Japan SEO

TL;DR — Search-engine usage and partnerships in Japan can change, so verify current market data before prioritizing platforms. The durable difference is the language: the same word can be written in several scripts (kanji, hiragana, katakana, or the Latin alphabet), and each version can be a separate keyword. And you need real localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native., not a quick translation.

Japan isn’t a “non-Google” market

If you’ve read our market-specific SEO hub, you know that some countries run on their own engines — China on Baidu, Russia on Yandex, South Korea on Naver. It’s natural to assume Japan belongs on that list because of Yahoo! Japan. It doesn’t.

Here’s the key fact: Yahoo! Japan’s organic search results are generated by Google’s technology, and have been since a 2010 agreement. Yahoo! Japan is a huge, popular portal — news, shopping, auctions, a Q&A service — but its actual web search results come from Google under the hood. So when someone tells you “you need to do Yahoo! Japan SEOJapan SEO is the practice of optimizing sites to rank in Japan — a market where Google leads and Yahoo! Japan has run on Google's search technology since 2010. The distinctive work is Japanese-language keyword research across kanji, hiragana, katakana, and romaji, plus genuine localization rather than translation. separately,” that advice is out of date. If you rank on Google in Japan, you generally rank on Yahoo! Japan too.

So what is different about Japan?

Two things, and neither is a separate algorithm:

1. The Japanese writing system. Japanese mixes four ways of writing:

• Kanji (漢字) — characters borrowed from Chinese (e.g. 寿司 for sushi).
• Hiragana (ひらがな) — a phonetic script for native words and grammar.
• Katakana (カタカナ) — a phonetic script mostly for foreign/loan words and brand names (e.g. コーヒー for “coffee”).
• Rōmaji — the Latin alphabet, used for some brand names and tech terms.

The same idea can be typed in more than one of these, and those versions aren’t always interchangeable. “Coffee shop” can be カフェ (a Western-style cafe) or 喫茶店 (a traditional Japanese coffee house) — different scripts, genuinely different meaning. That’s why Japanese keyword research means checking each variant separately, not picking one and moving on.

2. LocalizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. over translation. Machine-translating your English page into Japanese usually isn’t enough — and often reads as unnatural. Real localization adapts payment methods (Japanese shoppers expect options like convenience-store payment), page design (Japanese sites tend to be denser and more text-heavy than Western ones), and legal/trust details. There’s even a specific legal disclosure page (the “Tokushoho” page) that e-commerce sites selling to Japanese consumers are expected to publish. Our translation vs. localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. guide goes deeper on that distinction.

One thing to not get confused by

If you search “Japanese keyword hack,” you’ll find a lot of pages — but that’s a hacking topic, not a keyword-research one. It’s when attackers break into a site (often an outdated CMSA content management system (CMS) is software that lets users create, manage, and publish digital content — like blog posts and pages — without writing raw code. WordPress, Drupal, and Joomla are the most common open-source CMS platforms.) and inject spammy Japanese pages. If you foundA 302 (\"Found\") is a temporary redirect: it forwards users to a new URL while telling search engines the original URL should stay in the index. It's a weak canonicalization signal, not the zero-equity dead end of SEO folklore. that term because weird Japanese pages showed up on your own site, jump to the Advanced tab — there’s a section on cleaning it up.

Want the full version — the 2010 deal and why it still holds, the four-script keyword workflow, the current market-share numbers (Bing is surprisingly large in Japan now), and the localization checklist? Switch to the Advanced tab.

TL;DR — Treat search-engine share and syndication arrangements in Japan as time-sensitive inputs, not permanent assumptions. The durable Japan-specific work is (1) keyword research across kanji / hiragana / katakana / romaji, where variants carry different volumes and sometimes different intent, and (2) genuine localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. (payment, legal disclosures, page conventions), not translation. The “Japanese keyword hack” is a site-compromise attack, covered as an aside. As a For market-share claims, cite the measurement source and date and avoid assigning a cause the data does not establish.

Japan’s search landscape: Google, Yahoo! Japan, and the deal that ties them

The single most load-bearing fact for Japan SEOJapan SEO is the practice of optimizing sites to rank in Japan — a market where Google leads and Yahoo! Japan has run on Google's search technology since 2010. The distinctive work is Japanese-language keyword research across kanji, hiragana, katakana, and romaji, plus genuine localization rather than translation. is that Yahoo! Japan runs on Google’s search technology. This is not folklore — it’s documented from several independent angles.

