The Quiet Landings: Four Mid-Size Design, Food and Tech Companies Broke Into Japan This Year — and Not One of Them Started With a Flagship

The macro case for entering Japan has quietly become the strongest it has been in a generation. Inward foreign direct investment stock reached 53.3 trillion yen, a record, up 4.5% on the year. Greenfield investment — new operations rather than acquisitions — hit a record 31.6 billion US dollars, up 15.4%. In June 2025 the government raised its 2030 target for FDI stock from 100 trillion to 120 trillion yen, and signalled it wants 150 trillion as early as it can get there. That is not a slogan; it is a budget line with subsidies, tax treatment and free advisory capacity attached to it.

The company-level picture matches. JETRO's survey of foreign-affiliated firms in Japan, published on 26 February 2026 from 1,520 valid responses, found 61.6% expecting a profit for the current fiscal year — 3.6 times as many as expect a loss. Around 60% intend to strengthen or expand their Japanese operations, and retail is the sector most likely to say so. Just under half expect higher revenue this year, and more than half expect it next year. The complaints are real but specific: exchange-rate volatility, difficulty securing talent, worsening conditions for finding premises, and nearly 40% anticipating damage from US tariff policy.

What that data does not tell you is how anyone actually got in. Coverage of inbound investment is dominated by data centres, logistics parks and megadeals, because those are the ones with press offices. The mid-size entries — the design house doing 20 million euros, the regional restaurant group, the thirty-person deep-tech firm — get no coverage at all, which is why the folklore about entering Japan is roughly a decade out of date. Four entries completed or committed in 2026 are worth more than another survey, because they show the sequence, not the sentiment.

Manikomio DSGN is a handmade bag and accessories brand founded in Forlì in 2020 by Massimo Cabbia, entirely made in Italy. It now has more than 200 stockists and is expanding in Italy, Spain and — above all — Japan, which has become its leading overseas market. The route there had no flagship in it for five years. It ran on wholesale, then on pop-up counters inside Japanese department stores: Hankyu Men's in Osaka, Takashimaya in Tokyo. Only after those trials worked did a flagship appear on the calendar, in Ginza, at the end of 2026 — and it is the brand's local distributor opening it.

The department-store pop-up deserves more respect than Western brands give it. A two-week counter inside Hankyu Men's returns sell-through by SKU, price sensitivity you can read directly, a buyer's opinion that other buyers take seriously, and a customer list — for the cost of stock, a fixture and a few flights. It is the highest-quality short-term market test available in any developed retail market, it occasionally turns a profit while running, and most foreign brands skip it because a counter inside someone else's store feels like a demotion from the flagship in the pitch deck. Manikomio treated it as the experiment it is, and let the Japanese partner take the fixed-cost risk once the answer was known.

Bode arrived at the same discipline from the opposite direction. The New York label has been carried in the region for nearly a decade and sells through twenty high-end Asia-Pacific accounts, eight of them in Japan. On 20 February 2026 it opened its first company-owned Asian store — its fifth standalone store worldwide, after two in New York, one in Los Angeles and one in Paris. Emily Adams Bode Aujla's stated reason for going it alone rather than partnering with a wholesale account, the common route in Asia, was that they wanted to do it themselves. The point is that she could afford that choice, because ten years of wholesale had already answered the only question a first store cannot answer: do Japanese customers want this.

The address matters as much as the ownership. Bode skipped Aoyama and Ginza for Yoyogi-Uehara, on the ground floor of a residential building, designed with Green River Project. That is lower rent, a self-selecting customer who made a deliberate trip, and a store that argues the brand's own thesis — neighbourhood shops embedded in residential communities. In Japan the address is a statement about who you are, and the expensive address makes a statement many mid-size brands cannot back up on their second visit.

Fatburger entered Japan in 2018 with a Tokyo location and closed it in 2023. That is the ordinary shape of a foreign food failure in Japan, and most boards read it as a verdict on the market. In 2026 the brand came back — through a development agreement with Green Micro Factory Inc. for four restaurants in Okinawa over five years, the first of them now open in Kadena. Nothing about the product changed. Three things about the structure did.

First, ownership of the risk. A local development partner carries the capital expenditure, the hiring, the permits and the supplier relationships — which happen to be the four things that kill first attempts by foreign operators who underestimated how long each takes in Japan. Second, geography. Okinawa was chosen for tourism plus the steady foot traffic around the base presence: an English-comfortable, high-turnover customer base, rather than a Tokyo high street where an American burger brand has to out-execute Japan's own excellent burger chains on their home ground. Third, pace. Four units over five years is a schedule that survives one bad site; a land grab is not.

