“What will Japan actually cost us?” is the question boards ask first and get answered worst. Public numbers cover incorporation fees; they miss the localization line item, the payroll multipliers, and — most expensively — the cash-flow cost of a 6–18 month enterprise sales cycle.
This paper breaks entry costs into four layers — setup, ongoing operations, localization, and time — and closes with a build-vs-partner total-cost model. Every figure is stated in JPY with the assumptions visible, so your finance team can adapt it to your own scenario.
1. The opportunity in numbers
- Market size: nominal GDP ~USD 4.4 trillion; the SaaS segment alone is projected at ~USD 20–23 billion by 2029–2030 depending on the forecaster.
- Under-penetration: inward FDI stock of ¥53.3 trillion equals 8.7% of GDP, versus an OECD average of roughly 63% — the widest “wealth vs. foreign presence” gap among developed markets.
- Policy support: national target of ¥120 trillion FDI stock by 2030; record greenfield inflows of USD 31.6 billion in 2024.
In short: the demand is real and the field is uncrowded. The rest of this paper prices the ticket.
2. Entity setup costs
| Item | Kabushiki Kaisha (KK) | Godo Kaisha (GK) |
|---|---|---|
| Registration & license tax | From ¥150,000 (or 0.7% of capital, whichever is higher) | From ¥60,000 |
| Articles notarization | ¥30,000–50,000 | Not required |
| Professional fees (judicial scrivener, filings) | ¥100,000–300,000 | ¥80,000–200,000 |
| Registered address / virtual office (year 1) | ¥100,000–400,000 | ¥100,000–400,000 |
| Typical all-in (excl. capital) | ¥400,000–1,000,000 | ¥250,000–700,000 |
| Timeline | 4–8 weeks | 3–6 weeks |
Capital itself is legally ¥1, but practically ¥5–10 million is the floor for bank-account opening and buyer credibility. If a founder needs a business-manager visa, plan for the ¥30 million capital requirement in force since October 2025.
3. Ongoing operating costs (annual)
| Category | Lean setup (2–3 staff) | Established (6–10 staff) |
|---|---|---|
| Accounting, tax & statutory filings | ¥300,000–800,000 | ¥1.5–4 million |
| Office (Tokyo, serviced → leased) | ¥1–4 million | ¥6–18 million |
| Fully-loaded payroll per mid-level hire | ¥7–10 million | ¥7–12 million |
| Social insurance (employer share) | ~15–16% on top of gross salary | Same, plus child-rearing contribution changes from Apr 2026 |
| Payroll, legal & HR admin | ¥500,000–1.5 million | ¥1.5–4 million |
The dominant cost is people. A credible enterprise sales team — one country lead, one AE, one inside-sales rep, part-time CS — runs ¥35–55 million per year fully loaded before any marketing spend.
4. The localization line item
Localization is the budget line Western entrants most consistently underestimate, because it is treated as “translation” when it is actually product, marketing, and sales engineering combined:
- Product & UI localization: ¥2–10 million initial depending on surface area, plus continuous localization of releases.
- Website, collateral & case studies: ¥1–5 million for a properly transcreated (not translated) core set, including ringi-ready decision packets.
- Support & documentation: Japanese business hours coverage; buyers treat English-only support as a disqualifier in most enterprise deals.
- Why translation alone fails: with English proficiency ranked 96th of 123 countries (EF EPI 2025, “very low” band), your buyer's evaluation committee will read only the Japanese versions of everything.
5. The hidden cost of time
Japanese B2B enterprise cycles run 6–18 months, with 5–12 stakeholders participating in the internal ringi approval. The financial consequence is not just slower revenue — it is a working-capital requirement: you carry a full local cost base for 2–4 quarters before first significant invoices.
The compensation: retention. Once a Japanese enterprise adopts, churn is structurally lower than global benchmarks — the same consensus process that slowed the sale now protects the renewal.
6. Build vs. partner: a cost model
Three archetype scenarios for the first 18 months (illustrative, JPY millions):
| Cost block | A: Build everything in-house | B: Entry-support partner + lean team | C: Distributor-led, no entity |
|---|---|---|---|
| Entity & compliance | 1.0–1.5 | 0.8–1.2 | 0 |
| People (fully loaded) | 55–85 | 25–45 | 0–10 |
| Localization | 8–18 | 6–12 (shared/managed) | 3–8 (partner-driven) |
| Market research & GTM services | 3–8 (agencies ad hoc) | 8–15 (bundled: research, inside sales, deal support) | 2–5 |
| 18-month total | ≈67–112 | ≈40–73 | ≈5–23 |
| Margin & control | Full margin, full control | Full margin, high control | −30–50% margin, low control |
Scenario C is cheapest but caps the upside and surrenders the customer relationship. Scenario A is right for companies with committed nine-figure ambitions and patient capital. For most Western startups and mid-market entrants, Scenario B — a lean owned team leveraged by an entry-support partner for research, inside sales, localization, and deal support — delivers the best cost-to-learning ratio in the first 18 months.