うどん7軒ハシゴ男・田中渓が語る丸亀製麺の韓国リベンジ！新宿、渋谷…「魔境すぎる」日本の駅はなぜ生まれた？
Udon chain Marugame Seimen returned to South Korea in 2025 and, in under a year, opened four locations that draw constant queues—exceeding both sales and investment targets with plans to reach eight stores by end of 2026. The episode also traces why stations like Shinjuku and Shibuya became labyrinthine 'dungeon stations,' a structural inevitability born from 130-plus years of independent rail expansions since the Meiji era.
Listen to this episode Go to the show page
What this episode answers
What three changes did Marugame Seimen make for its South Korea re-entry?
First, they franchised with Korean conglomerate Lotte Group, handing local siting and operations to a local partner. Second, they abandoned street-level shops entirely and concentrated on shopping malls and food courts to guarantee foot traffic. Third, they added locally adapted menu items—such as bulgogi udon and spicy green-chili kakeudon—while keeping the noodle quality and preparation method unchanged.
How do you tell apart a failure that warrants retreat from one that warrants another attempt?
If the product itself does not fit the market, retreat is the right call; if the product is sound but the location, pricing, partner, or operations were wrong, that is a failure of execution and re-entry is worth trying; a third category is simply bad timing—and confusing the three types is the key mistake to avoid.
Why did 'dungeon stations' like Shinjuku and Shibuya come to exist?
Shinjuku Station opened in 1885 as a single-line stop, then different private operators—Kobu Railway (later the Chuo Line), Odakyu, Keio, Seibu, the Marunouchi Line, and others—each added their own gates, platforms, and signage over more than a century with no unified master plan, and layer upon layer of department stores, station buildings, underground malls, hotels, offices, and convenience stores were folded in on top.
What share of the world's busiest stations by ridership are located in Japan?
Nine of the world's top-ten stations by ridership are in Japan—Shinjuku, Shibuya, Ikebukuro, Umeda, Yokohama, Kita-Senju, Tokyo, and Nagoya among them—with India's Howrah Station as the sole exception.
Has Tanaka himself experienced failing, regrouping, and succeeding on a second attempt?
Yes—Tanaka (田中) entered the UTMB (Ultra-Trail du Mont-Blanc), a 170 km trail race crossing the French, Swiss, and Italian borders with 10,000 m of elevation gain and a 46-hour cutoff, and was eliminated at the 110 km mark in 2023; he then overhauled his training, race experience, gear, and nutrition and completed the race in 2024.
Marugame Seimen's Three Changes for Its South Korea Comeback
Marugame Seimen (丸亀製麺) first entered South Korea in 2012, but a consumer boycott driven by Japan-Korea tensions, compounded immediately by the COVID-19 pandemic, forced it to withdraw. When it returned in 2025, results were strikingly different: four stores operating in under a year, constant queues at all hours, performance ahead of both sales projections and investment plans, and an expansion target of eight stores before the end of 2026.
The first change was a **shift in partner strategy**. On the initial attempt, Marugame handled everything itself—from finding properties to running operations—but pinpointing the true heart of a high-traffic area in a foreign market proved elusive, and prime locations were never secured. This time the company franchised with Lotte Group, one of South Korea's largest conglomerates, delegating real-estate acquisition and day-to-day operations entirely to the local partner.
The second change was **where to open**. Street-level standalone shops were abandoned in favour of shopping malls and food courts, committing fully to the principle of capturing existing foot traffic before worrying about anything else.
The third change was **menu localisation**. Items such as bulgogi udon and spicy green-chili kakeudon were introduced to match Korean eating habits, and kimchi-based udon is said to be in the pipeline. Crucially, the noodle quality, kneading method, and boiling process were left untouched—a deliberate choice to draw a clear line between what would and would not change, which Tanaka praises as the right call.
Reading the Type of Failure
Tanaka (田中渓) organises business failures into three categories. The first is a product that simply does not fit the market—structurally difficult, and a signal to consider full withdrawal. The second is a sound product let down by the wrong location, pricing, partner, or operations—an execution problem that justifies another attempt. The third is nothing more than bad timing. Marugame's first Korea exit combined the second and third types; Korean consumers had not rejected udon itself.
