NiKirei

1. One-liner

Turns a phoned-in truck booking into the written terms Japan’s new freight law demands, before the driver rolls.

2. Trend signal — why now?

On 1 January 2026 Japan replaced 下請法 (the Subcontract Act) with 取適法 — the Act on Ensuring Fair Transactions for SME Subcontractors. Buried in that reform is a new transaction category, 特定運送委託 (specific transport commission), and it did something that had never been done before: it made the shipper — the manufacturer, wholesaler, or EC seller who books the truck — a directly liable party.

Before 2026, when a maker or wholesaler handed freight to a carrier, that transaction sat outside the Subcontract Act entirely. Regulators watched the practice curdle: carriers forced to wait unpaid at loading docks (荷待ち), made to do loading, labelling and shelf-stocking for free (荷役), squeezed on rates with no negotiation. So the reform pulled the shipper inside the fence.

What a covered shipper now owes, per freight order:

• 四条書面 — written terms issued immediately upon ordering, containing 12 mandatory fields: both party names, commission date, service content, delivery deadline, delivery location, inspection completion date, freight amount, payment date, and details of any promissory note, consolidated payment, electronic receivable, or supplied materials.
• No lump-sum pricing. “一式◯◯◯円” is explicitly non-compliant. Ancillary work — loading, labelling, inspection — must be itemised with its own price, separately from the freight charge.
• Per-shipment documents. A standing contract does not cover you; each shipment whose conditions vary needs its own written order.
• 2-year record retention of order documents, delivery completion dates, payment amounts and actual payment dates, any changes to terms with reasons, and negotiation histories.
• 14.6% annual late interest on anything unpaid past 60 days.

Enforcement escalates guidance → 勧告 (formal recommendation) → 社名公表 (public naming of the company), run jointly by the JFTC, the SME Agency, and MLIT’s トラック・物流Gメン.

And here is the gap. A 2026 industry survey found 32.2% of logistics-sector companies were doing nothing at all about the law reform — and among firms under 50 employees, half reported no response. Separately, 28.8% named “dependence on paper, FAX and telephone” as the blocker to digitalisation. Meanwhile the practitioner literature is blunt that “most shipper companies’ existing ordering systems do not meet the 四条書面 requirements.”

The obligation landed in January. The behaviour did not move. That gap is the product.

Provenance:
  - Signal 1 (demand): 32.2% of logistics companies taking no action on the 2026 law reform; half of sub-50-employee firms unprepared; 28.8% cite paper/FAX/phone dependence as the DX blocker — https://b-soudan.impress.co.jp/article/detail/1518 — 2026
  - Signal 2 (feasibility): 特定運送委託 added 1 Jan 2026 with 12 mandatory 四条書面 fields, immediate-issuance rule, itemised ancillary pricing, 2-year retention, 14.6% late interest, and 社名公表 enforcement — https://toritekihou.com/tokuteiunsoitaku/ and https://hacobu.jp/blog/archives/6209 — January 2026
  - Signal 3 (economic): Incumbent compliant dispatch systems (MOVO Vista) are quote-only enterprise TMS with undisclosed pricing and no self-serve SMB tier — https://boxil.jp/service/13038/ — 2026
  Category: Regulatory arbitrage

3. The opportunity

The compliant tooling that exists was built for the wrong customer.

MOVO Vista (Hacobu) is the named 四条書面-compliant dispatch system. It auto-generates properly formatted order documents from freight data, connects shippers to multiple carriers, and runs first-come or competitive-quote workflows. It is also a full inter-company logistics DX platform sold on a quote-only basis — BOXIL lists both its price and setup cost as undisclosed, with pricing varying by deployment and requiring a sales conversation. That is enterprise TMS shape: a procurement cycle, an integration project, and a number you have to ask for.

The company that actually got caught by this reform is a 40-person parts maker in Aichi that ships 15 pallets a week, books trucks by calling the same two carriers it has used for twenty years, and confirms by FAX. It does not have a TMS. It is not going to buy one. It cannot articulate what a 四条書面 is, and it is now one 勧告 away from having its name published by the JFTC.

The 10× is not better dispatch optimisation. It is meeting the shipper inside the workflow they already have — the phone call, the FAX, the LINE message — and emitting a compliant document out the back without asking them to adopt a logistics platform. The incumbent makes you change how you order freight. This makes the way you already order freight legal.

The itemisation rule is the specific wedge. A shipper who has always said “Osaka run, ¥45,000, all in” is now non-compliant by default, because the loading work folded into that number must be priced separately. Every legacy verbal arrangement in the country is a defect. Nobody is going to unpick those by hand.

