Gifu Ryokan Foliage Weekends: Oct 17 Rose From 84.8% at T-45 to 96.2%

Posted: 2026.10.19

Line up the booking curves for Gifu Prefecture ryokan (the whole prefecture, including Hida, Okuhida and Gero) for the three Saturdays that open the autumn-foliage season — October 10, 17 and 24 — from 45 days before the stay date to the last day, and a clear gap appears. October 10, when the autumn Takayama Festival coincided with a three-day weekend, had already reached an estimated OCC (OTA-listed inventory basis) of 92.2% at 45 days out, while the following Saturday, October 17, started selling from 84.8%, 7.4pt lower. October 17 narrowed the gap to 89.3% at 30 days out and 92.0% at 14 days out and reached 96.2% the day before the stay, but it never caught up with October 10’s 99.8%, finishing 3.6pt short. This article works backward through three checkpoints — 45, 30 and 14 days out — to show how far the “trough week” sandwiched between three-day weekends fell behind at each point and where it caught up, and sets out the sales-channel, plan and minimum-length-of-stay moves for the same week in future years.

Scope: Gifu Prefecture ryokan, N=132–136 properties with 3,442–3,497 rooms (a fixed cohort per stay date, based on listings on major booking sites). Occupancy in this article is an estimate based on OTA-listed inventory; price levels are not covered. Definitions are at the end of the article. Data as of October 18, 2026.

Key Takeaways

— −7.4pt at 45 days out — Gifu ryokan on Saturday, October 17 stood at an estimated OCC of 84.8% at 45 days out (as of September 2, 2026), below October 10 (92.2%), the first day of the three-day weekend and the Takayama Festival, and below October 24 (87.8%): sales started from a “trough.”
— Most of the gap closed by 30 days out — The gap with October 10 was −4.9pt at 30 days out, −4.7pt at 14 days out and −3.6pt the day before (96.2% vs 99.8%).
— Sold-out property rate 72.1% the day before — A 22.6pt gap remained against October 10’s 94.7%; the prefecture as a whole sold well, yet about three in ten inns still had inventory left.
— No Saturday premium — October 17 (96.2%) was level with Thursday (96.5%) and Friday (96.3%) of the same week. October 3 had formed a weekend peak +10.8pt above Thursday.
— A three-stage playbook — Identify the trough at 45 days out by comparing it side by side with the Saturdays before and after, set up plans and multi-night conditions by 30 days out, and sell through the inventory that comes back between 18 and 10 days out at the last minute.

Booking curves for three Saturdays — only October 17 started from a low position

First, look at the full curves for the three dates. Saturday, October 10 was the first day of a three-day weekend and also a day of the autumn Takayama Festival (Hachiman Festival) in Takayama City, held Friday, October 9 to Saturday, October 10 (Gifu Prefecture official tourism website). With the double demand of the festival and the holiday weekend, this date already stood at an estimated OCC of 92.2% at 45 days out (as of August 26, 2026) — above 90%. The backward-planning calendar for the first day of the three-day weekend itself is laid out in Gifu Ryokan Oct Long Weekend: 92.1% Est. OCC at T-45, Weekend 79.1%.

The following Saturday, October 17, by contrast, stood at only 84.8% at 45 days out, a 7.4pt gap with October 10. One week later, Saturday, October 24 was at 87.8%, so lining up the three dates puts October 17 alone in a lower “trough.” October 24, waiting for the foliage to peak, and October 10, with the festival and the holiday weekend, bracketed October 17, which at 45 days out does not appear to have been recognized as a destination for demand. As a follow-up to Gifu Ryokan Oct 17: 93.3% Est. OCC, 65.4% of Inns Sold Out at T-16, which captured this week in a cross-section 16 days before the stay date, this article tracks the curve all the way to the day-before landing.

Source: MetroEngines Research, compiled by HotelBank Editorial Team

The trough did not, however, stay equally deep to the end. October 17’s curve built steadily — +4.5pt between 45 and 30 days out and +2.7pt between 30 and 14 days out — and by 14 days out the gap with October 10 had narrowed to 4.7pt. October 24 rose to 93.4% at 30 days out, then flattened to 93.3% at 14 days out, and the latest observation (7 days out) was 92.5%. In other words, from 30 days out the two weeks diverged: October 17 was a “week that kept climbing,” October 24 a “week that built early and then stalled.”

