Choosing a Japan Cross-Border Fulfillment Partner: 3PL, Distributor, or Your Own Team

What This Decision Is — and What It Isn't

Three questions get collapsed into one when brands plan Japanese fulfillment, and collapsing them is what produces a partner nobody is happy with eighteen months later.

The first is where your inventory physically sits — every order shipped internationally from your existing warehouse, or a batch imported once and shipped domestically. If that is unsettled, start with Japan 3PL vs cross-border fulfillment, which lays out the landed cost, delivery speed, tax and channel-access differences and gives you a break-even to run on your own numbers.

The second is who sells to the Japanese end customer — a distributor buying your goods wholesale and reselling under its own relationships, or your own storefront under your own name. That is selling through a Japanese distributor vs running your own Japan store: margin, pricing control under the Antimonopoly Act, data ownership, and the trademark exposure of handing your brand to a counterparty.

This article is the third question, and it survives both of the others: once goods are going to sit in Japan, what kind of relationship do you want with the party that holds and ships them? A 3PL contract, a distributor taking fulfillment along with the channel, and your own leased operation allocate ownership, liability, data and exit rights in three incompatible ways. And unlike the stock-location question, this one is not settled by a break-even — it is settled by what you are willing to hand over.

If your question is specifically Amazon's warehouses versus an independent provider, that is a narrower comparison inside Model 1 below, covered in FBA vs 3PL for Japan logistics.

Three Partner Models, Defined by Who Holds Your Goods

Model 1: Contract 3PL — the goods stay yours

You import the goods, you remain the owner, and a Japanese logistics company holds them for you under a deposit arrangement — in Japanese law, 寄託, bailment. This is what most brands mean by "3PL." The inventory is still your asset; the operator has possession, not title. You instruct movements, you can withdraw the stock, and you see every order because your systems generate them.

What you take on is the importer of record obligation, the inventory risk, and a contract whose fine print is far more restrictive than most foreign brands expect — the subject of a section of its own below.

Model 2: Distributor-embedded fulfillment — the goods are no longer yours

A Japanese distributor buys your goods, imports them as its own, and fulfils orders out of its own inventory on its own systems. Fulfillment is not a service you bought; it is a consequence of having sold the goods. That is often the cleanest route when you have no capacity for Japanese operations: importer of record is solved, compliance sits with someone who does it routinely, and you carry no warehousing cost.

The costs are structural rather than invoiced. You cannot recall inventory you have already sold, you typically see the sell-through numbers the distributor chooses to share rather than order-level data, and because you are no longer the shipper none of the operational levers — delivery promise, packaging, insert, returns handling — are yours to pull. Whether that trade is worth it is a channel decision, so compare it against running your own store rather than against 3PL rate cards.

Model 3: Your own operation — you hold the lease and the payroll

You lease space, staff it, integrate a warehouse management system, and contract carriers directly. Legally this is the simplest of the three: storing your own goods in your own facility is not warehousing business, so no ministry registration applies. Operationally it is the hardest — every fixed cost arrives whether or not you sold anything that month, and you are hiring warehouse staff in a labour market where logistics companies with far better scale are hiring too.

For a cross-border brand this is almost never an entry point. It starts to make sense when volume is high enough that 3PL per-unit charges exceed the fully loaded cost of doing it yourself, when your handling is genuinely unusual — serialisation, assembly, regulated storage — and instructions to a third party keep breaking, or when fulfillment is itself part of your proposition.

The Legal Test That Sorts Real Partners From Informal Arrangements

Two registrations decide whether a company in front of you is legally equipped to do what it offers. Both are worth asking about directly, because a brand that has never imported into Japan has no other way to tell a logistics company from a company with a warehouse.

Warehousing. Article 2(2) of the Warehousing Business Act defines warehousing business as the business of storing, in a warehouse, goods received on deposit; Article 3 requires anyone carrying it on to register with the Minister of Land, Infrastructure, Transport and Tourism; and Article 11 requires a registered operator to appoint a warehouse management supervisor (倉庫管理主任者) for each warehouse, responsible for fire prevention and warehouse management. MLIT publishes a list of registered operators by building, so the claim is checkable rather than a matter of trust.

The mirror image of that rule is what makes Model 3 legally light: storing your own goods is not receiving goods on deposit, so a brand's own warehouse falls outside the registration regime entirely. It also means the middle ground brands drift into — a Japanese partner keeping your pallets in spare space as a favour or a line item — is not the same product as a registered operator's bailment service.

Transport arranging. A partner that contracts with you to move goods and then subcontracts the actual transport is doing utilisation transport. Article 3(1) of the Cargo Utilization Transport Business Act requires registration for first-class utilisation transport; second-class, which bundles road collection and delivery around a sea, air or rail line-haul, requires a licence under Article 20. "Are you registered, and under which class?" separates a logistics provider from a company that will hand your freight to whoever answers the phone.