4. Target market

Primary customer: Operations or 総務 (general affairs) manager at a Japanese manufacturer, wholesaler, or EC seller with ¥10M–¥300M capital or 100–500 employees, shipping 20–500 freight orders a month to a small stable of contracted carriers, with no TMS and no in-house legal function.

Why they buy: Not because they want software. Because the ordering side of a freight transaction became a named liability with a public-shaming enforcement step, their trade association told them so in a seminar, and they have no idea whether the FAX they sent yesterday satisfies twelve statutory fields. The practitioner guidance is explicit that existing ordering systems do not meet the requirement — so the honest ones already know they are exposed.

Rough TAM reasoning: Japan has roughly 3.3 million SMEs, of which manufacturing and wholesale — the two sectors most likely to commission outbound freight toward a customer — are a large minority. The covered set is narrower than the whole: it needs the shipper to be above threshold, the carrier below, and the freight to move goods toward delivery to the ordering party (internal factory-to-warehouse moves are excluded). A conservative reading puts serviceable prospects in the low hundreds of thousands. At the ¥12,000–30,000/month range this idea prices at, capturing 0.1% of that is a real business.

Why now for them: The obligation is live, retroactive interest accrues from the January implementation, and MLIT’s トラック・物流Gメン are actively conducting shipper-side investigations. There is no grace period left to wait out.

5. Product sketch (MVP)

• Order intake by voice, LINE, or FAX — the shipper books freight the way they always have. A phone call to a dedicated number is transcribed; a LINE message or a scanned FAX is parsed. No behaviour change required on day one.
• Twelve-field completeness check — the system extracts what was said, maps it to the 12 statutory fields, and flags exactly what is missing. “You didn’t state a payment date or an inspection completion date.”
• Lump-sum splitter — detects “一式” pricing and prompts the operator to itemise ancillary work (loading, labelling, inspection, waiting) with separate line prices, since bundled pricing is explicitly non-compliant.
• Compliant 四条書面 issued immediately — generated and delivered to the carrier by email the moment the order is confirmed. The 2026 reform permits electronic delivery without needing the carrier’s prior consent, which removes the last friction that made e-delivery impractical.
• Standing-terms templates — for the two or three carriers a shipper uses every week, pre-fill the stable fields so a repeat booking is one confirmation tap.
• 2-year evidence archive — every order document, delivery completion date, payment amount and actual payment date, term change with its stated reason, and negotiation history, retained and searchable in the statutory format.
• 60-day payment clock — tracks each order against its payment deadline and warns before 14.6% annual late interest starts running.
• Inspection-ready export — one bundle per carrier per period, formatted for a JFTC or Gメン document request.

6. AI angle — what’s load-bearing

Remove the AI and this is a form. Nobody with a FAX machine fills in a form.

The load-bearing work is turning unstructured, habitual freight ordering into twelve structured statutory fields. A dispatcher says “田中さん、明日の朝イチで大阪、パレット12枚、いつもの条件で” — and the system has to know that “いつもの条件” resolves to a specific standing arrangement, that “朝イチ” needs to become a stated delivery deadline, that pallet count implies a weight field, and that no payment date was mentioned at all. Japanese business speech is elliptical and highly context-dependent; the honorific register and dropped subjects are exactly what classical parsing fails at and what modern models handle.

The second AI job is detecting embedded ancillary work in a lump-sum price. When an order says “積み込み込みで45,000円,” the system must recognise that loading is bundled, that this is now non-compliant, and prompt for a split. That is a judgment call over natural language, not a lookup.

The third is FAX and handwritten-slip OCR — the input format 28.8% of this market is still stuck on.

This is a translation problem between how Japanese logistics actually talks and what the statute demands. That translation is the entire product.

7. Localization angle

This is Japan-only by construction and that is the point, not a limitation.

• Language: The product lives or dies on Japanese business-register comprehension — keigo, dropped subjects, industry shorthand, and place-name/route abbreviations. A generic English-first TMS with a Japanese UI layer cannot do the voice-intake job.
• Regulation: The 12 fields, the itemisation ban, the 2-year retention window, the 14.6% interest rate, and the 社名公表 escalation are all specific to 取適法. The product is a direct encoding of one statute.
• Channel: LINE is the default business messaging rail for Japanese SMEs, and FAX remains genuinely live in logistics — supporting it is a feature, not a legacy concession.
• Pricing: ¥12,000–30,000/month sits below the threshold where a Japanese SME needs a formal 稟議 (ringi) approval circuit, which is what makes self-serve possible in a market that normally buys software through relationships.
• Distribution: Trade associations (業界団体), 商工会議所 chambers, and the 中小企業診断士 consultant corps are the real channels, and they are already running 取適法 seminars.