Tanaka's view: 'As long as you have not fully exited a market, every failure along the way is a textbook for eventual success.' He emphasises keeping options open by monitoring a market even while retreating. He lives this philosophy in his own athletic life. He entered the UTMB—a 170 km trail race through the mountains of France, Switzerland, and Italy, with 10,000 m of cumulative elevation gain and a 46-hour cutoff—and was timed out at the 110 km mark in 2023. He then rebuilt his training, race experience, equipment, and nutrition from scratch and finished the race in 2024.
Udon Crawl and a New Year's Eve to Forget
Tanaka's passion for udon runs deep. Around twenty years ago he made a pilgrimage to Kagawa Prefecture—branded 'Udon Prefecture'—and visited seven shops in a single day. He also carries a grimmer memory: as a university student, after drinking through New Year's Eve into the morning and with nowhere else to go, he stumbled into Hanamaru Udon (花丸うどん) in Shibuya Center-gai. Short on cash, he ordered plain udon and piled the complimentary tenkasu (tempura crumbs) into a mountain. He then became so ill he could barely move in the middle of Center-gai, making it 'the worst New Year ever.' He has since become a committed soba devotee and says he rarely touches udon anymore.
Why Dungeon Stations Exist
Speed-skater Miho Takagi (高木美帆) posted on social media that Kita-Senju Station's transfers were 'like navigating a demon realm,' sparking widespread discussion. Stations like Shinjuku, Shibuya, and Ikebukuro—nicknamed 'dungeon stations'—were not the product of a misguided master plan; there was no master plan at all.
Take Shinjuku Station as an example. It opened in 1885 as a stop for a single line, and over the following century Kobu Railway (甲武鉄道, later the Chuo Line—named by combining the first characters of Kai and Musashi provinces), Odakyu, Keio, Seibu, and the Marunouchi Line each grafted on their own gates, platforms, concourses, and signage. Tanaka describes it as 'a single-storey house where someone later added a second floor, then an annex, then dug a basement, and a different construction company bolted each part onto the last.'
Mutual through-service agreements—Keisei with the Toei Asakusa Line, the Chuo Line with the Tozai Line, the Joban Line with the Chiyoda Line, the Tobu Line with the Hibiya Line—improved connectivity but made the station structures even more intricate. Department stores, station buildings, underground shopping arcades, hotels, offices, supermarkets, and convenience stores were then folded in, producing places that feel simultaneously like a station, a shopping centre, and a bus terminal.
The outcome shows in the numbers: nine of the world's ten busiest stations by ridership are in Japan—Shinjuku, Shibuya, Ikebukuro, Umeda, Yokohama, Kita-Senju, Tokyo, and Nagoya among them. The only exception is Howrah Station in India. Shinjuku once required passengers to detour far north or south to move between the east and west exits; the East-West Free Passage opened in 2020 and eased that problem. Shibuya, by contrast, has only grown more maze-like since the Fukutoshin Line opened. Tanaka closes with an anecdote: a high-school student once proposed at a business competition an app that shows your precise underground position in real time—an idea he notes remains as urgently needed today as ever.
Takeaways
- Marugame Seimen's South Korea comeback succeeded because it correctly diagnosed its first exit as a failure of execution—not of product—and fixed exactly that by franchising with Lotte Group, concentrating on mall locations, and localising the menu while keeping noodle quality intact.
- Failures come in three distinct types—wrong product, wrong execution, and wrong timing—and identifying which type you are facing before deciding to retreat or retry is the single most important diagnostic step.
- Japan's dungeon stations are not design failures but the inevitable result of more than 130 years of independent construction by competing rail operators, a complexity inseparable from the density that puts nine of the world's ten busiest stations on Japanese soil.
This article is edited from a machine transcript of the first 30 minutes of the episode. It is not an official summary — please listen to the episode itself.