Business hotels in Miyagi Prefecture held a high monthly average estimated OCC (based on OTA-listed inventory) of 91.6% in September 2026, yet broken down by day of week there is an 8.8-point step between Thursday at 95.2% and Sunday at 86.4%. Excluding the five-day Silver Week holiday (September 19–23), Sunday falls to 82.8%; the final night of the holiday, September 23 (Wed, national holiday), came in at 77.2%, and the following Sunday, September 27, at 78.8%. In every case, the month’s low points were concentrated on “nights followed by a working day.” In a market led by business demand, it is entirely natural for sell-outs to stack up on Thursdays while inventory remains on Sundays and on the last night of holidays. The real question is whether to write this trough off as unavoidable or to build it into how the month is sold. This article reviews day-of-week results for August through October, overlays booking curves (from 45 days before the stay date up to the latest observation) for upcoming Sundays and holiday final nights, and lays out a practical framework for filling those trough nights.

Scope: Miyagi Prefecture business hotels, N=58–62 properties (8,464–8,961 rooms, based on listings on major booking sites, September 2026). The price metric in this article is the estimated settled ADR (the transaction price level estimated from OTA and other sales data, tax-exclusive equivalent); occupancy is an estimate based on OTA-listed inventory. Both are defined at the end of the article. The day-of-week analysis covers occupancy only and does not address day-of-week pricing. Data as of: October 19, 2026.

Estimated OCC by Day of Week — A “Weekday-Type” Market Peaking on Thursday with a Sunday Trough

Averaging September’s 30 days by day of week, Thursday was the peak at 95.2% (4 days), followed by Saturday 93.8%, Tuesday 92.4%, Monday 92.3%, Friday 90.9% and Wednesday 90.1%, with Sunday the lowest at 86.4% (4 days). Saturday ranks second behind Thursday because leisure and family-visit stays are layered on top, so this Sendai-centred market is not a pure “weekday type” but a blend of weekday business and weekend leisure demand. Even so, Sunday alone dips clearly, most likely because Sunday night falls in the gap between demand flows: leisure guests who stayed Saturday head home on Sunday, while business travellers for the week starting Monday have not yet arrived.

That this pattern is not unique to September is confirmed by setting August and October (1st–19th) side by side. In August, Wednesday 96.0%, Thursday 95.8% and Tuesday 95.0% were high, while Monday 88.1% and Sunday 88.0% were low. In October (1st–19th), Thursday 97.1% and Saturday 95.8% were high, while Sunday 88.0% and Monday 85.1% were low. Months where Monday comes out low were largely driven by the “final night of a holiday or long break” falling on a Monday — such as August 31 (Mon, 79.7%) after the Obon period, and October 12 (Mon, Sports Day holiday, 78.4%). Setting the calendar aside, the structure of “Tuesday through Thursday high, Sunday (and holiday final nights) low” held throughout all three months. Aichi Prefecture showed the same kind of step in September, with Thursday at 92.5% versus 84.8% on Sunday excluding holidays; a comparison that includes the trough that also appears on Friday there is covered in Aichi Business Hotels: Thu 92.5%, Sun 84.8% OCC — A Friday Dip Too.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Miyagi Prefecture business hotels, N=56–62 properties; estimated OCC based on OTA-listed inventory; October covers the 1st–19th)

Source: MetroEngines Research; compiled by HotelBank Editorial Team (estimated OCC based on OTA-listed inventory. August N=56–58 properties, September N=58–62, October N=56–62. Figures in parentheses are the number of days for that weekday)

To add a monthly view on pricing, the finalized estimated settled ADR for Miyagi Prefecture business hotels in September 2026 was approximately ¥8,500 (N=142 properties), up 17.3% from approximately ¥7,300 in September 2025 (N=140). August, at approximately ¥8,600 (N=142), was likewise 6.5% above the prior-year figure of approximately ¥8,100 (N=140). In a market where rates are growing by double digits while occupancy holds in the 90% range, the remaining challenge is not “raising the average” but how to handle the inventory left over on one specific night: Sunday.

Daily Trend in September — Sunday Is 82.8% Excluding Holidays, and the Low Points Are “Nights Before a Working Day”

That said, September’s Sunday average of 86.4% includes September 20 (Sun, 97.2%), the middle of Silver Week. Excluding this five-day holiday (September 19–23), the remaining three Sundays came in at 83.9%, 85.8% and 78.8%, averaging just 82.8% — a 10.5-point gap versus the 93.3% average of the four Wednesdays, likewise excluding holidays. Looking only at day-of-week averages, Sunday appears merely “somewhat weak,” but once calendar effects are removed, the underlying Sunday consistently sat more than 10% below weekdays.

