At hotels in Chuo Ward, Sapporo, the Labor Thanksgiving Day long weekend (Saturday, November 21 to Monday, November 23, 2026, a national holiday) keeps showing the same pattern: the first night fills early and the second night does not follow. On the booking curve as of November 3 (estimated OCC, based on OTA-listed inventory), the first night, November 21 (Sat), had reached 90.6% for all hotel types (N=118 properties) 19 days before the stay date—already above the latest reading for November 7 (Sat), an ordinary Saturday, of 84.5% (5 days out). By contrast, the second night, November 22 (Sun), stood at 80.2% 20 days out, almost exactly where November 7 (Sat) was when it passed 20 days out (79.4%). In other words, the long weekend’s second night is not yet showing a holiday-style lift; it is simply tracing an ordinary Saturday’s curve. This article reads that gap by hotel type through a 3-point frame—45 days out, 30 days out and last minute—and sets out how to relax length-of-stay and minimum-night rules for the second night and what to base last-minute pricing decisions on.

Coverage: hotels in Chuo Ward, Sapporo (all hotel types N=117-118 properties / city hotels N=14 / business hotels N=73). Occupancy in this article is an estimate based on OTA-listed inventory (estimated OCC); the price metric is estimated settled ADR (the settled price level estimated from OTA and other sales data, tax-exclusive equivalent). Both are defined at the end of the article. Data as of November 3, 2026 (latest booking-curve observation: November 2).

Night 1 above 90% 19 days out, night 2 at ordinary-Saturday levels — the picture from the all-hotel-type curve

First, we overlay the curves for all hotel types (N=117-118 properties, 17,626-17,669 rooms, based on listings on major booking sites) for ordinary Saturdays and each night of the long weekend. Night 1, November 21 (Sat), added +8.5pt from 78.3% at 45 days out to 86.8% at 30 days out, and reached 90.6% at the latest reading, 19 days out. At the same 45-days-out point, the ordinary Saturdays of November 7 and November 14 were at 68.3% and 70.8%, so night 1 started about 10pt higher, and its 45→30-day gain of +8.5pt also exceeded the ordinary Saturdays’ +6.7pt (Nov 7) and +6.7pt (Nov 14). Night 1 was high early and kept gaining strongly—the typical shape of a holiday-weekend date.

Night 2, November 22 (Sun), meanwhile, moved from 67.5% at 45 days out to 75.6% at 30 days out and 80.2% at 20 days out. Its 45→30-day gain of +8.1pt is close to night 1’s, so the pace of pickup itself is not weak. The problem is the low starting point: 67.5% at 45 days out was below the 68.3% of ordinary Saturday November 7, and 80.2% at 20 days out was in line with the 79.4% November 7 passed at 20 days out and the 81.7% of November 14. High for a Sunday, but as the second night of a long weekend it is simply tracing the same curve as an ordinary Saturday, and the gap to night 1 had widened to about 10pt at 20 days out.

The final night, November 23 (Mon, holiday), is clearly the lowest of the three, moving from 56.0% at 45 days out to 61.3% at 30 days out and 64.7% at 21 days out. It is common for the last night of a long weekend, followed by a weekday, to lag, but in Chuo Ward the gap between night 1 and the final night reached about 26pt around 20 days out, suggesting that one-night stays—"stay on the first night, go home on the second day"—make up the majority. This pattern, in which only the holiday itself fails to fill, was also seen across Hokkaido over the October long weekend; Hokkaido October Long Weekend: Sat 88.3% vs Holiday Monday 69.6% OCC breaks it down at 45 days out, 30 days out and the latest reading. For reference, November 28 (Sat), the Saturday after the long weekend, started relatively strongly for an ordinary Saturday at 78.2% 30 days out and 79.7% 26 days out.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Figure 1: Estimated OCC (based on OTA-listed inventory) booking curves for Chuo Ward, Sapporo, all hotel types (N=117-118 properties). The horizontal axis shows days remaining until the stay date (45 days out → latest). Based on observations from 45 days before the stay date to the latest reading (observed November 2).

