When a department store or GMS reaches the end of its useful life, the site it leaves behind is one of the rarest development opportunities a Japanese city ever sees: a large, consolidated parcel right in front of the station. Land that held tens of thousands of square meters under a single retail use almost never returns to the market in one piece in a mature city center. This report tracks 14 large retail properties that closed from 2020 onward and whose redevelopment plans were disclosed and verifiable against primary sources, and quantifies how often a hotel use is chosen for those sites — and what the locations that do choose hotels have in common. It also back-solves, location by location, the ADR level a standalone 300-room hotel would need on a former retail site, putting numbers on the conditions for feasibility.

Share of former department store sites that choose a hotel — and what takes their place when they do not

Start with how the sample was built. Among department stores and large retail complexes that closed from 2020 onward, we included only those whose site reuse has been disclosed and could be verified against primary sources — operator press releases, official documents from redevelopment associations, or municipal publications. Projects with undisclosed reuse, and plans reported only in the press without corroboration, were excluded. This is therefore not an exhaustive census: because the search prioritized projects with a hotel use, the hotel share reads higher here than in the true population, which we flag up front.

On that basis, the 14 sites break down as follows. Six (42.9%) include a hotel use, but only one of those is a standalone hotel; in the other five the hotel is embedded in a mixed-use development alongside housing, offices or retail. Hotels are not the single destination for these sites — three are being renewed as retail and two go to residential, so uses are genuinely dispersed.

Uses diverge not because of the character of the site itself but because of the earning power required to carry the floor area. All six projects with a hotel sit in commercial zones within walking distance of a station, and they cluster on parcels where land prices are clearly rising: published land prices are up 20.0% YoY in Susukino, 10.7% at Hiroshima Station South Exit and 13.5% in Shibuya Dogenzaka. Where hotels dropped out of the plan, land price growth is comparatively muted.

What happens when the hotel drops out of the plan — the Fukushima Station East Exit record

Whether a hotel goes onto a site is often decided partway through the planning process. The Fukushima Station East Exit Type 1 urban redevelopment project, which includes the former Nakago Fukushima site, originally intended to incorporate a hotel in the retail block immediately adjacent to the station. In May 2024, at a plenary briefing of the city assembly, Fukushima City announced that it had abandoned the hotel recruitment. The reason given was that rising construction costs and general inflation had made operators cautious about new investment, and rent terms could not be agreed. The project outline the redevelopment association currently publishes describes a district area of roughly 2.0 ha and roughly 73,000 sqm of floor area with principal uses of “retail, offices, housing, public facilities, parking and so on” — no hotel. Completion and opening are scheduled for FY2028-2029.

The sequence carries a clear lesson for anyone assessing a former retail site. The presence of a large parcel is not by itself a sufficient condition for a hotel use; the real dividing line is whether the ADR achievable at that location balances against a total investment that now embeds higher construction costs. How many projects stall for the same reason between 2027 and 2029 is traced nationwide in Construction +41% & Labor Crunch Erase Japan’s 2027-2029 Hotel Supply. Below we put numbers on that balance.

The former-retail development pipeline in planning data

MLIT’s Building Activity Statistics Survey identifies 68 planned or under-construction hotels totaling 15,115 rooms as of the survey date. Of those, 20 are mixed-use projects listing retail, commercial, office or residential among their principal uses, and 5 have a joint-development entity — an urban redevelopment association, a superior-building development association and the like — among the project implementers. The count is small, but that does not mean former-retail development is rare: because building confirmation applications are typically filed one to two years before opening, counts for future years are structurally undercounted.

Of these five, only Ikegamicho in Utsunomiya is classified as a “rebuild”; three of the remaining four are new builds, and the Kumoidori project in Kobe has no registered work type. The configuration of 300 rooms in 13,000 sqm of floor area — that is, 43.3 sqm of gross floor area per room — serves as the observed benchmark for the back-solve below. Note that “rebuild” in the Building Activity Statistics refers to demolishing a previous building and erecting an equivalent one; it is a different concept from a conversion that reuses the existing structural frame. In fact, within this sample the only renewal that reused an existing frame is the retail refurbishment of the former Sogo Kawaguchi, and no hotel conversion reusing an existing frame was identified within the scope tracked here.

Where the tracked sites are — which station-front parcels are coming back to market

The catchment as a surface — the resident-population mesh around Ikegamicho, Utsunomiya

Former retail sites are casually lumped together as “prime station-front land,” but the demand structure around them differs sharply from project to project. Within a 1.5 km radius of Ikegamicho in Utsunomiya, the resident population is estimated at 35,117 for 2025 and is projected to fall 13.2% by 2040, with a senior share of 31.3%. At the same time, the 1 km radius holds 29,900 employees with an office-sector share of 23.3% — the prefectural capital retains real depth as a business center. A catchment whose residential base is slowly contracting while business demand persists is consistent with a design that anchors accommodation demand in business and wide-area visitation rather than leisure.

Back-solving feasibility — what ADR does a 300-room hotel on a former retail site require?

From here we back-solve the ADR required if an identically sized hotel were placed on each site. This is not any project’s business plan; it is a hypothetical designed to isolate the effect of location alone.

