Most foreign companies enter Japan with a translated website and a keyword list converted from English. Then they wait, rankings do not come, and the conclusion is that “Japanese SEO is different” in some mysterious way.
It is different, but not mysteriously. This guide covers what actually decides whether a foreign company ranks in Japan, in the order the decisions come up.
The one decision everything else follows from
Translated content rarely ranks in Japan. Not because translation quality is poor — often it is excellent — but because a translated page answers the English version of the question.
Three things are local, and all three break translation:
- Query patterns. The words Japanese users type are not the translations of the words English users type. They are often longer, frequently mix scripts, and cluster around different framings of the same need.
- Search intent. The same nominal keyword can carry a different intent in Japan — comparison-heavy where the English market is transactional, or vice versa.
- The evidence readers accept. Japanese readers look for different trust signals: named supervision by a qualified professional, explicit company details, specific figures over broad claims.
So the working rule is: research in Japanese, write in Japanese, and let the English site be a separate asset rather than the source text.
Step 1. Decide where the Japanese site lives
This is the decision that is expensive to reverse, so make it deliberately.
| Option | Works well when | Cost |
|---|---|---|
example.com/ja/ | You want the Japanese pages to inherit the authority of your main domain | Lowest. Usually the right default. |
example.jp or example.co.jp | Trust with Japanese buyers matters more than inherited authority — common for B2C and regulated categories | Starts from zero authority. .co.jp requires a registered Japanese entity. |
jp.example.com | Rarely the best of either | Splits authority without the local trust benefit |
For most companies entering Japan, a subfolder on the existing domain is the pragmatic choice. The .co.jp option carries genuine trust weight with Japanese buyers — it is restricted to registered Japanese companies, and Japanese users know that — but you are starting a new domain from nothing, which typically adds six to twelve months before organic traffic is meaningful.
Whichever you choose, set hreflang correctly between the English and Japanese versions. Mis-set hreflang is one of the most common technical faults we find on foreign companies’ Japanese sites, and it quietly suppresses the right page from showing to the right audience.
Step 2. Do keyword research in Japanese, not in translation
Japanese is written in three scripts that mix inside a single query, and users choose between them inconsistently. The same product can be searched as kanji, katakana, hiragana, romaji, or an English loanword — with different volumes and different intent behind each.
A practical example of the shape of the problem: a category term written in katakana may attract consumers comparing options, while the same concept in kanji attracts professionals looking for specifications. Translating your English keyword gives you exactly one of these, chosen at random.
What to actually do
- Start from Japanese suggest and related-query data, not from a translated seed list.
- Collect every script variant of your core terms and check volume for each separately.
- Read the top ten results for each target query. In Japan the SERP tells you the expected format — comparison table, ranking list, expert-supervised explainer — more reliably than any keyword metric.
- Map each query to one page. Japanese sites in competitive categories lose more to cannibalisation than to missing content.
Step 3. Produce the content natively
This is where most Japan programmes fail quietly. The page is grammatically perfect and reads like a translation, because the structure, the examples and the evidence were all chosen for a different reader.
For scale: we produce over 3,000 Japanese articles a year, and for YMYL categories — health, finance, legal — every piece is supervised by a licensed professional from a panel of more than 250 doctors, tax accountants and similar qualified experts. That supervision is not decoration. In Japanese YMYL search, named professional supervision is close to table stakes, and its absence is visible to both readers and search engines.
We cover the mechanics of Japanese content production separately in how to write SEO content for the Japanese market.
Step 4. Get the technical basics right for Japan
- Mobile. Japanese consumer search is overwhelmingly mobile. Your Japanese pages should be designed mobile-first, not adapted from a desktop layout.
- Speed from Japan. Test load times from a Japanese location, not from your head office. Origin distance and CDN configuration produce real differences here.
- Line breaking and typography. Japanese has no spaces between words. Long headings break in awkward places and read as machine output. This is a small detail that Japanese readers notice immediately.