Source: MetroEngines Research; compiled by HotelBank Editorial Team

City hotels: night 1 at 96%, night 2 at 80%; business hotels: night 1 at 89%, night 2 at 80% — the "night 1 fills first" pattern by hotel type

Split by hotel type, the gap between night 1 and night 2 shows up differently. For city hotels (N=14 properties / 4,118 rooms), November 21 (Sat) was already at 90.3% at 45 days out—above the latest reading for ordinary Saturday November 7 of 87.0% (5 days out). From there it reached 95.1% at 30 days out and 96.3% at 19 days out; the gains shrank to +4.8pt and +1.2pt, but this is likely because there is little sellable inventory left rather than because demand stopped. Night 2, November 22 (Sun), by contrast, went from 74.2% at 45 days out to 79.1% at 30 days out and 80.5% at 20 days out, about 16pt behind night 1 at 20 days out. City hotels’ night 1 is "nearly full," while night 2, at 80.5% 20 days out, is about 6pt below the 86.2% ordinary Saturday November 7 showed at 20 days out.

Business hotels (N=73 properties / 12,045 rooms) reach a lower level than city hotels on night 1 but gain more along the way. November 21 (Sat) went from 74.5% at 45 days out to 84.3% at 30 days out and 88.9% at 19 days out; the 45→30-day gain of +9.8pt was the largest of the six dates compared. Night 2, November 22 (Sun), also added +9.1pt from 45 to 30 days out (65.5% → 74.6%) and stood at 80.2% at 20 days out—a faster pace than city hotels’ night 2 (+4.9pt). Business hotels’ night 2 is now almost level with city hotels’ night 2 at the latest reading (80.2% vs 80.5%), but business hotels may still have last-minute pickup to come. For ordinary Saturday November 7, business hotels added +6.5pt from 77.7% at 20 days out to 84.2% at 5 days out, whereas city hotels moved only from 86.2% to 87.0% (+0.8pt) over the same window; the inventory that moves at the last minute is mainly business-hotel inventory.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Figure 2: 3-point comparison of estimated OCC (based on OTA-listed inventory) by hotel type. City hotels N=14 properties, business hotels N=73 properties. The latest reading is 19 days out for the Nov 21 stay and 20 days out for the Nov 22 stay (observed November 2).

Source: MetroEngines Research; compiled by HotelBank Editorial Team

To sum up, Chuo Ward’s long weekend has a three-layer structure: "city hotels’ night 1 fills first, business hotels’ night 1 follows, and night 2 stays on a curve at or below an ordinary Saturday." The night-1-fills-first pattern is not unique to Sapporo: our analysis of the same long weekend, Kyoto Nov 21-23 Long Weekend: Business Hotels 98.3%, 75.1% Sold Out, shows night 1 (Nov 21) reaching 98.3% for business hotels 30 days out while the middle and final nights lagged. With city hotels’ night 1 in the mid-90s, demand that cannot get a room on night 1 is now spilling over to business hotels and on to neighboring wards, which means business hotels in Chuo Ward can command rate on their remaining night-1 inventory. Night 2, conversely, has to be filled either by persuading night-1 guests to stay "one more night" or by attracting more guests who book night 2 on its own.

October’s Sports Day long weekend landed at 94.1% → 87.8% → 77.9%, September’s 5-day holiday peaked mid-run — how the second night landed in past long weekends

As a guide to how far the second night can rise, we look at how the two most recent long weekends landed (actuals just before the stay date, estimated OCC based on OTA-listed inventory). Over October’s Sports Day long weekend, night 1 on October 10 (Sat) landed at 94.1%, night 2 on October 11 (Sun) at 87.8%, and the final night, October 12 (Mon, holiday), at 77.9% (all hotel types, N=110-117 properties; 112 as of the first of the month). The middle Sunday was 6.3pt below night 1 and the final night a further 9.9pt lower—a staircase with the same "night 1 > night 2 > final night" ranking that the November long weekend shows at 20 days out.

The shape changes with the length of the holiday, however. Over September’s 5-day holiday (Saturday, September 19 to Wednesday, September 23, a national holiday), night 1 on September 19 (Sat) landed at 91.6%, day 2 on September 20 (Sun) at 95.6%, day 3 on September 21 (Mon, holiday) at 94.9%, day 4 on September 22 (Tue, holiday) at 88.6% and the final night, September 23 (Wed, holiday), at 81.0%—the two middle nights exceeded night 1. Longer holidays bring more multi-night stays that use Sapporo as a base for touring, so nights 2 and 3 build up after night 1. Over a 3-day weekend this mid-run lift is weak, and night 2 tends to land not as "a Sunday boosted by the long weekend" but as "a Sunday slightly higher than usual."