The holiday itself, on the other hand, was strong. September 19 (Sat) reached 95.9%, the 20th (Sun) 97.2%, the 21st (Mon, holiday) 96.3% and the 22nd (Tue, holiday) 94.4% — four consecutive nights above 94%. The share of properties with no listed inventory on OTAs and other channels (estimated) was also 59.7% on the 19th, 61.9% on the 20th and 57.1% on the 21st, above a normal Thursday (20–46%). Then, on the holiday’s final night, September 23 (Wed, holiday), occupancy plunged to 77.2% before rebounding to 97.4% on the 24th (Thu). This pattern of “only the final night dropping 16–17 points from the night before” recurred on October 12 (Mon, holiday) at 78.4%, suggesting that a holiday’s final night should be treated the same as a Sunday regardless of the day of week. How this five-day holiday built up night by night — with the first three nights and the last two nights diverging in booking pace — is analyzed in detail in Miyagi Sept Holiday: First 3 Nights 99% at T-14, Last 2 Gain 15pt+.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Miyagi Prefecture business hotels, N=58–62 properties, 8,464–8,961 rooms; estimated OCC based on OTA-listed inventory; September 1–30, 2026)

Looking at the trough nights from the inventory side, on the three September Sundays excluding the holiday, the share of properties with no listed inventory (estimated) was 7.9%, 13.4% and 10.9%, averaging 10.7% — just over one-third of the 29.5% average across the four Thursdays. In other words, Sunday is not a night when “only some properties fail to fill,” but one when nearly every property in the market reaches the stay date still holding inventory. On such nights, a single property cutting its rate is unlikely to absorb the market-wide vacancy, so the main battleground shifts to non-price levers: designing multi-night stays, bundling with the previous night (Saturday), and capturing corporate demand tied to Monday mornings.

Booking Curves for Upcoming Sundays and Holiday Final Nights — A Gap Open from T-45

So does this Sunday trough appear only close to the stay date, or is it visible early on? Among the stay dates observable as of October 19, the following curves overlay estimated OCC (based on OTA-listed inventory, N=62 properties / 9,214 rooms) from 45 days out to the latest observation for four dates: Thursday, October 22; Sunday, October 25; and, within the Labor Thanksgiving Day three-day weekend, November 22 (Sun, middle night) and November 23 (Mon, holiday, final night).

Thursday, October 22 stood at 67.8% at T-45, 79.8% at T-30 and 97.9% at the latest observation (T-4), building +30.1 points over 45 days. Sunday, October 25, by contrast, was at 61.8% at T-45, 68.2% at T-30 and 82.2% at T-7: already 6.0 points below Thursday at T-45, with the gap widening to 11.6 points by T-30. The Sunday trough shows up not as a last-minute stall but as a slower build from an early stage — by T-30, one can already read that “the whole market is at around 70% for this Sunday.”

The holiday pattern is even clearer. While the middle night, November 22 (Sun), rose steadily from 70.9% at T-45 to 79.2% at T-35, the final night, November 23 (Mon, holiday), was at 45.8% at T-45 and 50.1% at T-36 — a gap of roughly 29 points between adjacent nights of the same holiday. Since both dates are still more than 30 days away, we limit ourselves to describing current sales status, but it is clear that the “final-night dip” seen on September 23 and October 12 is already taking shape more than a month ahead for the November holiday. For background on the property-type difference within Miyagi, where the T-45 gap between city hotels and business hotels narrows toward the stay date, see Miyagi City vs Business Hotels: 29.5pt T-45 OCC Gap Narrows to 4.7pt (a comparison for stays on October 17, 2026).

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Miyagi Prefecture business hotels, N=62 properties / 9,214 rooms, based on listings on major booking sites; estimated OCC based on OTA-listed inventory; observations from 45 days before each stay date through October 19, 2026)

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Miyagi Prefecture business hotels, N=62 properties / 9,214 rooms, based on listings on major booking sites; estimated OCC based on OTA-listed inventory; as of October 19, 2026. “—” = not yet observed, as the stay date is more than 30 days away)

To sum up, three points are confirmed for the Miyagi Prefecture business hotel market: (1) the day-of-week shape — peaking on Thursday with a Sunday trough — was stable across all three months; (2) a holiday’s final night sinks as low as or lower than a Sunday, regardless of the day of week; and (3) that gap is already open 45 days before the stay date. Put differently, Sunday and holiday-final-night inventory is “predictable leftover” — not something to be dumped in a panic with last-minute discounts, but nights that can be assigned a role in advance within the month’s sales design.

Comparison with the Six Tohoku Prefectures and Major Metro Areas — Miyagi’s Sunday Trough Is “Mid-Depth”

Is Miyagi’s Sunday trough deep or shallow compared with other markets? Lining up September 2026 business hotel data, aggregated on the same definition, for the six Tohoku prefectures and major metro areas, the gap between Thursday and Sunday excluding holidays was 12.4 points in Miyagi — the second smallest in Tohoku after Aomori (8.0 points). Fukushima (18.8 points), Yamagata (19.3 points), Iwate (20.8 points) and Akita (24.5 points) all exceeded 18 points, showing a sharper weekday-to-Sunday drop than Miyagi.

