The translation question is usually too late.
When a SaaS company starts preparing for Japan, the first operational question is often: “Who should translate the website?” A better first question is whether the current website deserves to be translated as it is.
Language is only one layer of a market-facing decision. A page can be fluent Japanese and still ask a Japanese buyer to make the wrong comparison, trust the wrong proof, or act before they have enough certainty.
1. What category will a Japanese buyer put us in?
Your global category label may be clear to people who already know the space. In Japan, the same label may be unfamiliar, overlap with an older category, or create expectations your product does not meet.
Before translating the hero, decide what the buyer should compare you with. If you cannot answer that clearly, the wording is not ready.
2. Which problem should lead?
The problem that wins attention in your strongest market is not automatically the problem that should lead in Japan. A global site may lead with speed, disruption or autonomy. A Japanese B2B buyer may need the first screen to make the operational job, implementation fit or risk reduction more concrete.
This is not about making the copy “more Japanese.” It is about choosing the most useful starting point for the buying decision.
3. What proof does the buyer need before the CTA?
Many localization projects preserve the global page order: proposition, features, CTA, then trust. That order may be correct. It may also be too optimistic for an unfamiliar entrant.
Ask what must be believed before someone will request a demo, start a trial or share the product internally. Depending on the product, that may include security, administrative control, integrations, customer evidence, local support or clear company information.
4. Which parts should stay global?
Localization is not a license to rewrite everything. Strong product language, distinctive proof and brand character should stay when they still help the buyer understand and trust the product.
The useful decision is not “global or local.” It is: what is already doing its job, and what stops doing its job when the market changes?
5. Which parts need more explanation?
English-language SaaS sites often benefit from compression. A few words can carry a lot because the category, product pattern and buying motion are familiar.
In a new market, the same compression can become ambiguity. If the buyer has to infer what the product replaces, how it is controlled or where it fits, the Japanese page may need more context even if the global page does not.
6. What should the first CTA ask for?
“Start free” and “Book a demo” are not translation problems. They are confidence-level decisions.
If the visitor still needs to understand the product category or evaluate risk, a lower-commitment CTA such as “See how it works” may deserve more prominence. If the product truly supports self-serve onboarding and Japanese buyers are ready to act, the global CTA may remain exactly right.
The point is to review the hierarchy, not to replace the CTA by rule.
7. What should we not translate yet?
A full-site translation feels complete, but completeness is not the same as priority. Start with the pages that carry the buying decision: homepage, product or solution page, pricing, trust, company information, FAQ and the conversion path that matters most.
Pages that exist mainly for content volume can wait until the positioning and category language are stable. Otherwise, you risk producing more Japanese around a message you later decide to change.
The practical sequence
Before localization begins, settle the category, the lead value proposition, the proof order and the CTA hierarchy. Then decide which pages deserve production first.
Translation becomes easier after those choices because the team is no longer asking Japanese to solve a strategy problem.