Fukuoka Opens Global Expansion Program for Regional Deep-Tech Startups

Long before “global expansion” became the language of startup policy, Hakata merchants understood that a port is not a destination. It is an instrument for reaching the other shore. A gold seal sent from China in the first century, missions that sailed toward the continent, the foreign guests housed at the Kōrokan and the merchant networks of medieval Hakata all placed the city at the edge of Japan and the center of an exchange.

Fukuoka City is now trying to give that old outward instinct a scientific form. Its Deep Tech Expansion Program 2026 will choose roughly five startups or prospective founders and help them identify customers, research collaborators, clinical or industrial validation partners and other institutions in Singapore, Malaysia, the Philippines, Thailand, Indonesia, Vietnam or Taiwan. The route will be designed separately for each technology.

The program is deliberately narrower than the typical stage-and-pitch accelerator. The city’s procurement specification requires at least 20 prospective overseas counterparties and at least 10 business meetings for each participant. It asks the operator to stay after those meetings and work toward continuing negotiations, proof-of-concept projects, joint research, demonstrations and contracts. Leave a Nest’s recruitment notice names NDAs, letters of intent and memoranda of understanding among the possible intermediate results.

That emphasis reflects an awkward truth about deep tech. A software founder can sometimes translate an interface, buy digital advertising and test a foreign market within weeks. A company built on a new biomarker, fermentation process, crop treatment, carbon material or medical device may need a local laboratory, samples, ethics review, regulatory advice, manufacturing validation and a partner willing to run a trial. The first international sale can begin years before money changes hands.

What Fukuoka is buying

The distinction between a public contract and startup financing matters. Fukuoka City set a ceiling of ¥29.876 million, including tax, for the commissioned program through March 31, 2027. That sum pays for recruitment, analysis, mentoring, partner research, meeting coordination, overseas activity, follow-up, administration and evaluation. It is not a ¥29.876 million grant pool to be divided among five companies, and the announcement promises no equity investment.

Travel and accommodation for the overseas program are expected to be subsidized, with detailed eligible costs, limits and conditions given to selected participants. The city’s specification says the contractor will generally bear those costs within a per-company ceiling agreed with the city. Applicants should wait for the written terms before assuming that every traveler, ticket or extended stay will be covered.

The operator is Leave a Nest, selected as the highest-rated bidder after a city evaluation committee considered five applicants in July. The company began with science and engineering graduate students in 2002 and has built its TECH PLANTER platform across Japan and Southeast Asia. It maintains group operations in Singapore, Malaysia and the Philippines and has spent years connecting researchers, startups, manufacturers and investors across the six ASEAN countries named by the Fukuoka program.

That network is the asset the city is purchasing. A list of email addresses is not market access. A useful intermediary must know which hospital can approve a study, which university laboratory has the necessary equipment, which corporate division owns a problem, which distributor understands a regulated product and which contact is empowered to say yes. It must also know when a startup is not ready to travel.

Who can apply—and the local bargain

The program is open to a Fukuoka company or to an outside company willing to establish an office or other business base in the city by March 31, 2028. A startup should generally have been incorporated within about 15 years as of April 1, 2026. A prospective founder must plan to incorporate in Fukuoka by the same 2028 deadline and normally possess a product or prototype. A research seed without a prototype may still qualify if the reviewers believe the program can produce concrete progress.

All applicants need the intention and operating capacity to pursue overseas development during the program. This is more demanding than curiosity. A founder must make decisions, prepare material, travel or participate in an equivalent local program, respond after meetings and assign technical people when a partner asks detailed questions. For a five-person research company, that opportunity cost can be as consequential as airfare.

The location requirement reveals the city’s bargain. Public money is being used to help technologies that are already in Fukuoka or will plant an operating presence there. The program is regional in economic purpose, but it is not an unrestricted benefit for every Kyushu startup. A company outside the city must be prepared to root part of its future in Fukuoka.