8. Business model — path to $1M–$5M ARR

• Pricing: ¥12,000/mo (Starter, ≤50 orders/mo) · ¥30,000/mo (Standard, ≤300 orders/mo) · ¥60,000/mo (Multi-site, unlimited + multiple shipping locations)
• ACV: ¥300,000 ($2,000) blended
• Rough math to $1M ARR: 500 customers × ¥25,000/mo × 12 = ¥150M ≈ $1M
• Rough math to $5M ARR: ~2,000 customers at a slightly richer ¥31,000 blended ACV, which realistically requires the carrier side paying too (see expansion) and one adjacent statutory module
• Expansion path: (1) per-shipping-location pricing as multi-site manufacturers roll out; (2) a carrier-side receiving tier — the carrier who receives compliant documents from one shipper wants the same archive across all their shippers, and they are the party with the 14.6% interest claim; (3) the 60-day payment clock is a natural on-ramp to invoice reconciliation; (4) adjacent 取適法 modules for non-freight subcontracting, which shares the same 四条書面 machinery.

Gross margin is strong — voice transcription and model inference on 20–500 short orders a month is a rounding error against a ¥25,000 subscription.

9. Go-to-market wedge — first 100 customers

• Ride the seminar circuit that already exists. The SME Agency, JFTC and prefectural 商工会議所 are actively running 取適法 briefing sessions — METI published the reform explainer deck itself. These rooms are full of exactly the operations managers who just learned they are liable. Offer to run the free “what the 12 fields actually are” clinic at 20 chamber events across Aichi, Osaka, and Shizuoka manufacturing belts. Attendee lists convert because the fear is fresh and the room is pre-qualified.
• The 中小企業診断士 channel. Japan’s registered SME consultants are the trusted advisor layer for exactly this customer, they are being asked about 取適法 right now, and they have no tool to recommend below enterprise TMS. Recruit 30 as referral partners on a 20% first-year revenue share. Each has a book of 10–40 SME clients.
• Free 四条書面 audit as the wedge. Ask a prospect to send 10 recent freight orders — FAX scans, emails, whatever they have. Return a scored report: which of the 12 fields are missing on each, which prices are non-compliantly bundled, what their retroactive exposure looks like. This is a 20-minute AI job that produces a document nobody can look at calmly. It is the demo and the close in one artifact.
• Carrier-side referral loop. Carriers want compliant documents from their shippers — the written terms are the carrier’s protection and the basis of their 14.6% interest claim. Give carriers a free receiving portal and let them nudge their shippers onto it. The protected party becomes the distribution channel, which is unusual and cheap.
• Targeted search on the panic query. “四条書面 書き方”, “特定運送委託 対象”, “取適法 荷主 対応” are high-intent, low-competition Japanese queries with a compliance deadline behind them.

10. Build complexity — justification

Low-to-Medium. Everything is off-the-shelf: Japanese speech-to-text, a strong multilingual model for field extraction, standard document generation, email delivery, and object storage for the 2-year archive. There is no dispatch optimisation, no route planning, no carrier marketplace, no telematics integration — deliberately, because those are what make the incumbent heavy.

The real work is statutory precision, not engineering: getting the 12 fields exactly right, encoding what counts as compliant itemisation, and handling the covered/excluded boundary (goods moving toward the ordering party are in; internal factory-to-warehouse moves are out). That needs a 行政書士 or logistics-law advisor on retainer, not more developers. FAX ingestion is the one genuinely fiddly integration.

Realistic v1 for a technical founder plus a domain advisor: 10–14 weeks.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Compliance tooling for a public statute. No licence needed to generate documents; legal advice is the line not to cross.
Ethical — no harm / dark patterns	✅	The statute exists to protect small carriers from unpaid waiting and free labour. Making shippers comply helps the weaker party.
Market exists (evidence above)	✅	32.2% doing nothing, half of sub-50-employee firms unprepared, incumbent priced for enterprise only.
1–5 person team can build this	✅	Off-the-shelf AI + document generation. No optimisation engine.
Launchable with <$50K / ₹40L	✅	Main costs are the legal advisor retainer and seminar-circuit travel.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Live statutory liability with 社名公表 as the escalation — reputational, not just financial. Felt on every shipment, so 20–500×/month. Docked because enforcement against SME shippers is still early and no one has been named yet; fear is anticipatory rather than realised.
Demand evidence	15	12/15	Hard survey data (32.2% no action, 50% of small firms, 28.8% paper/FAX) plus a funded incumbent proving willingness to pay. Docked because that data is about logistics companies broadly, and I found no direct verbatim from a shipper asking for this specific tool.
Build feasibility	15	13/15	Off-the-shelf stack, no optimisation engine, 10–14 weeks. FAX ingestion and Japanese voice edge cases are the only real friction.
Distribution clarity	15	11/15	Named, pre-qualified channels — chamber seminars, 診断士 referrals, carrier-side loop. Docked because Japanese SME software sales are relationship-led and slower than the cold-outbound math that scores 14–15 elsewhere.
Revenue mechanics	15	11/15	¥12–60K/mo is benchmarked below the ringi threshold and far below enterprise TMS. 500 customers to $1M is achievable. Docked because the incumbent’s undisclosed pricing means the true ceiling is unverified.
Time to first revenue	10	8/10	The free audit converts fast against a live obligation, but Japanese SMEs rarely sign in week one. 6–10 weeks to first paid.
Defensibility	10	5/10	Execution and statutory encoding only. Hacobu could ship an SMB tier; a domestic競合 could clone the document logic in a quarter. The accumulating 2-year evidence archive creates real switching cost by month 12 — you cannot leave the vendor holding your inspection record — but month 3 is wide open.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required