Three checkpoints at 45, 30 and 14 days out — the gap narrowed until 30 days out

The figures at the three checkpoints are gathered in a single table. What matters here is less the level of estimated OCC itself than the interval in which the gap between dates narrowed.

Table 1: Estimated OCC on three Saturdays (Gifu Prefecture ryokan, 45 / 30 / 14 days out / latest, as of October 18, 2026)
Stay date	45 days out	30 days out	14 days out	Latest	45 days out → latest	Coverage
Sat, Oct 10 — first day of three-day weekend, Takayama Festival	92.2%	94.2%	96.7%	99.8% (day before)	+7.6pt	132 properties, 3,442 rooms
Sat, Oct 17	84.8%	89.3%	92.0%	96.2% (day before)	+11.4pt	136 properties, 3,491 rooms
Sat, Oct 24	87.8%	93.4%	93.3%	92.5% (7 days out)	+4.7pt	136 properties, 3,497 rooms
Gap vs Oct 10 (Oct 17)	−7.4pt	−4.9pt	−4.7pt	−3.6pt	—	—

Estimated OCC (OTA-listed inventory basis). October 24 is shown up to the latest observation as of October 18, 2026 (7 days before the stay date).

Source: MetroEngines Research, compiled by HotelBank Editorial Team

The gap between October 17 and October 10 narrowed by 2.5pt, from 7.4pt at 45 days out to 4.9pt at 30 days out. Between 30 and 14 days out, however, it narrowed only 0.2pt, from 4.9pt to 4.7pt, and the next real narrowing came in the final two weeks. Put differently, the filling of the trough week came in two waves: “up to 30 days out” and “the final week.”

The final week is worth tracking in detail. October 17’s estimated OCC rose +4.2pt, from 92.0% at 7 days out to 96.2% the day before, outpacing October 10 over the same interval (96.0% → 99.8%, +3.8pt). Measured by the last-minute gain, the trough week grew more.Before that, however, between 18 and 10 days out, October 17’s curve slipped from 93.2% to 91.7%. Listed inventory that came back to the market in this period was then sold out at the last minute, which suggests that how inns handled inventory returned through cancellations and released allotments shaped the trough week’s landing.

By sold-out property rate — the trough week had many properties that could not sell out

When estimated OCC is above 90%, “the share of properties with no listed inventory visible on OTAs and similar channels (sold-out property rate, estimated)” reflects differences between properties better than the average does. Across the three checkpoints, October 10 went 62.3% at 45 days out → 75.0% at 30 days out → 80.8% at 14 days out, with most properties nearing sell-out early, whereas October 17 went 42.2% → 48.9% → 61.2%, still below half at 30 days out.

Source: MetroEngines Research, compiled by HotelBank Editorial Team

October 17’s sold-out property rate rose to 72.1% the day before, but a 22.6pt gap remained against October 10 on its own day before (94.7%). In contrast to the estimated-OCC gap, which narrowed to 3.6pt, this shows that even as the prefecture as a whole sold well, about three in ten inns still had inventory left at the end. October 24 stood at 66.7% at 30 days out, 17.8pt above October 17 (48.9%): the closer a week was to peak foliage, the more properties sold out early.

For reference, the Saturdays before and after were lined up at the same three checkpoints. October 3 started from an even lower position than October 17, 79.3% at 45 days out, and rose to 96.6% the day before. October 31 and November 7 both stood at 91.8% at 30 days out, 2.5pt above October 17 at the same point (89.3%).

Table 2: Three-checkpoint comparison including the Saturdays before and after (stays October 3 – November 7, 2026)
Stay date	45 days out	30 days out	14 days out	Final (day before)	Coverage
Sat, Oct 3	79.3%	85.5%	89.3%	96.6%	126 properties, 3,371 rooms
Sat, Oct 17	84.8%	89.3%	92.0%	96.2%	136 properties, 3,491 rooms
Sat, Oct 31	87.9%	91.8%	93.6%	—	138 properties, 3,517 rooms
Sat, Nov 7	88.1%	91.8%	—	—	137 properties, 3,512 rooms

Estimated OCC (OTA-listed inventory basis). October 31 and November 7 are estimates based on sales status as of October 18, 2026 and may change.

