Hello, this is Ryuta Hamamoto from TIMEWELL.
Whenever I talk with founders of deep tech startups, the same lament always comes back: "We have the technology. We have the papers. We have the prototypes. But we don't have a first customer." Even when R&D-phase grants can somehow be secured, the bridge from there—through the valley of death and into real-world use and mass production—simply has not been built. This question of "who buys the first unit" has been, in my view, the single most structural bottleneck that has held Japanese deep tech back for years.
On April 24, 2026, the Startup Policy Promotion Subcommittee, which was set up under the Japan Growth Strategy Council, unveiled a document designed to take that structure apart and rebuild it head-on. It is titled the "Government Procurement Promotion Package for Startups in the 17 Strategic Sectors (Draft)"[^11]. Three days later, on April 27, Prime Minister Sanae Takaichi declared at the opening ceremony of SusHi Tech Tokyo 2026 that "we will create a system in which each ministry pilots startup technologies in its own areas of operation"[^6][^7]. Most of the news coverage focused on the working-hours debate, but for startup founders the real headline was here.
In my own work at TIMEWELL, supporting collaboration projects between large enterprises and startups, I have repeatedly seen on the ground that there are domains where "private-sector enterprises simply cannot become the first customer"—technical fields that will never scale unless the government buys. Defense, space, quantum, fusion, frontier biotech: in domains where private demand alone is not enough to ignite a market, this package will move the needle. In this article, I want to break down the contents of the draft package down to the level of the tables and figures, and then think through what it actually delivers on the ground.
What the 17 Strategic Sectors Are: The Outline of a "Market the State Underwrites"
As a starting point, the 17 strategic sectors refer to the cluster of domains—AI and semiconductors, quantum, fusion energy, biotech, aviation, space, shipbuilding, materials (critical minerals and components), food tech, and others—that the Takaichi administration has decided to prioritize from the dual angles of economic security and growth investment. At the third meeting of the Japan Growth Strategy Council in March 2026, 61 "key products and technologies" tied to these 17 sectors were organized, and priority frontier technologies were also designated[^1].
Three perspectives drive selection: "the necessity of reducing domestic risk," "the potential to capture overseas markets," and "the innovativeness of the related technologies." In other words, what was selected is not simply markets that are large in size, but the intersection between supply-chain criticality and the realistic potential for Japanese companies to be internationally competitive. Concrete numbers were also written into the public-private investment roadmap: the semiconductor market is projected to reach roughly JPY 190 trillion by 2035, and Japan is targeting JPY 40 trillion in domestic production value by 2040[^1].
What matters here is that these 17 sectors are positioned not merely as a list of "support targets," but as "domains in which the government will be involved on the demand side." This is where the new procurement package fits in. The government will not just hand out R&D subsidies; it will buy the technologies that emerge from them. The state, in other words, is being designed as the first customer for strategic domains. In the United States, it is well known that the Department of Defense (DOD) helped grow SpaceX, Anduril, and Palantir. The Japanese version of "the government as a lead customer" is now being seriously built—and this is the policy intent we should read into the package.
Six Pillars of the Draft Package: Closing the Gap between R&D, Pilot Deployment, and Full-Scale Procurement
Reading the materials shown at the subcommittee, the package redesigns the entire flow from R&D to full-scale procurement, broken into three phases. The existing SBIR (Small Business Innovation Research) program is structured around Phase 1 (PoC and feasibility study support), Phase 2 (commercialization development support), and Phase 3 (technology demonstration and the like). The problem was always what happened after Phase 3. The technology demonstration is done. So how do you connect that to "pilot deployment and operation," and from there to "full-scale procurement"? That is exactly where the institutional vacuum sat.
The current issues raised in the materials are blunt: "We are not adequately responding to new R&D themes," "Budget allocation and procedures take too long, making outlooks unclear," "Long-term procurement outlooks are not presented, making it difficult for startups to invest and raise capital," and "Procurement-side capacity within ministries is limited." All of these have been pointed out by the startup community for years. From the government side too, the honest reality has been that "we want to buy, but we don't have the framework to do so."
In response to this diagnosis, six policy directions are laid out in the draft as the "directions for response"[^11].
First, strengthening large-scale technology demonstration support. This can be read as envisioning demonstration projects beyond the scale of the existing Phase 3—creating frameworks where demonstrations worth several hundred million to several billion yen can be entrusted to startups.
