooms.
For reference, city hotels (N=14 properties, 1,662 rooms) averaged 90.4% in August: Wednesday 94.0%, Tuesday 92.6% and Thursday 92.6% against Sunday 87.1% and Monday 85.8%, a maximum day-of-week spread of 8.2pt—about half that of business hotels. City hotels, with a higher leisure share, see stays stretch into Sunday night more easily, whereas business hotels’ corporate demand concentrates on Tuesday–Thursday; that structure shows up as the day-of-week spread.
Daily Trend in August — the Start-of-Week Trough Arrives Every Week
Following August day by day shows the trough is not confined to particular weeks. After 88.7% on August 1 (Sat), occupancy fell to 77.5% on August 2 (Sun) and 82.9% on the 3rd (Mon), then jumped back to 97.0% on the 4th (Tue). After 90.5% on August 22 (Sat), the 23rd (Sun) was 76.5%, the 24th (Mon) 84.3% and the 25th (Tue) 91.2%—the same shape again. August 30 (Sun) at 70.7% and 31 (Mon) at 69.6% at month-end were the two lowest nights of the month, recovering through the start of the week to 82.1% on September 1 (Tue) and 92.3% on the 2nd (Wed).
The exception is the Obon week. Occupancy held high from 95.8% on August 8 (Sat) to 92.6% on the 9th (Sun) and 91.7% on the 10th (Mon), with no trough through 89.7% on the 11th (Tue, Mountain Day). Since the trough disappears in weeks such as Obon and holiday weekends, when leisure stays stretch into Sunday night, the trough in ordinary weeks can be read as two gaps overlapping: “Saturday leisure guests go home on Sunday” and “business travellers’ arrivals shift to Monday night or later”.
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Figure 2: Daily estimated OCC, August 2026 (based on OTA-listed inventory). Business hotels N=93 properties, 11,025 rooms; city hotels N=14 properties, 1,662 rooms (both modes).
October Booking Pace — Sunday and Monday Start Low at 45 Days Out, and the Gap Does Not Close
The start-of-week trough shows up not only in past months’ results but also in the booking pace of stay dates still to come. Comparing booking curves from 45 days out to the latest observation (N=98 properties, 11,141 rooms) for four nights—October 15 (Thu), 17 (Sat), 18 (Sun) and 19 (Mon)—Saturday stood at 86.7% at 45 days out against Thursday 66.2%, Sunday 68.8% and Monday 64.5%; at 30 days out, Saturday 90.1%, Thursday 74.6%, Sunday 76.5% and Monday 72.4%; and at the latest observation (as of September 30), Saturday 94.7% (17 days out), Thursday 86.1% (15 days out), Sunday 81.4% (18 days out) and Monday 77.8% (19 days out).
That week’s Thursday built up +19.9pt from 45 days out to the latest observation, a larger pickup than Sunday (+12.6pt) or Monday (+13.3pt). In other words, weekdays fill with last-minute business bookings, while Sunday and Monday stay sluggish to the end—the same trough as in August’s results is being reproduced in October. The following week, October 22 (Thu), 24 (Sat), 25 (Sun) and 26 (Mon), looks the same: at 30 days out, Saturday 81.0% and Thursday 72.1% against Sunday 66.3% and Monday 62.0%, with Sunday and Monday at the bottom. For the average shape of Hiroshima’s curves, our fixed-point tracking of business and city hotels put estimated OCC at 72.2% at 45 days out, 76.3% at 30 days out and 91.8% just before the stay date.
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Figure 3: Booking curves for October 15 (Thu), 17 (Sat), 18 (Sun) and 19 (Mon), 2026 (estimated OCC, based on OTA-listed inventory). The x-axis is days remaining until the stay date (45 days out → latest). N=98 properties, 11,141 rooms. Based on observations from 45 days before the stay date to the latest.
| Stay date | 45 days out | 30 days out | Latest observation | Pickup, 45 days out → latest |
|---|---|---|---|---|
| Oct 15 (Thu) | 66.2% | 74.6% | 86.1% (15 days out) | +19.9pt |
| Oct 17 (Sat) | 86.7% | 90.1% | 94.7% (17 days out) | +8.0pt |
| Oct 18 (Sun) | 68.8% | 76.5% | 81.4% (18 days out) | +12.6pt |
| Oct 19 (Mon) | 64.5% | 72.4% | 77.8% (19 days out) | +13.3pt |
| Oct 22 (Thu) | 63.6% | 72.1% | 78.1% (22 days out) | +14.5pt |
| Oct 24 (Sat) | 73.0% | 81.0% | 84.1% (24 days out) | +11.1pt |
| Oct 25 (Sun) | 59.2% | 66.3% | 68.1% (25 days out) | +8.9pt |
| Oct 26 (Mon) | 53.5% | 62.0% | 64.0% (26 days out) | +10.5pt |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Table 2: Booking pace for eight October nights (estimated OCC, based on OTA-listed inventory). N=98 properties, 11,141 rooms. The latest observation is as of September 29–30, 2026, so days remaining differ by stay date. Sales status changes daily.