Neither registration tells you the operator is good. Both tell you it is accountable to a regulator — a different and far cheaper thing to verify than service quality.

What a Japanese 3PL Contract Actually Promises You

Most Japanese warehousing operators do not draft bespoke storage terms. Under Article 8 of the Warehousing Business Act an operator must file its bailment terms with the minister — but one that adopts terms identical to the ministry's standard terms is deemed to have filed. So the Standard Warehouse Bailment Terms are, in effect, the industry's default contract, and they were substantially revised with effect from 1 April 2026 — the first large-scale revision since the original 1959 terms. Four provisions change how you should price the relationship; references are to the Type B terms now in force.

• Liability is limited to intent or gross negligence, and you carry the burden of proof. Article 39 limits the operator's liability to losses caused by its own or its employees' intent or gross negligence, and requires the party claiming compensation to prove it. Ordinary carelessness is not a compensable event under the default terms — and proving gross negligence inside someone else's warehouse is not a position you want to be in.
• Consequential loss is now capped at the fees you have paid. Article 43(1) assesses compensation for loss of or damage to the goods at market value, capped at the fire insurance sum or the declared bailment value. Article 43(2), new in the 2026 revision, caps compensation for any other loss at the total fees already charged on those goods — the ministry's explanatory material is explicit that claims had broadened to matters such as lost sales from goods released in the wrong expiry order, at amounts disproportionate to the storage fee. The cost of a fulfillment error that ruins a launch window sits with you, not the warehouse.
• Fire cover is only as good as the value you declared. Article 33 has the operator insure the goods against fire on the depositor's behalf unless told otherwise, and Article 34 sets the insured sum at the declared bailment value. Article 41(2) excludes fire damage beyond the settlement amount determined under Article 35, and excludes it entirely where the depositor asked not to be insured. Declaring a low bailment value to save a premium is a decision about your own downside.
• The value-added work you assumed was included is chargeable. Article 4, also new in 2026, defines manual handling inside the vehicle, sorting, full or unpacking inspection and label application as ancillary services the operator may charge for separately, and lets it charge for inbound or outbound instructions issued without adequate lead time. If your operation depends on relabelling, kitting or same-day cut-offs, that is a priced line item, not goodwill.

Two further points are worth reading in your own provider's terms: Article 42 makes clear the operator does not inspect contents, so a discrepancy between carton and paperwork is not its problem; and Article 45 extinguishes liability once you or your agent take the goods back without reservation. None of this makes a Japanese 3PL a bad deal — it makes it a deal whose residual risk belongs in your own cargo and stock insurance, the same conclusion we reach from the freight side in cargo insurance for cross-border import into Japan.

What Building Your Own Actually Costs

Two current numbers frame Model 3 honestly.

Space. CBRE's Japan logistics market view for the second quarter of 2026, released 29 July 2026, puts vacancy for large multi-tenant logistics facilities in Greater Tokyo at 7.8% with effective rent at ¥4,600 per tsubo, up 1.5% on the quarter; Kinki stood at 2.4% vacancy and ¥4,390 per tsubo. Those are the modern facilities a serious own-operation would want, leased at a scale and term an early-stage cross-border brand cannot absorb — which is precisely why the 3PL model exists. Rents rising in all four regions CBRE tracks, with Greater Tokyo vacancy falling, is also the wrong moment to take a first Japanese lease.

Labour. Japan's regional minimum wages for FY2026, as recommended by the prefectural councils in September 2026, give a national weighted average of ¥1,177 per hour, up from ¥1,121, scheduled to take effect in stages between 1 October and 2 December 2026. That is a floor, not a warehouse wage, and it has risen every year recently — so a build-your-own model priced on today's payroll is priced against a rising baseline.

There is also a staffing trap specific to foreign brands that want control. Engage a contractor to supply warehouse labour but direct those workers yourself day to day, and you are not buying a contracted service — you are running disguised worker dispatch. Whether an arrangement is dispatch or a genuine subcontract is judged on substance under the 1986 Ministry of Labour notice that sets out the distinction, not on what the contract is called, and getting it wrong breaches the Worker Dispatch Act. Employ the people, or let the contractor manage its own staff. Wanting both is the failure mode.

The Obligations That Follow the Goods, Not the Contract

One shift in Japanese logistics regulation cuts across all three models and is barely covered in English. The revised Act on Promotion of Logistics Efficiency imposes, from FY2025, duties of effort on all shippers and logistics operators to make logistics more efficient — reducing waiting and loading times, improving load rates. From FY2026, entities above a size threshold are designated as specified entities, must submit a medium-to-long-term plan and periodic reports, and specified shippers must appoint a logistics officer (物流統括管理者, or CLO) from among those in a managerial position participating in important business decisions. The thresholds are 90,000 tonnes handled annually for specified shippers, 150 vehicles for specified trucking operators and 700,000 tonnes stored for specified warehousing operators.