October’s day-of-week actuals confirm this. Excluding the long weekend, estimated OCC for all hotel types on October’s four Saturdays (10/3, 10/17, 10/24, 10/31) averaged 92.7%, and on the three Sundays excluding the long-weekend middle night (10/4, 10/18, 10/25) it averaged 84.7%—a Saturday-Sunday gap of about 8pt. The day-of-week structure in which Sundays and Mondays are the troughs for Hokkaido business hotels can also be seen in the day-of-week estimated OCC in Hokkaido Business Hotels Aren’t Weekend-Led. The long-weekend middle night, October 11 (Sun), at 87.8%, was 3.1pt above an ordinary Sunday but did not reach the ordinary-Saturday average. Given that November 22 (Sun) is at 80.2% 20 days out—"the same as an ordinary Saturday at 20 days out"—a scenario in which it lands, if nothing is done, in the high 80s like October’s middle night, or at worst near ordinary-Sunday levels, cannot be ruled out. How to sell the second night will be decided over the roughly three weeks from here to the last minute.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Figure 3: Daily estimated OCC (based on OTA-listed inventory, observed just before each stay date) for October 2026, Chuo Ward, Sapporo, all hotel types (N=110-117 properties). Monthly average: 90.5%.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Indicative landings are October 2026 observations just before each stay date (all hotel types, N=110-117 properties); 80.2% at 20 days out is from the November fixed cohort (N=118 properties). The property sets do not fully match, so this is a reference comparison, not a forecast. For comparison, ordinary Saturday Nov 7 (Sat) gained +5.1pt from 79.4% at 20 days out to 84.5% at 5 days out; a gain of the same size would leave night 2 near the downside scenario.

Estimated settled ADR bottoms out for the second half of the year in November — the yardstick for last-minute pricing

We also look at the pricing context. Estimated settled ADR in Chuo Ward, Sapporo (tax-exclusive equivalent, area median) for city hotels (N=25-26 properties) was a finalized ¥16,000 in November 2025 and a finalized ¥18,000 in October 2026. For business hotels (N=100-105 properties) it was ¥12,000 in November 2025 and ¥14,000 in October 2026. Year on year between finalized months, October 2026 was +1.0% for city hotels and +14.7% for business hotels, so the rise in business-hotel rates stands out. November falls in the gap after the autumn leaves and before the Snow Festival; in the 2025 finalized figures, too, city hotels fell from ¥17,800 to ¥16,000 and business hotels from ¥12,200 to ¥12,000 from October to November, making it the month that enters the trough for the second half of the year (July-December).

The current estimate for November 2026 (based on current sales and subject to change) is ¥13,400 for city hotels (N=26) and ¥10,600 for business hotels (N=105). A straight comparison with last year’s finalized figures will have to wait for month-end, but when selling the long weekend’s second night, there is good reason to be cautious about keeping the Sunday rate above the month’s overall level just "because it’s a long weekend," given that the curve is tracking an ordinary Saturday. Night 1, November 21 (Sat), on the other hand, is in the mid-90s for city hotels and at 90% for all hotel types, so there is no reason to cut it.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Finalized = finalized value for a past month. Current estimate = estimate based on current sales for the current or a future month, which may change before the month is finalized. Year-on-year comparisons are only between finalized months (October 2026 / October 2025).

For revenue managers running city and business hotels in Chuo Ward, Sapporo — implications and action plan

Summary — a yardstick for reading a long weekend through the 3-point frame

Viewed at 45 days out, 30 days out and the last minute, the Labor Thanksgiving long weekend in Chuo Ward, Sapporo comes down to three points. First, night 1, November 21 (Sat), went from 78.3% at 45 days out to 86.8% at 30 days out and 90.6% at 19 days out for all hotel types, already above the latest readings for ordinary Saturdays, and city hotels are almost full at 96.3%. Second, night 2, November 22 (Sun), at 80.2% 20 days out, is simply tracing almost the same curve as an ordinary Saturday at the same point, about 10pt behind night 1. Third, in past long weekends the middle night of a 3-day weekend landed 6.3pt below night 1, while the middle nights of a 5-day holiday exceeded night 1, so the second night of a 3-day weekend will not rise like a holiday night on its own. The steps for selling the second night can be ordered as: remove multi-night restrictions anchored on night 1 → offer second-half (Sun-Mon) multi-night products → at the last minute, set prices by the gap to October’s Sunday landings.

About the Data

Source: MetroEngines Research; compiled by HotelBank Editorial Team