For Revenue Managers Running Business Hotels in Hiroshima Prefecture — Implications and Action Plan
Operational insights
1. Treat the trough as a two-night set. August’s Sunday 79.6% and Monday 83.2% are not two separate low-occupancy days but one continuous trough between Saturday’s 90.0% and the return to 93.2% on Tuesday. Line up your own day-of-week sales on the same yardstick (the four-night sequence Sat → Sun → Mon → Tue) and first establish whether your Sunday or your Monday falls further below the market. A property where only Sunday is deep should focus on extending Saturday guests; one where only Monday is deep should focus on business travellers arriving the night before a Tuesday-or-later trip.
2. Sunday’s 2,241 unsold rooms are not inventory that “fills if you cut the price”. In the October snapshot, Sunday and Monday were already low at 45 days out, and their pickup to the latest observation was smaller than Thursday’s. Moving rates alone on days where demand itself is thin risks eroding the week’s overall revenue, which is underpinned by high weekday occupancy. What needs filling is room nights, so prioritise designs that reshape demand through selling conditions (multi-night stays, check-in-day restrictions) rather than price.
3. Thursday at 93.6% with a 40.2% sold-out share puts mid-week on the defensive. Tuesday to Thursday is a seller’s market where the sold-out property share reaches 40%; opening multi-night discounts broadly there risks diluting high-rate single-night demand. Limiting multi-night conditions to combinations that include a Sunday or Monday check-in, and not applying them to multi-night stays checking in on Tuesday or later, lifts the trough while holding mid-week rates.
4. Split the calendar on the assumption that the trough disappears in holiday weeks. About once a month there is a week, like Obon (August 9 at 92.6%, August 10 at 91.7%) or September’s five-day holiday, in which Sunday and Monday run at high occupancy. Calendar management is needed so that multi-night conditions and Sunday-check-in offers are “valid in ordinary weeks only”, with holiday weeks reverting to standard single-night selling.
Action plan (with timeline)
| Timeline | Action | Decision trigger (check against the figures in this article) | Aim |
|---|---|---|---|
| Today to this week | Compare your own booking pace for October 18 (Sun), 19 (Mon), 25 (Sun) and 26 (Mon) against the market curve | The market stands at 80.7% for October 18 (Sun) and 77.4% for 19 (Mon) at 20 days out, 68.1% for 25 (Sun) at 25 days out, and 62.0% for 26 (Mon) at 30 days out. If you are below these, open multi-night conditions on Sunday and Monday immediately | Do not leave the start-of-week two-night trough unattended until the last minute |
| Today to this week | Offer a two-night stay checking in on Saturday (Sat and Sun nights) on better terms than a Saturday single night | Saturday 90.0% vs Sunday 79.6% in August. Fill part of Sunday’s 2,241 unsold rooms by extending Saturday leisure guests by one night | Stretch Saturday demand into Sunday night |
| Within two weeks | Set up two-night (Sun–Mon nights) and three-night (Sun–Tue nights) stays restricted to Sunday check-in | Sunday 79.6% and Monday 83.2% leave 4,091 unsold rooms across the two nights. Tuesday at 93.2% fills with single nights, so restrict the check-in days for multi-night conditions to Sunday and Monday | Capture business travellers arriving the night before, from Sunday night |
| Within two weeks | Do not apply multi-night conditions to stays checking in on Tuesday or later (protect mid-week single-night sales) | Tuesday–Thursday sold-out shares are in the 40% range, with 700–820 unsold rooms. There is no need to open multi-night discounts there | Hold mid-week rates |
| Toward next month | In holiday weeks (such as November’s three-day weekends), suspend Sunday/Monday multi-night conditions and revert to standard selling | In August’s Obon week Sunday was 92.6% and Monday 91.7%; in September’s five-day holiday both Sunday and Monday were 100.0%. Leaving multi-night conditions in place in weeks without a trough means lost opportunity | Separate ordinary weeks from holiday weeks on the calendar |
| Toward next month | Make estimated OCC and unsold rooms by day of week a fixed monthly report, and track the depth of the Sunday/Monday trough (the gap to Thursday) | Using August’s Thursday−Sunday gap of 14.0pt and July’s 16.9pt as baselines, check whether your own gap is narrowing | Measure the effect of measures by the narrowing of the day-of-week spread |
Table 3: Action plan. These are not guaranteed outcomes but material to consider after checking the figures in this article against your own property’s situation.