Applications are due August 28. Selection and notification are planned for early to mid-September, followed by a joint kickoff in late September. Individual interviews, strategy and meeting preparation run through September and October. Overseas programs are planned for November 2026 and January 2027; follow-up continues into February, when participants are expected to report results in Fukuoka. An information session will be held August 19 at F Lab Kyushu University Hospital.

Deep tech is not a compliment; it is a financing problem

“Deep tech” is often applied so broadly that it can mean any startup with a scientist. The useful definition is operational. These companies build on formal research or substantial engineering; face technical uncertainty that cannot be removed by ordinary product design; require long development and expensive validation; and often must create a supply chain, regulatory path or new market before revenue can scale.

Japan’s New Energy and Industrial Technology Development Organization describes the category in similar terms: long research and commercialization periods, large capital needs, high uncertainty and business models that cannot simply be copied from existing industries. These traits make private capital hesitate precisely when the technology still needs expensive evidence. That gap is commonly called the valley of death, but there is rarely only one valley.

There is a technical valley between a laboratory result and a repeatable prototype. A manufacturing valley separates the prototype from a product made at cost and quality. A regulatory valley stands between a plausible medical benefit and permission to test or sell. A market valley lies between solving an interesting problem and solving a problem for which a named customer will pay. International expansion adds language, intellectual-property strategy, data rules, standards, import requirements, reimbursement and local trust.

A strong overseas program cannot eliminate those valleys. It can help a company discover the correct one before spending scarce capital on the wrong bridge. A hospital that refuses a proposed pilot may reveal a missing endpoint. A food company may say the formulation works but the local price does not. A university collaborator may identify a validation method that regulators will accept. Negative evidence can be a productive outcome if it arrives early and changes the plan.

Seven destinations, not one “Asian market”

Fukuoka’s geography makes the southward and westward orientation intuitive. The city government describes an airport, port, central station and convention district concentrated within roughly 2.5 kilometers; the subway ride from airport to city center takes about five minutes. Across the water lie markets with young populations, expanding cities, stressed food and health systems, industrial supply chains and climate risks that can make research-based solutions valuable.

But Southeast Asia is not a single test market. Singapore offers concentrated capital, biomedical research, corporate regional headquarters and a demanding regulatory environment. Malaysia combines advanced manufacturing with different state and national institutions. Indonesia and the Philippines are archipelagic markets where logistics and price can dominate adoption. Thailand has deep automotive, food and medical networks. Vietnam’s manufacturing base and fast-growing cities create different partnerships. Taiwan adds globally connected electronics, precision manufacturing, medicine and agricultural technology.

The right destination therefore depends on the next risk a company must remove. A medical startup may need a clinical collaborator rather than a distributor. An agricultural biology company may need tropical field data. A materials venture may seek a pilot line, while a carbon technology needs a source of feedstock and a buyer for verified reductions. Choosing a country because it hosts a famous startup event is weaker than choosing it because one institution can answer the next decisive question.

Regulation also crosses borders imperfectly. Clinical data accepted in one jurisdiction may not be sufficient in another. Biosafety, genetic resources, data localization, medical-device classification, food labeling and environmental claims vary. Contracts must define background intellectual property, newly generated results, publication rights and the law governing disputes. A memorandum can open a door, but it cannot substitute for those terms.

Fukuoka’s long experiment with openness

The city’s startup policy did not begin with this program. It began with an economic question: could a compact regional capital create new headquarters and growth companies instead of remaining primarily a branch-office economy? In 2012, Mayor Soichiro Takashima issued the “Startup City Fukuoka” declaration. The city placed entrepreneurship inside a larger story about youth, urban convenience and connection to Asia.

In May 2014, the national government designated Fukuoka as a National Strategic Special Zone for global startups and job creation. The mechanism mattered because cities alone cannot rewrite immigration, labor or national tax rules. The special-zone framework allowed Fukuoka to propose and test regulatory changes, pair local services with national exemptions and make entrepreneurship an administrative priority.

In 2015, Fukuoka became the first Japanese city to use the special-zone startup visa. Foreign founders who could not yet meet the normal Business Manager visa’s office, staffing or capital requirements received a preparation period under municipal review. The policy treated immigration not as an afterthought but as part of the startup infrastructure.