Among major metro areas, by contrast, the gaps in Fukuoka (3.8 points), Tokyo (4.8 points) and Aichi (7.7 points) were small, Hokkaido (10.5 points) was close to Miyagi, and Hiroshima (20.4 points) was as wide as the rural Tohoku prefectures. Miyagi sits between the “big-city type, where Sundays also fill with tourism and leisure stays” and the “regional type, relying solely on weekday business,” and because Sunday demand is not zero, there remains room to capture it through multi-night and pre/post-stay design. Note that Akita (N=22–23 properties) and Yamagata (N=21–25) have small samples, so a single property’s movement can swing the day-of-week averages.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (fixed cohort of business hotels in each prefecture; estimated OCC based on OTA-listed inventory; September 1–30, 2026. “Ex-holiday” is the average of the three Sundays excluding Silver Week, September 19–23. Gaps are in points, calculated from the displayed values)

Sensitivity Analysis — How Monthly Occupancy Moves When the Four Trough Nights Are Filled

Defining September’s “trough nights” as the three Sundays excluding the holiday (September 6 at 83.9%, the 13th at 85.8% and the 27th at 78.8%) plus the holiday’s final night on September 23 (77.2%), their average is 81.4%, roughly 10 points below the monthly average of 91.6%. Table 4 shows, arithmetically, the effect of uniformly lifting only these four nights. Because they are just four of the month’s 30 nights, filling the trough by 6 points adds only +0.8 points to the monthly average, and even filling it by 9 points to lift the trough average to 90.4% adds just +1.2 points.

Small as that looks on a monthly-average basis, it becomes a meaningful volume at the property level. For a 150-room property — close to the cohort’s average size — filling the four trough nights by 6 points equals an extra 36 room-nights per month, and 9 points equals 54 room-nights. In a market where weekdays are already capped in the 90% range, much of the remaining upside is concentrated in these four nights, so the multi-night design described above and redefining “Sunday-treatment nights” can be positioned as levers to capture these few dozen room-nights. As this article does not address day-of-week pricing, no conversion to revenue is made here.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Miyagi Prefecture business hotels, N=58–62 properties; estimated OCC for September 2026 based on OTA-listed inventory. Four trough nights = three Sundays excluding the holiday [9/6, 9/13, 9/27] + holiday final night [9/23]. Contribution to monthly average = improvement × 4 nights ÷ 30 days; property-level figures are additional rooms sold across the 4 nights in one month = rooms × improvement × 4)

For Revenue Managers of Business Hotels in Miyagi — Implications and Action Plan

All figures in the action plan are market values presented in this article (Source: MetroEngines Research; compiled by HotelBank Editorial Team). This is a framework for consideration, premised on comparison with your own property’s results, and does not guarantee any outcome.

Summary — Three Yardsticks for Measuring the Day-of-Week Trough

Miyagi Prefecture’s business hotel market combines monthly occupancy above 90% with an estimated settled ADR up 17.3% year on year (finalized September figure). Within it, the remaining challenges come down to the day-of-week step between Thursday at 95.2% and Sunday at 86.4% (82.8% excluding holidays), and the dip on holiday final nights. Three reusable yardsticks can help measure your own property’s position.

First, the day-of-week step — is the gap between your Thursday and Sunday larger or smaller than the market’s 8.8 points (10.5 points versus Wednesday, excluding holidays)? Second, the holiday final-night drop — does your calendar already account for the market pattern in which a holiday’s final night sinks 16–17 points from the night before (September 23 at 77.2%, October 12 at 78.4%)? Third, the T-45 and T-30 checkpoints — where does your pace sit against the market bands of Sundays in the 60% range at T-45 and around 70% at T-30, and Thursdays around 70% at T-45 and around 80% at T-30? Sunday inventory does not appear suddenly; it is visible a month and a half ahead. A trough you can see is a trough you can still act on while you see it.

About the Data

Source: MetroEngines Research; compiled by HotelBank Editorial Team

■ Data Sources

MetroEngines Research OTA-listed inventory and sales status data: a fixed cohort of Miyagi Prefecture business hotels (monthly estimated OCC: August N=56–58 properties, September N=58–62, October N=56–62), booking curves by stay date (N=62 properties / 9,214 rooms), and finalized monthly estimated settled ADR (N=140–142 properties). Table 3 covers business hotels in each prefecture aggregated on the same definition (September 2026).

■ Estimation Assumptions

Table 4 is arithmetic for uniformly lifting estimated OCC on September 2026’s four trough nights (three Sundays excluding the holiday plus the holiday’s final night on September 23; average 81.4%). Contribution to monthly average = improvement × 4 nights ÷ 30 days; additional rooms sold per property = rooms × improvement × 4 nights. In line with this article’s policy of not addressing day-of-week pricing, no conversion to revenue is made. Gaps between days of week and between prefectures are differences between displayed values (one decimal place).

■ Limitations and Caveats

Estimated OCC is an estimate based on OTA-listed inventory and reads higher than actual room occupancy because unlisted rooms are counted as sold. October covers the 1st–19th, and dates more than 30 days away are mid-sales observations. Day-of-week averages are simple averages of 2–5 days per weekday; in prefectures with only around 20 properties (such as Akita and Yamagata), a single property’s movement can swing the figures. These figures indicate market trends and do not guarantee results for any individual property.