Japan market entry: the five routes in, and what each one actually costs

Every foreign company entering Japan picks one of five structures, whether or not they know they are picking. The choice determines how fast you learn, how much you spend before you learn it, and how hard it is to reverse when you are wrong.

Most of the published advice compares these routes on cost and control. That framing is a trap. The variable that matters at the start is commitment before evidence — how much capital and how many months you have to put in before the market tells you whether it wants your product.

The five routes

1. Direct export

You ship from your existing operation and sell to Japanese buyers who come to you. No entity, no local staff, no fixed cost.

This works when demand already exists and is finding you — inbound enquiries, a distributor knocking on your door, orders through a global marketplace. It stops working the moment you need to create demand, because you have no one who can pick up the phone in Japanese, negotiate a retail listing, or run a campaign that lands.

Commitment before evidence: almost none. Ceiling: low.

2. Distributor or agent

You appoint a Japanese company to sell on your behalf. They know the channel, they carry the inventory risk, and they absorb the language and relationship work.

The cost is not the margin — it is the loss of the feedback loop. Your distributor tells you what sold. They do not tell you why the deals you lost were lost, what objection kept coming up, or which segment would have converted with a different price. You also inherit their priorities: your product sits in a portfolio, and it gets attention proportional to its contribution to their revenue, not yours.

Exclusivity terms are where this route goes wrong most often. A three-year exclusive signed before you understand the market can lock you out of your own territory.

Commitment before evidence: low financially, high contractually. Ceiling: capped by your partner's incentives.

3. Branch office (支店)

A registered presence of the foreign parent. Faster and cheaper to establish than a subsidiary, and it can conduct business and open a bank account.

The catch is liability: a branch is not a separate legal person, so the parent carries the obligations directly. Many Japanese counterparties — landlords, large corporate customers, some banks — treat a branch as less established than a locally incorporated company, which shows up as friction in exactly the negotiations you need to win.

Commitment before evidence: moderate. Ceiling: medium, with counterparty friction.

4. Subsidiary — KK or GK

A Japanese company that you own. This is the endpoint most companies eventually reach, and it is the correct structure once Japan is a real revenue line.

It is also the most common premature decision. Incorporating brings a registered address, a resident representative, corporate tax filings, social insurance enrollment, statutory accounting, and — once you hire — Japanese employment law, under which terminating a permanent employee is materially harder than in the US or UK. You are taking on a fixed cost base and a compliance calendar to test a hypothesis.

The honest sequence is: prove demand, then incorporate to capture it. Not the reverse.

Commitment before evidence: high, and slow to unwind. Ceiling: unlimited.

5. Outsourced go-to-market

A local team operates the Japanese market under your direction, using their entity, their staff and their relationships. You get execution in Japanese — negotiation, campaigns, marketplace operations, events, retail conversations — without incorporating or hiring.

This is the structure we operate, so treat the recommendation with appropriate scepticism. The reason it exists is specific: it is the only route that gives you real execution capacity while keeping commitment reversible. If Japan does not work, you stop. If it does, you have a proven playbook and a set of live relationships to hand to your first local hires.

It is not a permanent answer. It is the answer for the period between "we think Japan might work" and "Japan is a line on the board pack."

Commitment before evidence: low. Ceiling: medium-high, and designed to be handed over.

The comparison that matters

Route	Time to first revenue	Reversible?	Local execution capacity
Direct export	Weeks	Yes	None
Distributor	2–6 months	Contract-dependent	Theirs, not yours
Branch	2–3 months	Moderate	Only what you staff
Subsidiary	3–6 months	Slow and costly	Only what you staff
Outsourced GTM	Weeks	Yes	Full, from day one

How to actually choose

Ask three questions in order.

Does demand already exist and find you? If yes, direct export and stop reading. Do not build infrastructure for a problem you do not have.

Do you need to create demand, and is the outcome uncertain? Then the priority is learning speed at low commitment. That is outsourced GTM, or a non-exclusive distributor arrangement with a hard review date.

Is Japan already producing revenue you can forecast? Incorporate. At that point the fixed cost is justified, the compliance load is worth carrying, and permanent local employees will outperform any external arrangement.

The failure mode we see most often is companies answering question three's answer to question two — incorporating and hiring a country manager to find out whether Japan works. Eighteen months and a seven-figure spend later, the answer arrives, and it is frequently that the offer needed repositioning, which a three-month test would have shown.

What this does not cover

Regulated sectors — financial services, medical devices, pharmaceuticals, food and alcohol — have licensing requirements that can force an early entity decision regardless of commercial logic. If you are in one of those, the sequencing above still applies to the commercial test, but the regulatory path runs in parallel and starts earlier.

If you want a view on which route fits your specific product and stage, tell us what you sell and we will give you a straight answer, including when that answer is "not yet."