Those thresholds are far out of an ecommerce brand's reach, so the reporting regime is not your problem. Two things about it are. If you hold inventory in Japan and instruct its movement, you are a shipper carrying the duty of effort — so same-day outbound instructions, or collections that leave trucks waiting, are now something your partner can point at a statute about. And in Model 2 you are not the shipper at all; the distributor is. The obligation follows the goods and the instruction, not the invoice.

Behind all of it is the capacity constraint Japanese logistics has absorbed since the annual overtime cap of 960 hours took effect for truck drivers in April 2024. Whatever model you choose, domestic delivery capacity is tighter and more expensive than it was — and a partner with volume negotiates carrier rates you cannot.

Side by Side

How to Decide: Five Questions

1. Who do you want to own the Japanese customer relationship? If the answer is you, Model 2 is already out whatever its convenience — you cannot own the relationship while someone else owns the goods and the order record. If you are content to be a supplier to Japan, it is the cheapest way to be one.
2. How many channels does fulfillment have to serve? A single Amazon.co.jp channel narrows the question. Rakuten Ichiba, Yahoo! Shopping and your own Shopify store together need a partner that ships to all of them on their respective service expectations — in practice a contract 3PL with marketplace integrations, not the cheapest storage quote.
3. What does a bad week cost you? Compare the commercial loss from a week of mis-picks or a missed launch against what Article 43(2) will pay: the fees charged on those goods. If the gap is large, the answer is insurance and contractual variation negotiated up front, not a better provider.
4. Is your handling genuinely unusual? Serialisation, cold chain, regulated storage, assembly-to-order and heavy relabelling all raise the cost of instructing a third party and are the honest arguments for Model 3. Ordinary pick-and-pack is not.
5. How long is your Japan commitment? A two-year test and a ten-year build justify different fixed costs. A lease and a warehouse payroll are the least reversible commitments here; a bailment is the most reversible.

A Vetting Checklist for Any Japanese Fulfillment Partner

• Ask for the registrations. Warehousing registration, plus first- or second-class utilisation transport if they arrange freight. Check the warehousing claim against MLIT's published list of registered operators.
• Ask for their filed bailment terms, not a summary. Identical to the ministry standard terms means you know exactly what you are getting; any deviation is the negotiation.
• Settle the declared bailment value. It sets the fire insurance sum and caps compensation. Agree how it is reviewed as your stock value changes.
• Price the ancillary work explicitly. Labelling, full inspection, kitting and short-notice orders are chargeable under the 2026 terms. Get them on the rate card before launch, not after the first campaign.
• Confirm who is importer of record and whether import consumption tax is recoverable by you — see whether you need a Japanese entity to sell ecommerce, and the boundary between an operations agency and a licensed customs broker in what an agency can and cannot do at customs.
• Agree the returns process in the same contract. Receiving, inspecting, restocking and disposing of returns is a separate scope with its own rates; the options are in Japan ecommerce returns: outsource or handle in-house.
• Ask what happens on exit. Notice period, withdrawal charges, data handover, and who pays to move pallets to the next operator.

Common Mistakes

• Comparing rate cards across models. A distributor's wholesale discount and a 3PL's per-order fee are not the same kind of number. One buys market access and risk transfer; the other buys pick and pack.
• Assuming the warehouse insures your downside. Under the standard terms it insures fire at the value you declared, pays for lost or damaged goods only on proven intent or gross negligence, and caps everything else at the fees you have paid.
• Treating labelling and inspection as included. Since April 2026 they are defined ancillary services the operator may charge for — as are your same-day instructions.
• Letting the partner be importer of record for convenience. It simplifies the first shipment and can make import consumption tax unrecoverable in your own filings.
• Building your own to get control, then directing contracted staff. That is disguised dispatch. Employ the people or buy the service; not both at once.
• Choosing before the stock-location decision is settled. If you still ship every order from abroad, none of these models is live yet — our cross-border ecommerce service covers the route that needs no Japanese stock.

Where LAUNOVA Fits

We state the boundary plainly, because vagueness here costs brands money. LAUNOVA runs Japanese storefront and marketplace operations for overseas brands across Amazon Japan, Rakuten Ichiba, Yahoo! Shopping and Shopify. We are not a warehousing operator, a utilisation transport operator, a customs broker, a distributor or an importer of record. We do not hold, pick, pack or ship your goods, and we do not source or negotiate distributor relationships.

What we do here is the part next to the channel: reading your channel plan for the constraints that settle the model before cost does, briefing candidate partners in Japanese against the questions above, checking the arrangement can serve every channel you sell on, and running the storefront side once stock is landed — which is where our Japan ecommerce operation agency service sits. Contracts, liability terms, customs status and insurance belong with a Japanese lawyer, a licensed customs broker and your insurer, and we will say so rather than improvise. Tell us your category, volume and channel mix and we will tell you which model your situation supports. Pricing is quoted against the work, not published as a rate card.