The useful lesson for a mid-size board is that Japan does not remember a small foreign brand's first failure. Consumers do not track chain exits. Landlords and distributors do — but what they read is why it failed, and a first attempt that died of undercapitalised expansion in the wrong location reads very differently from one that died of a product Japan rejected. What ends second attempts is repeating the first attempt's structure with a bigger budget.

There is a caveat inside the Okinawa logic worth stating plainly, because it applies to every foreign consumer business currently modelling Japan on tourist volume. The first half of 2026 brought Japan's first decline in visitor numbers in five years, and JTB projects 41.4 million for the full year, down 2.8%. Spending, though, still rose: 2.3 trillion yen in the first quarter, up 2.5%, and roughly 2.51 trillion in the second, up 0.2%, with shopping and food both growing. Fewer visitors, each spending more. A concept that only works at peak tourist volume is not a Japan business — it is a currency trade with a lease attached.

KREATiS was founded in 2014 near Lyon and does computational — in silico — assessment of chemical safety. Its iSafeRat platform supports REACH compliance, generates regulatory data, assists R&D and trains practitioners, all positioned as an alternative to animal testing. In April 2026 it registered KREATiS Japan K.K. in Chuo-ku, Tokyo, as the first step of an Asia-Pacific expansion. JETRO's Invest Japan Business Support Center handled the company registration, the tax and labour questions, the service-provider referrals and the PR. The go-to-market plan is to build a customer base primarily through academic conferences, aimed at researchers and industry specialists.

Read casually, that sounds like a company too small to hire salespeople. Read properly, it is the correct channel choice for a regulated technical market. Japanese buyers in compliance and R&D functions do not buy from a pitch; they buy from a reference they can cite internally. A conference paper co-presented with a Japanese researcher is not marketing — it is distribution, because it manufactures the third-party credibility a section manager needs in order to defend the purchase to the person above them. A sales team hired before that credibility exists spends its first eighteen months producing it at well over 15 million yen per head per year, and usually leaves before it lands.

The second half of the KREATiS story is the free infrastructure. JETRO's support centre removes most of the fixed set-up cost that mid-size boards use as their reason not to try: entity registration, tax, labour, and introductions to the accountants and lawyers who would otherwise take three months to find. The cost of discovering whether Japan works for you has collapsed. The cost of staffing Japan has not — this year's 4.9% rise in the minimum wage floor and the record run of labour-shortage bankruptcies are the other half of that sentence, and they are the reason the sequencing in these four cases matters more than it did five years ago.

Strip away the sectors and the same four decisions appear. Every one of these companies had evidence of Japanese demand before it signed a lease or hired a country manager — wholesale numbers, pop-up sell-through, a franchise partner's own market read, or a body of published science. In every case someone local carried a cost: a distributor, a franchisee, a wholesale network, or a government agency doing the paperwork for free. In three of the four, the address that worked was the one the pitch deck would have rejected — a residential street, a prefecture 1,500 kilometres from Tokyo, a conference hall. And all of them ran on a schedule measured in years rather than quarters, which is the honest shape of Japanese trust.

That is worth stating as arithmetic, because the failure mode is expensive and familiar. Expensive-first entry — the prestige lease, the imported country manager, the translated deck, all committed before a single data point exists — burns the runway during exactly the months when a foreign brand is least able to explain itself to Japanese customers. A credible mid-scale entry is six to eight months of operational work, not a ninety-day launch, and the committed version typically lands somewhere between 250,000 and 600,000 US dollars, most of it spent on localisation and one or two bilingual hires rather than on square metres.

This is the part of the work we do at Medusa Japan, and it is why we build the localisation layer before anything else: the site, the catalogue, the product language and the sales materials a Japanese buyer can actually act on. Both routes above depend on it. A distributor cannot amplify what it cannot explain, and a company-owned store cannot open ahead of demand nobody measured. Localisation is not the translation step at the end of the plan; it is the instrument that produces the evidence the rest of the plan is supposed to be based on.

One last note on timing. The support has never been better funded, and the cost of finding out has never been lower — but the cost of operating is rising with the wage floor, talent is the constraint every survey now names first, and the simplest route into an established Japanese business, buying one, faces a tighter screening regime from 2027. The quiet entries of 2026 are going to look obvious in 2028, and the companies still writing the business case will be paying 2028 prices for 2026 information.