Second, strengthening R&D, pilot deployment, evaluation, and operation premised on full-scale government procurement (commissioned form). This phrase "premised on full-scale procurement" is decisive. It means designing R&D from the outset with procurement as the goal. Until now, R&D and procurement have been run under different systems, different budgets, and different departments, so even when R&D succeeded, it would not connect to procurement.
Third, establishing operational guidelines for ministry contracting practices. We will return to this below, but it is a horizontal initiative aimed at creating unified operational guidelines across all ministries and independent administrative agencies.
Fourth, building hands-on support systems that accompany startups all the way to procurement. This is a mechanism to prevent startups from getting lost after winning a grant, asking, "Now how do we get into the ministry's procurement window?"
Fifth, promoting ministry improvement plans and the use of catalogs such as the Digital Marketplace. Each ministry will have its own improvement plan, and the pathway by which startup products listed in government catalogs are actually used will be widened.
Sixth, expanding startup participation opportunities, including as subcontractors, and promoting disclosure. Even when a large company is awarded a contract as the prime contractor, transparency will be increased so that startups can enter as subcontractors.
When you line up all six, you can see the design intent: institutional mechanisms tailored to the realities of startups are being embedded into each of the three phases—the entry point (R&D support), the middle (pilot deployment), and the exit (full-scale procurement).
Looking for AI training and consulting?
Learn about WARP training programs and consulting services in our materials.
Six Items in the Draft Operational Guideline: Can the "Paperwork Hell" Be Cut Down?
Paired with the draft package, another document was presented: "Establishing Operational Guidelines for Contracting and Related Practices to Accelerate Public Procurement from Startups." It is a fairly bold proposal: to create unified operational guidelines across all ministries and independent administrative agencies. In tandem, the materials also discuss creating a "Startup Procurement Model Contract" as a contract template, and a "Startup Procurement Administrative Manual" that simplifies expense processing.
The fact that the model contract and administrative manual would be standardized across all ministries is, to anyone who knows the realities of public-sector procurement, a major impact. Until now, contract formats and expense-processing customs have differed subtly between ministries, so a startup trying to do business with two or three ministries faced an explosion of administrative cost. If standardized, the back-office burden on small startups could fall dramatically.
The six items raised in the draft operational guideline can be organized along the original document as follows.
(1) Bidding participation opportunities: The direction is to relax bidding qualification requirements for startups, including J-Startup-selected companies, and to confirm that the expanded qualifications are actually reflected in tender notices and applied in practice. Unless it shows up in the tender notice, the field will not move—so I think enforcing this consistently is critical.
(2) Procurement method: Active use of the "Startup Technology Proposal Evaluation Method," which permits negotiated contracts with startups that hold advanced and unique new technologies. This method was established at the Ministry of Defense (MOD) in 2024, and in May 2025, InfoStellar concluded the first contract under this method across all ministries—a satellite radio frequency analysis demonstration project[^2][^3]. It is a mechanism to solve the structural problem that emerging companies cannot win when track records dominate competitive bidding, by routing them through a separate channel of negotiated contracts.
(3) Contract performance bond: This one matters a lot in the field. The direction is to waive performance bonds for parties holding the unified all-ministry qualification. Performance bonds, which require pre-payment of a percentage of the contract amount, can run into tens of millions of yen on large projects—a body blow to startups with thin cash flow. Waiving these is the kind of move that signals the people drafting the guideline really understand the field.
(4) Plan changes: When thresholds for plan changes are too low, even minor modifications get caught up in approval processes that stall the project. The direction is to set thresholds appropriate to the project, which addresses a real bottleneck in operations.
(5) Payments: Payment-on-completion practices choke startup cash flow. The direction is to actively use estimated, advance, and partial payments. If pre-payment of a portion at the time of order becomes more common, startups will be able to run projects without leaning on bank loans or bridge financing.
(6) Re-contracting at full-scale procurement: This may look low-key, but it cuts to the heart of the problem. The conventional practice has been that even for procurement based on the results of national R&D programs, in principle a fresh round of competitive bidding is required at the full-scale procurement stage. After painstakingly running R&D together and refining the technology, a re-bid happens at the full-scale procurement stage and—too often—a similar Chinese or Indian product wins on price. Such laughable-but-real cases have actually happened. The direction is to use negotiated contracts for full-scale procurement based on R&D outcomes, finally giving institutional recognition to the continuity between R&D and procurement.