Source: MetroEngines Research, compiled by HotelBank Editorial Team

From early October to early November, the Saturday level at 45 days out rose step by step: 79.3% → 84.8% → 87.8% → 87.9% → 88.1% (excluding the October 10 three-day weekend). The way the foliage season moves down from high-altitude Hida and Okuhida appears to show up directly in how bookings come in. October 17 was a week partway up this seasonal climb — and one whose demand had been pulled forward by the three-day weekend just before it.

Daily estimated OCC in October — in the trough week, Saturday could not break away from weekdays

To check the trough week from another angle, daily estimated OCC for October 1–18 (observed the day before each stay date) is lined up below. Focusing on the height of the peaks by day of week makes October 17’s character clearer.

Source: MetroEngines Research, compiled by HotelBank Editorial Team

Saturday, October 3 (96.6%) formed a clear weekend peak, 10.8pt above the Thursday two days earlier (85.8%). Saturday, October 17 (96.2%), by contrast, stayed almost level with Thursday (October 15, 96.5%) and Friday (October 16, 96.3%) of the same week. Because the weekdays after the three-day weekend, October 13–16, were all high, in the 95–96% range, Saturday did not stand out.

What this suggests is that the trough week’s Saturday may have been “a day filled by the same guests as weekdays” rather than “a day of weak demand.” On foliage-season weekdays, demand from guests who can travel regardless of holidays comes in thick. The failure to build a weekend-style premium on Saturday likely showed up as the sluggish rise in the sold-out property rate.

For revenue managers running ryokan in Gifu Prefecture — implications and action plan

Implication 1: The trough week is visible at 45 days out. October 17 was 7.4pt below October 10 at 45 days out, and the gap persisted until 30 days out. For the weekend after a three-day weekend or a major festival, simply checking progress at 45 days out side by side with the weeks before and after is enough to identify a trough early. Overlay your own pace for the same week on the market curve and separate whether you are lagging along with the market or lagging on your own.

Implication 2: The run-up to 30 days out is the first decisive window. Across the market, the gap between October 17 and October 10 narrowed 2.5pt by 30 days out and barely moved until 14 days out. Adding plans or revisiting sales channels for the trough week fits the market’s movement better if it is all in place before 30 days out.

Implication 3: Inventory returning between 18 and 10 days out decides the landing. October 17’s curve dipped from 93.2% at 18 days out to 91.7% at 10 days out, then recovered to 96.2% the day before. Which channel rooms returned through cancellations are re-released to at the last minute was the point that decided whether an inn ended up on the sold-out side the day before.

Implication 4: Read the trough week’s Saturday as “an extension of the weekdays.” Saturday, October 17 was level with the surrounding weekdays. Imposing a weekend minimum length of stay on Saturday alone risks losing guests who travel mainly on weekdays. Review multi-night conditions in the trough week with a view to loosening them.

Turned into time-bound actions for the same week in future years (the weekend after a three-day weekend or festival), this becomes the plan below. Every decision criterion is a market figure shown in this article.

Table 3: Action plan for the same week in future years
Timing	Action	Trigger (figures from this article)	Aim
45 days out	Check progress for the Saturdays one, two and three weeks after the three-day weekend side by side	Your progress for the following Saturday lags well behind the first day of the three-day weekend (market: −7.4pt)	Identify the trough week early and secure lead time for measures
45 days out	Add 2+ night foliage-walk plans that include weekdays for the trough week	The trough is deeper than the Saturdays before and after (market at 45 days out: 84.8% vs 87.8%)	Connect guests with strong weekday demand to the days around Saturday
30 days out	Consider removing or loosening the minimum length of stay for the trough week’s Saturday	Sold-out property rate is still below 50% at 30 days out (market: 48.9%)	Limit lost bookings caused by multi-night conditions
30 days out	Review inventory allocation by sales channel and shift it toward the channels selling best	The gap with the first day of the three-day weekend is not narrowing (market: −4.9pt at 30 days out)	Capture the full build-up period through 30 days out
18–10 days out	Re-list rooms returned through cancellations immediately and decide in advance where returned inventory will be sold	Estimated OCC starts to fall (market: 93.2% at 18 days out → 91.7% at 10 days out)	Capture the last-minute gain (market: +4.2pt from 7 days out to the day before)
Last minute	Keep inventory on terms suited to last-minute bookings, such as one night with dinner and breakfast	Inventory remains up to the day before (market sold-out property rate the day before: 72.1%)	End up among the properties that sell out

All of these are guidelines built from market-wide movements; the best timing and moves will vary with an inn’s location (central Takayama, Okuhida Onsen-go, Gero Onsen and so on) and room types. There is room to test conditions one at a time against how your own property moved in the same week in past years.

