Localization is not translation: what actually has to change for Japan

A translated website is table stakes. It is also, on its own, close to worthless as a predictor of whether a foreign product will sell in Japan.

We have watched companies spend six figures on translation and creative adaptation, launch, and convert at a fraction of their home-market rate. The copy was not the problem. The problem was everywhere else.

Here is what actually has to change, ordered by how much it moves revenue.

1. Proof, and who is allowed to provide it

In the US, a strong claim plus a free trial is often enough to start a purchase. In Japan, an unsupported claim from an unknown foreign company is a reason to disengage, not to evaluate.

What substitutes for that trust:

• Named Japanese customers. One recognizable domestic logo outperforms twenty foreign ones. Companies routinely ask "which Japanese companies use this?" as the first qualifying question, and "none yet" ends a meaningful number of conversations.
• A named, reachable local counterpart. A Japanese contact with a Japanese phone number and a real title. Not a support address routed to another timezone.
• Specificity over superlatives. "Reduces processing time by 34% across 12 deployments" travels. "The world's leading platform" does not — it reads as unverified and slightly suspect.

This is why the first Japanese reference customer is disproportionately valuable, and why it is worth taking a bad commercial deal to get one.

2. The sales process, not the sales pitch

The standard Western B2B motion — discovery call, demo, proposal, close with a single decision-maker — maps badly onto Japanese enterprise buying.

What is different:

• Consensus over authority. The person you are talking to is frequently not deciding; they are assembling internal agreement. Material that helps them sell internally (a Japanese-language one-pager they can circulate, a risk-and-mitigation summary) does more than a better demo.
• Longer cycles, higher retention. Deals take longer to close and then churn less. Forecasting on home-market cycle lengths produces a pipeline that looks broken when it is merely Japanese.
• Written follow-up matters. A detailed written summary after a meeting is expected, and its absence reads as disorganisation.

3. Pricing and how money moves

Price points transfer badly. So does the payment mechanism.

• Consumption tax is 10% and buyers expect to see tax-inclusive or clearly tax-separated pricing. Ambiguity here creates friction at exactly the wrong moment.
• Invoice-and-bank-transfer remains standard in B2B. Credit-card-only checkout will lose enterprise deals outright. Many companies pay on a fixed monthly closing cycle, which affects your cash flow assumptions.
• Consumer payments fragment differently. Convenience store payment, carrier billing and QR-code wallets carry real share alongside cards. A checkout that only accepts international credit cards silently loses conversions.
• Annual-prepay discounting is less effective where budgets are set on a fiscal year starting in April. Aligning your contract cycle to the Japanese fiscal year is a small change with a measurable effect on close rates.

4. Support expectations

The service baseline is higher, and it is not negotiable through better product.

Response-time expectations are shorter, the tolerance for a support agent who cannot resolve the issue is lower, and machine-translated support replies are immediately identifiable and damaging. If you cannot staff Japanese-language support at launch, the honest options are to limit the segment you sell to, or to route support through a partner who can.

5. The product itself — usually less than you think

This is the one companies over-invest in. Genuine product changes that matter:

• Name and address handling. Family name first, no middle name, postal codes in the 〒000-0000 format, prefecture as a required field, and address lines that do not assume Western ordering.
• Character handling. Full-width and half-width characters, kana input, and text expansion in UI — Japanese strings frequently break layouts designed around English.
• Date, era and fiscal conventions. Year-month-day ordering, and the Japanese imperial era appearing in official documents.
• Vertical text and print output, if you produce documents that will be printed or filed.

What usually does not need to change: the core feature set. Companies convince themselves Japan needs a different product when what it needs is different proof and a different sales process.

6. Creative, last

Creative adaptation matters, and it is genuinely different work from translation — density of information, the role of illustration, how directly a claim can be stated. But it sits last on this list because it amplifies whatever the first five items have established. Excellent Japanese creative in front of a product with no local proof, no local payment method and no local support converts poorly.

The sequencing that works

In practice we run this order: proof → sales process → payments → support → product details → creative. Each step is cheap relative to the one before it in wasted spend if skipped.

The test of whether you have localized, rather than translated, is simple: could a Japanese buyer complete a purchase, get a question answered, and justify the decision internally — entirely in Japanese, without ever touching your home organization? Until that is true, you have a translated website.

If you want an assessment of where your product currently sits against that list, get in touch.