In April 2017, Fukuoka Growth Next opened in the former Daimyo Elementary School. The symbolism was almost too neat: classrooms built for an earlier generation became offices, mentoring rooms and a startup café for the next. A global startup center brought lawyers, accountants and advisers into a one-stop environment. The city later combined that physical hub with demonstration support, corporate matching, rent assistance and overseas partnerships.

From creating companies to growing difficult ones

Fukuoka’s own policy review shows measurable breadth. In fiscal 2023, its startup café handled 4,117 consultations. The cumulative number of companies supported as tenants in startup facilities reached 635. Sixty-one local startups were valued at ¥1 billion or more, up from 51 the year before. The city reported 101 demonstration projects supported through its public–private one-stop desk on a cumulative basis.

Those numbers demonstrate activity, not necessarily global scale. The same review recorded ¥5.3 billion raised by 26 resident companies in fiscal 2023, down from ¥11.1 billion by 29 companies the prior year. Funding moves unevenly, and company valuations do not equal revenue, scientific success or durable employment. A mature ecosystem must examine outcomes as carefully as it celebrates entries.

Deep tech exposes the next stage of the challenge. General startup support can help a founder register a company, rent a desk and learn to pitch. It cannot by itself finance a clinical trial, qualify a new material, build a fermentation plant or negotiate access to a foreign hospital. The required network reaches universities, manufacturers, regulators, specialized investors and large customers that can absorb technical risk.

Kyushu University gives the region an unusually broad research base—from medicine, biotechnology and agriculture to hydrogen, organic electronics, materials and advanced computing. In 2021 the university reported 124 affiliated startups as of fiscal 2020, including four that had completed initial public offerings. It had also built a gap fund to test university research as a business.

PARKS—the Platform for All Regions of Kyushu & Okinawa for Startup-ecosystem—expanded that effort in 2022. Led by Kyushu University and Kyushu Institute of Technology, it connects universities from Yamaguchi through Kyushu to Okinawa with FFG Venture Business Partners. National support awarded in 2024 aims to develop a system in which talent, knowledge and capital circulate, including links with Taiwan and Singapore and a planned inter-university structure by fiscal 2029.

The new Fukuoka program sits downstream from that research pipeline. PARKS and university gap funds try to create investable companies from research seeds. Deep Tech Expansion Program 2026 asks what happens when one of those companies must find evidence and counterparties beyond Japan. The two stages are complementary, but the handoff must be deliberate: too early, and the founder tours overseas without a testable offer; too late, and the product may be built around assumptions from the domestic market.

A city program inside a national wager

Japan’s government made startups a national growth priority in its 2022 five-year plan. It set an ambition to increase annual startup investment more than tenfold to ¥10 trillion by fiscal 2027 and ultimately create 100 unicorns and 100,000 startups. A ¥100 billion university-startup fund, expanded global support and new deep-tech programs became parts of that wager.

Large national funds address research and capital. Cities address a different layer: density and coordination. A founder needs to encounter university technology-transfer staff, a corporate engineer, a specialized investor, a regulator, a foreign entrepreneur and a mayoral office without navigating six disconnected systems. OECD work on startup internationalization emphasizes networks, links to established companies, dedicated overseas programs, integration with research institutions and serious impact evaluation.

Fukuoka’s program mirrors those recommendations unusually closely. It is publicly commissioned but privately operated. It is specialized by sector, tied to research and designed around foreign intermediaries. It requires aftercare and data collection, including an analysis of why supported cases did or did not produce outcomes. That last requirement may be the most important. Accelerator policy often counts events, attendees and introductions because they are easy to count. Deep-tech impact emerges later and resists a photograph.

The contract’s activity KPIs—20 candidate counterparties and 10 meetings per company—provide accountability, but they can also create perverse incentives. A startup does not benefit from ten weak meetings more than from three decisive ones. The operator must set outcome KPIs for continuing discussions, joint research, pilots and contracts. The quality of the funnel matters more than its width.