- Structured data. Organization, Article, FAQ and — where relevant — LocalBusiness markup, with Japanese values rather than transliterated English ones.
Step 5. Build the trust signals Japanese buyers look for
E-E-A-T is not a Japan-specific concept, but the signals that satisfy it are local.
- A proper 会社概要 (company profile) page with registered name, address, representative and founding year. Sparse company pages read as untrustworthy in Japan in a way they do not in the US.
- Named authors and named supervisors, with credentials. Anonymous corporate content underperforms.
- Specific numbers with dates and sources. Japanese readers are comparison-driven and treat vague superlatives as noise.
- Consistent entity information across your site, Japanese business directories and any listings, so search engines can resolve you as one organisation.
Step 6. Plan for AI answers, not just rankings
AI Overviews launched in Japanese on 15 August 2024, and AI Mode became available in Japanese on 9 September 2025. A meaningful share of Japanese informational queries now resolves above the organic results.
Because Yahoo! JAPAN’s organic results are supplied by Google, these changes reach most of the Japanese market at once — see Google vs. Yahoo! JAPAN for how that relationship works.
We monitor AI citations weekly on our own Japanese media property. Over a recent six-month window, traffic arriving from AI assistants grew roughly 5.8x, and the property is currently cited about 14,000 times a month in AI answers. The pages that get cited share a pattern: one clear claim per section, a date, a source, and unambiguous topical scope.
Step 7. Measure it properly
- Search Console is your primary source. Because Yahoo! JAPAN runs on Google’s index, it covers the large majority of Japanese organic search.
- Segment by country and language. A global property will bury Japan.
- Track AI citation presence for your priority queries, on a schedule. Rankings alone increasingly under-report the outcome.
- Record a baseline before you change anything. Obvious, routinely skipped, and impossible to reconstruct later.
How long this takes, realistically
Japanese SEO is not faster than SEO anywhere else, and anyone promising otherwise is selling something.
One recent B2B engagement is representative of a normal good outcome: monthly clicks from Google grew from 56 to 751 over six months. That was a category with real competition, on a site that already existed. A new .co.jp domain in a crowded consumer category would be slower.
Since 2020 we have worked with 129 clients across 44 industries, all Japan-based, including Japanese subsidiaries of global companies. The pattern across them is consistent: the programmes that work commit to native production for at least two quarters before judging results.
The four ways Japan programmes usually fail
- Translating the English site and calling it Japanese SEO. The most common, and the most expensive.
- Running Japan through a global agency with no Japanese-language research capability. The deliverables look right in English and are unanchored in Japanese search data.
- Publishing anonymous corporate content in a category that requires named expertise. In YMYL this is close to fatal.
- Judging at 90 days. Japanese organic programmes rarely show their shape before two quarters.
Common questions
Do I need a .co.jp domain to rank in Japan?
No. A subfolder on your existing domain can rank well. .co.jp buys trust with Japanese buyers, not ranking ability, and it costs you the authority your main domain already has.
Can I use machine translation with human post-editing?
For documentation and support content, often yes. For content meant to rank and convert, it inherits the structure and evidence of the English original, which is the actual problem — not the wording.
Do I need to optimize separately for Yahoo! JAPAN?
Not for organic rankings — Yahoo! JAPAN’s results come from Google’s index. Its advertising, shopping and local listing systems are separate and do need separate work.
How much content do we need?
Fewer, better pages beat volume in Japanese search, particularly since AI answers began absorbing thin informational queries. Map one page per intent and resist publishing near-duplicates.
Working on Japan?
We are a small Japanese SEO practice working with companies entering the Japanese market. Research, content, entity and structured data work are all produced natively in Japanese, and we measure both rankings and citations in Japanese AI answers.
Sources
- AI Overviews Japanese launch, 15 August 2024 — 窓の杜
- Yahoo! JAPAN’s use of Google search technology and the extension to 31 March 2027 — Web担当者Forum
Last reviewed: August 2026.