Summary — three yardsticks for measuring the trough week

In Gifu Prefecture’s ryokan market in October 2026, October 10, when the autumn Takayama Festival overlapped with a three-day weekend, was above 90% from 45 days out, while the following Saturday, October 17, started from a trough of 84.8% at 45 days out and landed at 96.2% the day before. From this sequence, three yardsticks emerge that can be used in future years as well.

Table 4: Three yardsticks for measuring the trough week (results for October 17, 2026)
Yardstick	What to look at	Result for October 17, 2026
1. Side-by-side gap at 45 days out	Estimated OCC gap between the three-day weekend / festival weekend and the following weekend	−7.4pt (84.8% vs 92.2%)
2. Narrowing by 30 days out	How far the gap narrowed (it narrows little after 30 days out)	Narrowed to −4.9pt (by 2.5pt)
3. Sold-out property rate the day before	The share of properties that sold out, not prefecture-wide OCC	72.1% (October 10: 94.7%)

Source: MetroEngines Research, compiled by HotelBank Editorial Team

A trough week is not a week without demand but a week in which demand comes in differently. Approached in three stages — find the trough at 45 days out, set conditions by 30 days out, and capture the inventory that returns at the last minute — the weekend after a three-day weekend may be turned from “a week that doesn’t quite fill” into “a week filled by weekday demand.”

About the data

Table 5: Data definitions and scope
Definition of estimated OCC	OTA-listed inventory occupancy = 100 − 100 × OTA-listed rooms remaining ÷ total rooms. The cohort is fixed and consists of properties listing at least 30% of their total rooms on OTAs (the same set of properties for each stay date; property counts are shown in the “Coverage” column). It is an estimate based on how inventory sold on OTAs is being absorbed and differs in definition from actual room occupancy (unlisted rooms are counted as absorbed, so it reads higher).
Booking curve	Based on observations from 45 days before the stay date to the latest point. Checkpoints are 45, 30 and 14 days out and the final reading (the day before).
Sold-out property rate	The share of properties with no listed inventory visible on OTAs and similar channels (estimated).
Coverage	Gifu Prefecture ryokan. October 10 = 132 properties / 3,442 rooms; October 17 = 136 properties / 3,491 rooms; October 24 = 136 properties / 3,497 rooms; October 3 = 126 properties / 3,371 rooms; October 31 = 138 properties / 3,517 rooms; November 7 = 137 properties / 3,512 rooms (all fixed cohorts per stay date, based on listings on major booking sites). Daily estimated OCC covers 124–138 properties per day (October 1–18).
Data as of	October 18, 2026. Sales status and inventory change daily, so the figures in this article are a snapshot at the time of retrieval. Stay dates from October 24 onward are estimates based on current sales status.

References and sources

Gifu Prefecture official tourism website, “Gifu Travel Guide”

Autumn Takayama Festival (Hachiman Festival) (in Japanese)

■ Data source

Listed-inventory and availability data observed daily by MetroEngines Research on major booking sites (Gifu Prefecture ryokan, fixed cohort per stay date, N=126–138 properties, 3,371–3,517 rooms). Autumn Takayama Festival dates are from the Gifu Prefecture official tourism website.

■ Calculation assumptions

Estimated OCC = 100 − 100 × OTA-listed rooms remaining ÷ total rooms. Covers properties listing at least 30% of total rooms on OTAs, tracking the same set of properties for each stay date. Checkpoints are 45, 30, 14 and 7 days before the stay date and the day before. Sold-out property rate is the share of properties with no listed inventory visible. Daily estimated OCC for October is the value observed the day before each stay date (October 1–18, 2026).

■ Limitations

These are estimates based on OTA-listed inventory; unlisted rooms are counted as absorbed, so they read higher than actual room occupancy. Prices and settled rates are not covered. Stay dates from October 24 onward are interim readings as of October 18, 2026 and may change. Figures are prefecture-wide and do not separate regional differences such as central Takayama, Okuhida Onsen-go and Gero Onsen. “Returned inventory” is an interpretation read from increases and decreases in listed inventory; cancellation counts themselves are not observed.