Where Seoul's foreign visitors go — and where the growth is

A flexible-living room in Seoul sells a furnished stay of weeks to months to visitors, remote workers and tenants between leases.⁵ An underwriter needs to know three things about the visitor demand behind such a room: where it is concentrated, how it swings through the year, and how large a base its growth starts from. In the twelve months to July 2026 Seoul's 25 districts recorded 108.2 million foreign visitor-days, 24.0% more than the 87.3 million of the twelve months before.¹ The base had already recovered nationally: Korea's 2025 arrivals were 8.2% above the 2019 peak.³ Most of that volume is in five districts. The fastest growth lies outside them, and one district, Seongdong-gu, combines a sizeable base with fast growth to a degree no other district does.

The series counts only foreign visitors, measured in visitor-days and including day visitors.¹ The average foreign visitor to Korea stays 7.8 days,³ a stay of the kind sold by the night or the week rather than by the month. Tenants between leases and students, who take rooms for weeks or months, are covered elsewhere on this site. The series therefore captures the visitor side of flexible living. Of the four routes into the market that this site sets out, it bears on the two that sell stays to visitors.⁵ The flexible-living route converts ageing offices near metro stations into furnished units let for weeks or months; the lodging-conversion route repositions ageing lodging buildings under an operator and pools them through structured finance. Visitor demand matters for how capital enters the market. The 2025 tax and lending measures closed off buying flats to let and squeezed build-to-rent, so the routes still open run through conversion and operators,⁵ and the choice of building to convert depends on the demand it will serve.

Where it is deep, and what is already there

Jung-gu, Jongno, Mapo, Gangnam and Yongsan hold 63% of the city's foreign visitor-days, and Jung-gu alone almost a quarter.² Jung-gu leads every month, averaging 2.09 million visitor-days a month.² Three of the five grew more slowly than the city: Jung-gu by 21.2%, Mapo by 20.9%, Jongno by 23.5%.¹ Gangnam and Yongsan, both at 25.9%,¹ beat Seoul's pace by less than two points.² Together the five are holding their share of a larger market.

The operators are already there. Of the 131 institutional living assets that Remarble Earth, our asset register, records in Seoul, the five districts hold 45%, and the districts with the most are Jung-gu (20), Gangnam (15), Mapo (12) and Yeongdeungpo (12). The wider lodging market is there too: the same five districts hold 52% of Seoul's 10,629 operating lodging licences. In the established core the question is who runs the building next door, and what it is registered as. In Remarble's view, change there comes from repositioning existing stock rather than adding to it.

Where it is growing, and what is there

Seongdong-gu is where depth and pace meet. Its foreign visitor-days rose 57.4% to 5.4 million,¹ more than twice the city's rate,² and it is the only district with at least a tenth of Jung-gu's visitor volume to outpace the city by more than two points.² On the index, where Jung-gu's monthly average is 100, it stood at 16 in August 2025 against Jung-gu's 86, and at 29 in July 2026 against Jung-gu's 125.² The growth rests on a single year-on-year comparison.

Seongdong ranks sixth of the 25 districts for foreign visitor-days² and 11th for operating lodging licences, with 323. Earth records 5 institutional living assets there, 2 of them planned, against 15 in Gangnam and 20 in Jung-gu. Those are raw counts, planned assets included, and Gangnam and Jung-gu draw more visitors. Per million foreign visitor-days, Seongdong has 0.6 operating institutional assets, 0.9 counting planned ones, and 60 operating lodging licences; Jung-gu has 0.8 and 36, and Seoul as a whole 1.1 and 98.² Scaled to its visitors, Seongdong's supply is thinner than Seoul's on both measures and thinner than Jung-gu's in institutional assets only. Of the two measures, the institutional count is the closer guide to the operator-run competition a new building would face. The licence count covers licensed lodging of every kind: across Seoul, 68% of the licences are foreign-tourist urban homestays, in which, as the law defines them, a resident provides board and lodging to foreign tourists in the home they live in. In Remarble's view, licensing does not bind operators, who work as property managers; where they need a licence at all, it is not hard to obtain.⁵

In Korea's survey of departing foreign visitors, a growing share of respondents name Seongsu-dong, a neighbourhood in Seongdong-gu, as the place they liked best. The share was 5.0% in 2025, against 2.5% in 2024 and 1.1% in 2023;⁶ in the second quarter of 2026 it was 9.6%, against 5.2% a year earlier, second only to Myeong-dong's 18.3%.⁷ Seoul reports foreign card spending spreading beyond Gangnam and Myeong-dong to Hongdae and Seongsu-dong.⁸ The survey records preference and Seoul's report covers spending; neither counts visitors or identifies a cause.

In the core the operators are already there. In Seongdong the demand is growing fastest.

After Seongdong, the fastest growth is elsewhere in the city's northeast, as Seoul's master plan defines the region:⁴ Gwangjin at 37.0%, Seongbuk at 34.6% and Gangbuk at 31.5%, with Dongdaemun, at 26.1%, also ahead of Seoul's pace.¹ Five of the region's eight districts grew faster than the city; of the seven in the southwest, only Gwanak did.² The bases are small: Seongbuk's 1.3 million visitor-days for the whole year¹ are fewer than Jung-gu recorded in its quietest month.² Dongdaemun is a different case. Earth has recorded 6 institutional deals there since June 2024, by 5 investors, and those this site has examined were underwritten on student demand and a dormitory shortfall rather than on visitors. Visitors add a second source of demand there, too small to underwrite on by itself. Gwanak has the licences but not the visitors: 605 operating lodging licences, 5th of 25 districts, against a visitor index of 5.² The register does not say whom they serve.