The paper staircase from meeting to market

The recruitment notice lists a progression that will be familiar to founders: NDA, LOI, MOU, proof of concept, joint research and perhaps a commercial agreement. These documents can mark genuine movement. An NDA allows technical disclosure. A letter of intent records a proposed direction. A memorandum assigns areas of cooperation. A PoC generates evidence. Joint research can divide work, cost and intellectual property.

Yet paper is not revenue. NDAs are routine and may lead nowhere. Many MOUs express goodwill without budget or enforceable purchase obligations. A PoC can become “pilot purgatory,” repeated without a path to procurement. A meaningful result identifies the decision-maker, success criteria, resources, timeline and next commercial gate. If a hospital trial succeeds, who submits to the regulator? If a factory test works, who pays for scale-up? If a field demonstration improves yield, who becomes the buyer?

The six-month program period is short by biotechnology or materials standards. Fukuoka should not judge every company by whether it books international sales before March 2027. A fair evaluation would distinguish verified learning from ceremonial activity: a signed study protocol, secured sample access, an approved pilot budget, regulatory classification, a co-development contract, licensed data, follow-on financing linked to the overseas work or a documented decision not to enter an unsuitable market.

Risks the program cannot ignore

Selection is the first risk. A company can be scientifically brilliant and still be unready for this timetable. Conversely, a polished English pitch can conceal weak intellectual property or poor reproducibility. Reviewers need to test the founder’s capacity, ownership of the underlying research, freedom to operate, quality system, evidence package and the specificity of the overseas question—not just presentation quality.

Market selection is the second. Sending every company to the same two fashionable cities would simplify logistics and defeat the individualized premise. The correct partner may be in a provincial agricultural institute, a specialized hospital, an industrial estate or a manufacturing cluster away from the usual conference circuit. Leave a Nest’s value will be measured by the relevance of the introductions, not the recognizability of the venues.

Confidentiality and intellectual property are the third. Early disclosure can damage patent rights in some jurisdictions. Joint data may become difficult to separate. Universities and founders may hold overlapping rights. Before a detailed overseas conversation, participants need a disclosure map: what is public, what is confidential, what requires university approval and what claims remain unfiled.

Continuity is the fourth. International partnerships fail in the weeks after travel, when founders return to product crises and counterparties return to their own priorities. Every meeting needs an owner on each side, a written next action and a date. The program promises follow-up through February; the harder question is who maintains the relationship after the public contract ends in March.

Finally, there is policy additionality: would the same company and partner have met without city support? Public value is highest when the program supplies trust, specialist access or coordination the market would not otherwise provide. Subsidizing ordinary business travel for already international companies would create less value than helping a technically credible but network-poor team cross its first border.

A bridge must carry weight

Fukuoka has spent fourteen years building the visible architecture of a startup city: a declaration, a special zone, a visa, a converted school, consultation desks, demonstration programs, overseas alliances and a citywide technology week. The infrastructure has helped make entrepreneurship legible and accessible. The next test is not whether another event can fill a room. It is whether a small research company can use that system to cross from evidence to adoption.

The city’s history offers the right metaphor. Hakata prospered not because ships looked impressive in port, but because exchange moved through them—books and medicines, tools and food, language and capital. A modern innovation bridge must also carry something measurable in both directions. Fukuoka technology needs overseas problems, samples, partners and customers. Local laboratories need to absorb knowledge about regulation, price and use. Foreign collaborators need a reason to remain connected after the delegation leaves.

Deep Tech Expansion Program 2026 is small enough to be personal. Five participants allow a serious operator to learn each technology rather than provide generic export lectures. The required partner lists and meetings impose discipline. Travel support lowers a real barrier. The focus on aftercare recognizes that a business relationship begins where a pitch tour ends.

It is also small enough that every selection matters. The contract ceiling is modest beside the capital required to develop a drug, material or climate system. Six months cannot compress biological validation or factory qualification into software speed. An MOU may make a good photograph and a poor business. Those limits should define the program’s honesty, not diminish its ambition.