The results show clearly how location conditions bite. Ranked by how many times the market median the required ADR is, Sapporo Susukino is lowest at 1.37x, followed by Shibuya Dogenzaka at 1.46x, Fukushima Station East Exit at 1.77x, Utsunomiya Ikegamicho at 2.08x and Okayama Nodayacho at 2.20x, with Fukuoka Tenjin at 2.94x and Hiroshima Station South Exit at 3.11x highest. Susukino and Shibuya come out low because an FAR of 800% lets the same floor area sit on a smaller site, and because the estimated transacted ADR in those markets is already high. And indeed both of these sites have hotels planned or built: 436 rooms in Susukino, and a hotel alongside housing and retail in Shibuya. The same back-solve applied to general development land — not limited to former retail sites — across eight major markets is set out in Policy Rate 1.00%: Break-Even ADR Across Japan’s 8 Hotel Markets.

More important is land cost as a share of total investment. It is just 5% in Utsunomiya and 6% in Fukushima, but reaches 53% at Hiroshima Station South Exit, 59% in Shibuya and 70% at Tenjin in Fukuoka. Where land accounts for more than half of total investment, 300 hotel rooms alone cannot carry the land. That is precisely why, at this price level, former retail sites share their floor area with offices, retail and housing, and the hotel is planned as one component of a mixed-use scheme occupying part of the upper floors. Of the six projects with a hotel use, the only standalone hotel is the one in Utsunomiya, where land is cheapest — the actual choice of use matches what the model implies.

Sensitivity — how the required ADR moves when the assumptions move

The base case assumes OCC 80% (a modeling assumption as of August 2026, not an observed occupancy rate), a GOP margin of 35% and a 5.0% target NOI yield, giving a required ADR of ¥15,171. If the GOP margin improves to 40% it falls to ¥13,275, and combined with OCC 85% (again an assumption) it drops to ¥12,494. Conversely, at a 30% GOP margin and OCC 75% (same basis) it takes ¥18,880. The spread reaches roughly ¥6,386, which shows that designing for operating efficiency governs feasibility as much as — or more than — site selection. It is also why pairing with an operator that can hold down operating costs in a limited-service format keeps recurring in former-retail hotel projects.

The sensitivity above is confined to the single Utsunomiya location. Applying the same three sets of operating assumptions to all seven locations shows that while the ranking between locations does not move, the sheer difficulty of making a project work changes substantially with the assumptions.

On a multiple basis, the optimistic case brings Sapporo Susukino down to 1.01x and Shibuya Dogenzaka to 1.08x — essentially level with the market median, meaning that if operations run to plan the project works as an extension of standard price points. Under the pessimistic case, however, even the most favorable location, Sapporo, needs 1.87x, while Hiroshima Station South Exit jumps to 4.25x and Fukuoka Tenjin to 4.02x. The relative merits of locations do not reorder when the assumptions move, but the feasibility line moves by nearly a factor of two. Before deciding whether to acquire a site, you need to settle first on how adverse an operating case you can absorb.

ADR trends in the markets these sites sit in — can they absorb the assumed room count?

Whether a market can carry the required ADR also depends on whether its rates are rising. Overlaying monthly estimated transacted ADR for three of the site markets, all three show clear peaks in summer and around the New Year, and the past year sits above the prior-year level in the same months. The trailing 13-month median is ¥7,300 for Utsunomiya City (N=45 properties), ¥7,818 for Kita-ku Okayama (N=49) and ¥14,704 for Chuo-ku Sapporo (N=136).

What matters here is that the required ADR is being compared against the market median. The median reflects the level of a typical property in that city, and a newly built 300-room hotel would naturally aim above it. Utsunomiya’s required ADR of ¥15,171 is 2.08x the market median, but that means “within reach if the product can consistently capture the upper price band of the market” — not that the project is infeasible. Large station-front parcels are among the few development opportunities with the locational qualities needed to reach for that upper band.

What makes a site likely to become a hotel — five conditions

Conclusion

Tracking 14 large retail properties that closed from 2020 onward against primary sources, six include a hotel use and only one is a standalone hotel. These sites are absorbed by retail, housing and offices alike; a hotel is not the automatic answer. The dividing line is not the size of the parcel but the balance between the ADR achievable at that location and a total investment that now embeds relentlessly rising construction costs.

Back-solving with the common assumption of 300 rooms in 13,000 sqm across seven locations, the required ADR ranges from 1.37x to 3.11x the market median. Where land accounts for less than 30% of total investment, a standalone hotel is chosen; above 50%, the hotel appears as one component of a mixed-use scheme — and the actual choices of use line up with that structure. Cases like Fukushima Station East Exit, where the hotel drops out midway through planning, are explained by the same mechanism: rent terms could not keep pace with construction cost inflation.

Rising construction costs also suppress new supply and thereby raise the scarcity value of existing stock. Large consolidated parcels do not come to market in front of stations very often. Before deciding the use, pin down in numbers the price band achievable at that location and the land cost ratio — that is where a former-retail hotel project starts.