The season, by the night and by the month

Across the 25 districts, January 2026 was the quietest month at 6.59 million visitor-days and July 2026 the busiest at 11.45 million,¹ so the trough ran at 58% of the peak.² The seasonal swing is not uniform, and a building is underwritten on its own district's pattern. On the index Jung-gu ran 78 in January against 125 in July; Seongdong 16 against 29; Gangnam 28 in February against 52; Yongsan 25 in February against 47.² For a product sold by the night, the swing shapes the revenue line, and the underwriting has to hold in the trough month. For a product let by the week or the month, a room in January depends on tenants this series does not count. One is the tenant between leases. Seoul is moving from jeonse, the lump-sum deposit lease, to monthly rent, which accounted for 53.4% of the new and renewed leases signed in the city in 2025 on apartments, row and multi-family houses and officetels, according to the land ministry's lease records; detached houses are not in that count. The other is the student: dormitory beds covered 15.3% of enrolment in 2025 at the 44 Seoul universities that disclose both.

The level is rising too. Every month from April to July 2026 topped 10.5 million, higher than any month from August to December 2025.¹ Twelve months are too few to separate trend from season; next year's figures for the same months will show which it is.

Why they come, and whether it holds

Korea's survey of departing foreign visitors asks those travelling for leisure or to see friends and family what drew their interest in Korea. In 2025 the most common answer, at 40.1%, was encountering Hallyu content (Korean popular culture). That share was 38.3% in 2024 and 32.1% in 2023, the first year the question was asked.⁶ The answer skews young and female: 52.8% of those in their 20s against 15.9% at 60 and over, 44.6% of women against 29.8% of men.⁶ Fewer act on it. Of the same visitors, 17.1% counted K-pop venues or drama and film locations among the activities they weighed, up from 14.2% in 2024, the largest rise among those activities; 13.5% of all respondents visited one, against 11.7%.⁶ A study published by KOFICE, the culture ministry's cultural-exchange agency, applies that 17.1% to all arrivals and counts 3.238 million "Hallyu tourists" in 2025, up 39.3%, while arrivals as a whole rose 15.7%.⁹ The link is drawn for individual sites too. The press credited K-content such as KPop Demon Hunters with the National Museum of Korea's record 2025; in the first half of 2026 the Yongsan museum drew 165,404 foreign visitors, 68.8% more than a year earlier.¹⁰

Whether it holds is less settled. In the provisional 2026 results the answer, reworded as K-content, stood at 33.3% and 34.5% in the first two quarters, against 41.9% and 38.1% a year earlier. It was still the most common answer, though its lead over the next answer was two points or less, and the rewording rules out a strict comparison.⁷ Arrivals kept rising, to 15.05 million in January to August 2026, 21.6% more than a year earlier, and China supplied 32% of them.¹¹ China is also where the Hallyu answer runs high, 45.2% in 2025 against 40.0% in 2024; in Japan it fell to 39.8% from 45.1%.⁶ Chinese tour groups have entered visa-free since 29 September 2025, under a waiver now set to end on 31 December 2026.¹² In July 2026 Chinese visits to Korea rose 29.0% on a year earlier while those to Japan fell 56.1%, a fall the Korea Culture and Tourism Institute put down to the Chinese government's advice against travel there.¹³ For Japanese arrivals the culture ministry names airlines' shift to short-haul routes as the trigger and lists K-culture among other conditions without weighting any of them.¹⁴ The Hallyu answer, the visa waiver and China's advice against travel to Japan all bear on the same market; none of these sources separates their effects, and neither does this piece.

What this cannot tell an investment committee

• Visitor-days are not visitors and not nights. A week's stay counts seven times and a day visitor once. Nothing here converts to room-nights, occupancy, a headcount or a market size.
• They come from one carrier's roaming records. The carrier's share of inbound roaming is not stated; if it moved between the two windows, every growth rate moved with it.
• They count foreign visitors only. The remote worker on a local SIM, the domestic tenant between leases and the student are invisible to them.
• They do not split by nationality or purpose, and cannot separate visitors who stay from those in transit. Gangseo, home to Gimpo airport, sits in the established core at 19 on the index;² how much of that reflects overnight stays there is unknown. The visitor survey cited here reports purpose for Korea as a whole and for stays of 90 days or less, never by district.
• The visitor data go no finer than the district, while a licence attaches to a single parcel: the Seongsu neighbourhood cannot be separated from the rest of Seongdong, nor Itaewon from Yongsan.
• A licence is not a room count. Rooms are reported for 4,684 of Seoul's 10,629 operating licences, and 68% of the licences are urban foreign guesthouses (foreign-tourist urban homestays, in which a resident lets rooms in their own home to foreign tourists).
• The visitor series is entered by hand rather than drawn from Earth, so the boards built on it are marked Cited; the licence and asset figures are Earth's own.

The lodging, districts and inbound boards on the Seoul page show the supply data and the national arrivals figures, and the operator map shows who runs what.