When you line up these items, you can see they are constructed from an extremely practical problem awareness: cut down both "paperwork hell" and "cash flow hell" at the same time. This is not paperwork-for-paperwork's-sake guidance; it is guidance that bites into cash flow and decision speed.
Fundamental SBIR Reform and the "New Pilot Deployment Framework": The Reach of the Prime Minister's Instructions
At the backbone of the package is the fundamental strengthening of the SBIR system itself. At the fourth meeting of the Japan Growth Strategy Council on April 22, 2026, Prime Minister Takaichi clearly instructed Minister Kiuchi and Minister Akazawa to "fundamentally strengthen the SBIR system, and study the creation of a new pilot deployment framework that goes beyond conventional R&D support and connects to full-scale procurement"[^4].
Two keywords stand out: "fundamental strengthening" and "a new pilot deployment framework that connects to full-scale procurement." The former means a root-and-branch review across both budget scale and program design. The latter is a clear signal of intent that "we will not let things end at PoC."
Japan's SBIR was originally introduced in 1999 with reference to the U.S. SBIR, but the original American philosophy of "demand creation by the government plus startup development" never fully transferred. The U.S. SBIR program (launched in 1982) has produced many of today's household-name growth companies, including Qualcomm, iRobot, and Gilead Sciences[^5]. The U.S. SBIR budget for FY2019 was approximately USD 3.29 billion (then about JPY 430 billion at the prevailing exchange rate). Of this, the Department of Defense (DOD) accounted for 47.9% and the Department of Health and Human Services (HHS) for 31.2%. The cycle is alive: the government directly funds R&D for the technologies it itself uses, and then buys the resulting outputs.
Japan's SBIR has long had the flavor of small- and medium-business support, with the U.S.-style axis of "continuous support based on problem-setting from administrative needs" being weak[^5]. The system was reviewed in 2020 and reborn as the new SBIR program led by the Cabinet Office, but the perennial issue—"there's nothing beyond PoC"—remained. This new procurement package is designed to swing the hammer directly at that remaining issue.
Particularly noteworthy is the phrase the Prime Minister herself used at SusHi Tech Tokyo 2026: "We will create a system in which each ministry pilots startup technologies in its own areas of operation"[^6][^7]. This is a distributed approach: rather than the Cabinet Office or METI handling everything, each ministry will trial startup technologies within its own work. The Ministry of Health, Labour and Welfare (MHLW) for healthcare and elderly-care settings; the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) for roads and airports; the Ministry of Agriculture, Forestry and Fisheries (MAFF) for plant factories and land-based aquaculture; the Ministry of Defense (MOD) for defense equipment, and so on. Startup technologies will be absorbed in line with the actual administrative challenges of each field. Once this starts to function, government procurement becomes far more diverse and three-dimensional.
Possibilities and Issues Visible from Leading Cases: InfoStellar, Food Tech, and Defense
Talking about institutional design is easy, but the cases of startups actually contracting with the government have been quietly accumulating.
The most iconic is the contract between InfoStellar and the Ministry of Defense, announced in May 2025[^2][^3]. It was the first-ever contract across all ministries under the Startup Technology Proposal Evaluation Method—a satellite radio frequency analysis demonstration project. This is more than just a "first case." It proves that the technology proposal evaluation method, a new framework for negotiated contracts, actually functions in practice. Since the war in Ukraine, the use of startups in defense tech has accelerated globally. It is well known that the U.S. firm Anduril Industries has built a track record in border surveillance and battlefield monitoring, and in Japan too, defense startups—from cardboard drones to artificial satellites—have started to emerge[^8].
Things are also moving in food tech. In March 2026, at the second meeting of the food tech working group at the Ministry of Agriculture, Forestry and Fisheries (MAFF), plant factories and land-based aquaculture were explicitly identified as priority targets for advance study[^9]. Plant factories are positioned as a technology that is resistant to climate volatility and capable of producing agricultural products at fixed times, fixed volumes, fixed prices, and fixed quality. The vision is to weaponize Japan's strength in modular, fully closed plant factories, with the government creating demand and then selling globally.