This needs someone who can ship Japanese voice/OCR extraction and who has, or can rent, real 取適法 fluency. Japanese-language capability is non-negotiable — this is not a market you enter through translation.

Key assumptions to validate (3–5)

1. Assumption: SME shippers know they are now liable and are not waiting for their carriers to complain first. How to test: Attend 3 chamber 取適法 seminars, interview 20 attendee shippers, ask what they have changed since January. If most say “nothing, and nobody’s asked,” problem intensity drops.
2. Assumption: Operations managers will accept AI-transcribed voice orders as the basis of a legal document. How to test: Run the free audit on 15 firms’ real FAX/call records; measure whether they accept the extracted fields without manual re-entry. If they re-key everything, the wedge collapses into a form.
3. Assumption: ¥25,000/mo clears without a ringi circuit. How to test: Price-test across 20 prospects; track how many need approval beyond the ops manager.
4. Assumption: Hacobu will not launch a self-serve SMB tier within 12 months. How to test: Monitor their pricing page and BOXIL listing; talk to their resellers.

Risk flags

1. Incumbent downmarket move: Hacobu already has the compliant document logic, the carrier network, and the brand. An SMB self-serve tier from them is the single most likely way this dies. Speed and the FAX/voice intake they will not want to build are the defence.
2. Enforcement softness: If the JFTC and Gメン spend 2026–27 on guidance rather than 勧告 and 社名公表, urgency decays and this becomes a vitamin. Watch for the first published company name — that is the demand accelerant.
3. Regulatory drift: The statute is seven months old. Field definitions, thresholds, and Q&A guidance will move. The product must track official interpretation or it will confidently generate non-compliant documents — which is worse than no product.
4. Market-entry difficulty: Japanese SME software distribution rewards incumbency and relationships. A foreign or unknown founder without a Japanese partner will find the seminar circuit closed.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Japanese-fluent technical founder with a 行政書士 or logistics-law advisor on retainer
Time to revenue:        6–10 weeks
Capital to launch:      ¥2–4M (~$15–27K) — mostly legal advisory retainer and seminar-circuit travel
Top 3 assumptions to validate first:
  1. SME shippers feel the liability now — 20 interviews at 3 chamber 取適法 seminars
  2. AI field-extraction from real FAX/voice orders is accepted without re-keying — free audit on 15 firms
  3. ¥25,000/mo clears without a ringi approval circuit — price-test 20 prospects
Kill criteria:
  - Abandon if <5 of 20 seminar-attendee shippers say they have changed nothing and feel exposed
  - Abandon if Hacobu or a domestic competitor launches a self-serve sub-¥40,000/mo 四条書面 tier before v1 ships
  - Abandon if free audits on 15 firms produce <3 paid conversions within 60 days

15. Next step — 1-week validation sprint

• Day 1–2: Pull the schedule of 取適法 briefing sessions from prefectural 商工会議所 and the SME Agency across the Aichi/Osaka/Shizuoka manufacturing belts. Register for three. Build a 10-question interview script aimed at one thing: what has actually changed in your freight ordering since 1 January.
• Day 3–4: Attend and interview 20 shipper-side attendees. Ask each for 10 recent freight order records — FAX scans, email threads, call notes. Offer the free audit in exchange.
• Day 5: Run the 12-field extraction over every record collected. Produce a per-firm defect report. Take it back to all 20 and ask for ¥25,000/mo for a pilot starting in 8 weeks.

Falsifiable outcome: Of 20 audited shippers, at least 5 verbally commit to a paid pilot and at least 12 show a real defect rate above 30% of fields missing or bundled. Below either number, the obligation is not being felt yet — shelve until the first 社名公表 lands and revisit.