Summary — Three Yardsticks for the Start-of-Week Trough
The day-of-week pattern of Hiroshima Prefecture’s business hotels has been stable for three months in the shape “Thursday at the peak, Sunday at the bottom, Monday on the way back”. When reviewing your own day-of-week strategy, keep the following three points as yardsticks.
| Yardstick | Market value (August 2026, N=93 properties) | What to look at |
|---|---|---|
| Day-of-week spread (Thursday − Sunday) | 93.6% − 79.6% = 14.0pt | If your spread is wider than the market, review Sunday selling conditions; if narrower, review mid-week rates |
| Unsold inventory on the first two nights of the week | Sunday 2,241 rooms + Monday 1,850 rooms = 4,091 rooms (1.8 times the Tuesday–Thursday three-night total of 2,276 rooms) | The scale of “room nights” to fill with multi-night conditions |
| Holiday-week exception | Obon week: Sunday 92.6%, Monday 91.7% | Limit multi-night conditions to ordinary weeks and revert to standard selling in holiday weeks |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
Because the Sunday–Monday trough stems from thin demand, the first lever is a design that lengthens stays through selling conditions rather than price. Consider building the three elements—Saturday extensions, Sunday-check-in multi-night stays and an exclusion that protects mid-week—as a set, and measuring the effect by how much the day-of-week spread narrows. Neighbouring Okayama also showed a 17.8pt Sunday trough in August (Wednesday 92.4% vs Sunday 74.6%), which reads as a structure shared across the Setouchi business-travel demand area.
Frequently Asked Questions
Q. Which day of the week has the lowest occupancy at business hotels in Hiroshima Prefecture?
A. Sunday and Monday. In August 2026, estimated OCC (based on OTA-listed inventory, N=93 properties) was 79.6% on Sunday and 83.2% on Monday, more than 10pt below Thursday’s 93.6% and Tuesday’s 93.2%. July had the same shape, with Sunday at 72.8% and Monday at 75.2%.
Q. Can the Sunday–Monday trough be filled by cutting prices?
A. The data in this article make that hard to conclude. In October booking pace (as of September 30, 2026), Sunday and Monday were already low at 45 days out and their pickup to the latest observation was smaller than Thursday’s, suggesting days where demand itself is thin. There is room to prioritise designs that lengthen stays through selling conditions—multi-night stays or check-in-day restrictions—rather than price.
Q. Which days should multi-night conditions apply to?
A. One option is to limit them to combinations that include a Sunday or Monday check-in. Tuesday to Thursday is a seller’s market where the sold-out property share reached about 40% in August 2026, and opening multi-night discounts broadly there risks diluting single-night demand.
Q. Do the same measures work in holiday weeks?
A. Holiday weeks need separate treatment because the trough disappears. In August 2026’s Obon week Sunday was 92.6% and Monday 91.7%, and in September’s five-day holiday both Sunday and Monday were 100.0%. Calendar management is needed so that multi-night conditions are valid in ordinary weeks only, with holiday weeks reverting to standard single-night selling.
About the Data
| Definition of estimated OCC | Occupancy based on OTA-listed inventory = 100 − 100 × OTA-listed unsold rooms ÷ total rooms. Covers a fixed cohort of properties that list at least 30% of their total rooms on OTAs (the same set of properties on each stay date; N is that number of properties). It is an estimate based on how inventory sold on OTAs is being absorbed and is defined differently from actual room occupancy (rooms not listed are counted as sold, so it reads high). |
| Booking curve | Based on observations from 45 days before the stay date to the latest. Day-of-week figures average each month’s daily results (observed just before the stay date) by day of week. |
| Sold-out property share | The share of properties with no listed inventory visible on OTAs and other channels (estimated). Unsold rooms is the average of the total rooms left as OTA-listed inventory across the whole cohort. |
| Scope | Hiroshima Prefecture business hotels. July and August 2026: N=93 properties, 11,025 rooms; September: N=95 properties, 11,118 rooms (both the mode of daily N; the count varies by 1–6 properties from day to day). October booking curves: N=98 properties, 11,141 rooms. City hotels, shown for reference: N=14 properties, 1,662 rooms. Based on listings on major booking sites. |
| Rates and year-on-year | This article covers occupancy only and does not cover rate levels or year-on-year change. |
| Data as of | Data as of: September 30, 2026. Sales status and inventory change daily, so the figures in this article are a snapshot at the time of collection. |
Source: MetroEngines Research; compiled by the HotelBank Editorial Team
References and Sources
■ Data source
Daily observations of listed inventory on major booking sites collected by MetroEngines Research. Hiroshima Prefecture business hotels N=93 properties, 11,025 rooms (mode for July and August 2026); September N=95 properties, 11,118 rooms; October booking curves N=98 properties, 11,141 rooms. City hotels, shown for reference, N=14 properties, 1,662 rooms.
■ Calculation assumptions
Estimated OCC = 100 − 100 × OTA-listed unsold rooms ÷ total rooms. Day-of-week values are simple averages, by day of week, of each month’s daily results (observed just before the stay date). Unsold rooms is the daily total of rooms left as OTA-listed inventory, averaged by day of week. October booking pace was compared at three points: 45 days out, 30 days out and the latest observation (September 29–30, 2026).
■ Limitations and caveats
This is an estimate based on OTA-listed inventory and is defined differently from actual room occupancy (rooms not listed are counted as sold, so it reads high). The number of properties covered varies by 1–6 from day to day. Final occupancy for October stay dates is not yet settled. Rate levels and year-on-year change are not covered.