In the space sector, the guidebook records cases in which water-recycling shower equipment and care-assist suits developed by startups were used in disaster-stricken areas after the Noto Peninsula Earthquake[^10]. Startup participation is also visible in satellite constellations and satellite data utilization solutions. As these kinds of examples accumulate, the perception that "ministries can use startups" spreads inside ministries themselves. The first step is harder than imagined, but the second and third steps are surprisingly fast.
That said, there are issues to be honest about. Differences in literacy and motivation among procurement officers across ministries cannot be solved by program design alone. I have heard procurement officers say, more than once, that "I know there is something called the Startup Technology Proposal Evaluation Method, but I have never used it in my section. It is scary to take a flyer on a method with no precedent." That is exactly why the package's elements—"establishing hands-on support systems" and "ministry improvement plans"—matter. Unless each ministry holds a target like "this fiscal year, we will sign X number of startup contracts on this theme," with an execution-support system running alongside, the program risks becoming a paper tiger.
What Startups Need to Prepare: The Skill of "Making the Government a Customer"
So far we have mostly been talking about the government side, but there are also changes required on the startup side. Bluntly put, "making the government a customer" is not easy. It demands a different skill set from corporate B2B sales.
First, the ability to translate your technology into the language of "which administrative challenge it solves." For example, if an image-recognition startup says "our accuracy is 99%," nothing in a ministry will move. Proposals need to be translated into terms tied to administrative problem-setting and budget structures: "a technology that cuts the JPY XX billion annual cost of nationwide bridge inspections," or "completing post-disaster damage assessment in half the time of current methods."
Second, the ability to understand each ministry's decision-making process and annual cycle. If you go in selling without understanding the rhythm of budget request (August), budget approval (December), and execution (from the following April), you will not get results. You need to grasp which ministry is moving on which budget item at which time, and you need to chase tender information with your feet[^6].
Third, an organization that can absorb the administrative burden of tender documents and expense processing. Many startups operate at 5 to 20 people, where the back office consists of the founder and a single accounting hire. Pile public-sector paperwork on top of that, and the CEO ends up buried in document drafting. The package's direction toward common model contracts and a common administrative manual should help ease this pain point.
Fourth, the long-term perspective of cultivating R&D outcomes into "technologies that get chosen at the full-scale procurement stage." Even if you build a relationship with the government during R&D, when a re-bid happens at full-scale procurement, you can still lose to overseas players on price. The draft operational guideline indicates that negotiated contracts will be used, but which projects qualify will still be a field-level judgment. Securing the "uniqueness and difficulty of imitation" of your own technology—including patent strategy and the capture of first-mover advantage—remains the startup's own job.
Speaking from the position of someone supporting large-enterprise / startup collaboration projects at TIMEWELL, let me add one more thing: for startups that are eyeing government procurement, partner selection becomes especially important. Building demonstration track records together with major firms, then entering government procurement under their umbrella, is one royal road for tech-driven startups. The government is also using this package to back the design of having startups enter as subcontractors of large prime contractors—so this trend is, if anything, a tailwind.
Three Changes I Particularly Look Forward To
Personally, let me name the three changes in this package that I think will have the biggest impact.
The first is the normalization of fundraising through the visualization of "long-term procurement outlooks." The single biggest reason deep tech startups have struggled with fundraising is that the exit has not been visible. "Commercializing this technology takes five years and JPY 1 billion. But who the final customer will be is unclear." Faced with that, VCs naturally turn cautious. If the package's direction—"the government presents a long-term outlook premised on full-scale procurement"—takes hold, capital will flow more easily as "investment in technologies with a procurement commitment." That is not just a story about helping startups; it is the priming flow that lets private risk capital move toward deep tech with confidence.
The second is the change in cash-flow structure via performance bond waivers and the use of upfront payments. This sounds modest, but on the ground its effect is dramatic. Just two changes—not having to put down tens of millions of yen at contract signing, and receiving an estimated payment at project kickoff—lift the size of project a startup can take on by an entire tier. What used to be "we cannot bid because cash will not turn" becomes "we can bid." This is not a story about institutions; it is a story about the reality of running a business.
The third is the "domestic retention effect" of using negotiated contracts to convert R&D outcomes into full-scale procurement. Strategically, this is the most important to me. Letting technologies developed with Japanese tax money be carried away by overseas players in full-scale procurement re-bids—if you leave that structure intact while investing in 17 strategic sectors, the leakage of national wealth never stops. Designing a system in which a single startup can run all the way from R&D to full-scale procurement is rational from an economic security standpoint as well.
That said—just to be careful—I do not think this should degenerate into mere "domestic preference." The essence is "continuity between R&D and procurement," not "we buy because it's a Japanese company." The objective is to build a system in which startups with superior technology are evaluated consistently from the research stage through full-scale procurement, not the exclusion of foreign companies. Confusing the two creates a separate set of problems, including consistency with the WTO Government Procurement Agreement.
The Starting Line of "The Government as First Customer" Era
Let me quote, once more, a passage Prime Minister Takaichi delivered at the opening ceremony of SusHi Tech Tokyo 2026.
"For the creation and development of deep tech startups, we will provide seamless support across 17 strategic sectors—including AI, semiconductors, and quantum—from R&D through commercialization and social implementation. In particular, government procurement that creates demand for startup products and services—that is, the strengthening of SBIR—is essential. To that end, in addition to R&D support for startups, we will create a system in which each ministry pilots startup technologies in its own areas of operation, further lowering the bar to government procurement."
The two phrases "seamless support" and "government procurement to create demand" represent, in my reading, a landmark policy declaration in the postwar history of Japanese startup support. Up to now, METI's J-Startup, the SME Agency's Monozukuri subsidy, NEDO's large-scale R&D, and the SBIR programs of various ministries had been running in disjointed fashion. The new package threads them together with a single shared goal: "the government becomes the first customer."
Of course, programs do not run themselves just because they have been designed. A change in mindset among ministry procurement officers, an upgrade in startups' ability to propose, and ecosystem formation that pulls in large enterprises—these have to advance in parallel, or the package ends on paper. As we head into the summer's growth strategy compilation, I will be watching how far concrete program design and budget measures actually go.
Even so, the April 2026 subcommittee materials and the Prime Minister's speech were, I believe, a definitive moment for Japanese deep tech startups. A demand-side declaration that "the government will buy," packaged with this much specific operational-guideline detail, is probably without precedent. For startup founders who have survived by threading the gaps in the system, the time is finally arriving when they can walk straight down the middle of the institutional road.
At TIMEWELL, against the backdrop of this policy shift, we want to accelerate hands-on support that connects large enterprises, startups, and government as a trio. To startup founders eyeing government procurement, and to executives at large companies looking to combine their technologies with startups in proposals for government projects: let's build the concrete shape of "the era when the government becomes the first customer" together.
For specific business plans and procurement strategies, we are happy to discuss them individually through TIMEWELL's AI consulting service WARP. You can start with a 30-minute online consultation. For related policy and startup trends, see also our SusHi Tech Tokyo 2026 Comprehensive Summary Report, the Keynote Report, and Discontinuous Growth and Japanese Risk-Taking.
Appendix: Analytical Framework for the Government Procurement Package
In writing this article, I evaluated the policy contents against the analytical framework below (compliance checklist, evaluation matrix, procurement policy), incorporating the perspectives of public procurement and RFP/RFQ practice. Use it as a reference when considering how to position your own technology for government procurement as a deep tech startup founder.
Six Operational Guideline Items × RFP/RFQ Practical-Impact Matrix
| Item | Issue under current practice | Direction of response (draft) | Pre-work on the startup side |
|---|---|---|---|
| (1) Bidding participation | J-Startup and other relaxations are not reflected in tender notices | Reliable reflection of expanded eligibility in notices | Acquire J-Startup status / unified qualification |
| (2) Procurement method | Track-record bias keeps emerging firms from winning | Active use of the Startup Technology Proposal Evaluation Method (a form of negotiated contract) | Articulate and patent your unique technology |
| (3) Contract performance bond | Tens of millions of yen in upfront cash strain | Waiver for holders of all-ministry unified qualification | Acquire unified qualification |
| (4) Plan changes | Even minor changes face approval delays | Thresholds set appropriately to the project | Maintain change-history records |
| (5) Payments | Cash flow worsens under payment-on-completion | Active use of estimated, advance, and partial payments | Build invoicing infrastructure |
| (6) Re-contracting at full-scale procurement | Re-bidding required even with R&D outcomes | Use of negotiated contracts based on R&D results | Build track record from the R&D stage |
Shifts in the Vendor (Startup) Evaluation Matrix
In conventional government procurement, vendor evaluation has emphasized three points: past track record, financial stability, and price. Under the Startup Technology Proposal Evaluation Method, the evaluation is reframed along the following six axes.
| Evaluation axis | Weight (example) | Evaluation perspective |
|---|---|---|
| Technical uniqueness | 30% | Patents, papers, difficulty of imitation |
| Fit with administrative challenges | 20% | Alignment with ministry-level issues |
| Implementation feasibility | 15% | Progress in prototyping and demonstration |
| Team and organization | 15% | CTO, development structure |
| Cost | 10% | Fit with budget |
| Compliance | 10% | Security trade controls, economic security |
Compliance Checklist
A compliance checklist for startups entering government procurement in the 17 strategic sectors.
- System for confirming security trade controls (item classification check) in place
- Compliance with the Economic Security Promotion Act (Specified Critical Technologies, etc.)
- J-Startup status or unified bidding qualification acquired
- Applicability of the Startup Technology Proposal Evaluation Method assessed
- Compliance with specified-user requirements (defense projects, etc.)
- Conflict minerals / supply-chain due-diligence framework in place
- Compliance with the Subcontractor Act / Trade Fairness Act (January 2026 amendment)
Disclaimer
This article is a commentary based on publicly disclosed government materials and reporting, and is not intended as investment advice, legal advice, or procurement advice. The contents are based on information as of April 30, 2026, and the final substance of the policy may change through subsequent deliberation and decision-making processes. For specific decisions on bidding participation, contract terms, item classification, and the like, please consult the relevant ministry notices, applicable laws, and qualified experts (lawyers, tax accountants, security trade control specialists, etc.) and act at your own discretion. TIMEWELL Inc. and the author bear no responsibility for any actions or decisions taken based on the contents of this article.
References
[^1]: Asahi Shimbun SMBIZ, "List of 61 products and technologies the government will give priority support to: permanent magnets and perovskite solar cells included" — https://smbiz.asahi.com/article/16415793
[^2]: Tokyo Shimbun Advertising Bureau (PRTimes), "InfoStellar signs Ministry of Defense 'Satellite Radio Frequency Analysis Demonstration' project" — https://adv.tokyo-np.co.jp/prtimes/article28686/
[^3]: InfoStellar, "InfoStellar signs Ministry of Defense 'Satellite Radio Frequency Analysis Demonstration' project" — https://www.infostellar.net/jp/news/JMoD-RF-Analysis-demo
[^4]: Cabinet Secretariat, "Directions for Cross-Cutting Issues (4th Japan Growth Strategy Council, Document 2)," April 2026 — https://www.cas.go.jp/jp/seisaku/nipponseichosenryaku/kaigi/dai4/shiryou2.pdf
[^5]: Cabinet Office, Bureau of Science, Technology and Innovation, "On the New SBIR System," March 2023 — https://www8.cao.go.jp/cstp/gaiyo/sip/230316/sanko6.pdf
[^6]: Nikkei, "Prime Minister Takaichi: 'Use startup technologies in government operations'—supporting AI, quantum, and more," April 27, 2026 — https://www.nikkei.com/article/DGXZQOUA277BN0X20C26A4000000/
[^7]: Yomiuri Shimbun, "Prime Minister Takaichi: 'Startups should be the bearers who turn research outcomes into reality,'" April 27, 2026 — https://www.yomiuri.co.jp/economy/20260427-GYT1T00279/
[^8]: Jitsugyo no Nihon Forum, "From cardboard drones to artificial satellites: the rising era of 'space-self-defense-electronics' defense startups in Japan too," August 27, 2025 — https://forum.j-n.co.jp/narrative/8592/
[^9]: Foovo, "Government designates priority technologies in 17 strategic sectors / In food tech, plant factories and land-based aquaculture given priority study," April 6, 2026 — https://foodtech-japan.com/2026/04/06/japan-growth-strategy-3/
[^10]: Cabinet Office, Bureau of Science, Technology and Innovation, "Guidebook on Measures to Promote Public Procurement from Startups [Reference Materials]," January 2025 — https://www8.cao.go.jp/cstp/openinnovation/procurement/guidebook/02_sanko.pdf
[^11]: Cabinet Secretariat, "Government Procurement Promotion Package (Draft)" (3rd Startup Policy Promotion Subcommittee, Document 1), April 2026 — https://www.cas.go.jp/jp/seisaku/nipponseichosenryaku/startup/dai3/shiryou1.pdf