The 2010 deal itself: contemporaneous tech press reported that “Yahoo search will be powered by Google in the future” and that “users of the service will be served both paid and algorithmic search results generated by Google technology” (TechCrunch, July 27, 2010). Google’s own announcement was published the same day on the Google Japan Blog, authored by Daniel Alegre, then Google’s VP for Asia Pacific and Japan (Google Japan Blog, July 27, 2010) — it’s in Japanese, so I’m paraphrasing rather than quoting it in English: it confirms Google had previously supplied search technology to Yahoo! Japan from 2001 to 2004, and that Yahoo! Japan’s auction, shopping, and Chiebukuro (Q&A) content would feed into Google’s Japanese search indexStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed.. On the record from Yahoo! Japan’s side, its then-CEO explained the choice of Google over Microsoft plainly: “We had many discussions with Yahoo Inc. and in the end, they recognized that this was the right strategy for us in Japan” (Masahiro Inoue, quoted in Fortune, July 27, 2010).

Is it still true? Yes — and here’s the part I like, because it’s a government document rather than a blog post. In 2024 the Japan Fair Trade Commission approved a commitment plan from Google that describes Google having “restricted the provision of technologies for search engines and search advertising to Yahoo” from September 2, 2015 through October 31, 2022 (JFTC, April 22, 2024). The same filing pins the original Google Services Agreement to July 27, 2010 (matching the press date exactly) and an amendment to November 1, 2014. So we have official, dated proof the arrangement was still operative into late 2022 — and the remedy actually preserves Yahoo/LINE Yahoo’s ongoing access to that technology. Read the JFTC action carefully, though: it’s not “Google was punished for powering Yahoo! Japan.” It’s the opposite direction — Google was foundA 302 (\"Found\") is a temporary redirect: it forwards users to a new URL while telling search engines the original URL should stay in the index. It's a weak canonicalization signal, not the zero-equity dead end of SEO folklore. to have limited how much of that shared technology Yahoo could use to compete in mobile search-ad syndication, and the filing explicitly notes approval “does not mean that Google LLC’s conduct violated the Act.” Even as recently as 2025, US Department of Justice lawyers cited the 2010 Yahoo! Japan arrangement in Google’s US antitrust remedies case as an example of Google having voluntarily shared data before — with the useful precision, per Bloomberg Law’s reporting, that it applied only to Google’s Japanese-language index, not its global index.

There’s a more current — and more directly useful — answer to “is the deal still active” than the JFTC filing, too. At LINE Yahoo’s FY2023 full-year earnings briefing in May 2024, the company confirmed it had extended its commercial search- technology supply contract with Google by two years. The original contract was due to expire at the end of March 2025; LINE Yahoo had reportedly been evaluating a switch to Naver’s search technology — Naver co-owns LY Corporation’s parent via the SoftBank/Naver joint venture described above — as the contract wound down, but shelved that plan because it couldn’t resolve data-security and personal-information-handling concerns. The result, per contemporaneous Japanese business-press coverage citing Nikkei: “Yahoo! JAPANの検索サービスは、少なくとも2027年3月末まではグーグルから技術供与される” (“Yahoo! JAPAN’s search service will continue to receive technology from Google through at least the end of March 2027”) (Web担当者Forum, May 17, 2024, citing 日本経済新聞/Nikkei). As of this pass (mid-2026) I found no newer reporting of a further renewal, change, or termination — the March 2027 floor is still the most current commercial-contract fact available. (This is a separate data point from the JFTC filing above: the JFTC filing documents a restriction period Google was found to have imposed within an otherwise-ongoing relationship; this earnings disclosure documents the underlying commercial contract’s term. Both are true and they reinforce the same conclusion from different angles.)

A few clarifications that matter for accuracy:

• This is not the US Yahoo. yahoo.com in the US runs on Bing. Yahoo! Japan is a separate corporate lineage, today a brand operated by LY Corporation (renamed from Z Holdings in October 2023, itself renamed from Yahoo! Japan Corporation in 2021). LY Corporation is majority-owned via A Holdings, a 50/50 joint venture between SoftBank and Naver Corporation — the same Naver behind Korea’s NaverNaver SEO is optimizing for Naver, South Korea's dominant portal-style search engine. Naver favors content published on its own properties (Naver Blog, Cafe, Knowledge iN) and ranks it with its own C-Rank and D.I.A. systems, so the work is as much about getting inside the ecosystem as it is about a standalone website.. That’s a genuinely odd connective-tissue fact: the company running Yahoo! Japan is half-owned by the parent of Korea’s dominant search portal.
• It was a hybrid, not a pure mirror. Google generates the core web results; Yahoo! Japan wraps them in its own portal shell with its own properties (News, Chiebukuro, Shopping) layered in. And reportedly, LINE’s in-app search — LINE being the messaging app most of Japan uses — was switched to Yahoo! Japan Search after the merger (industry-sourced, not first-party-confirmed). If accurate, all three of Japan’s dominant consumer search surfaces resolve to the same Google-powered results.

Current market share (dated — and Bing is the surprise)

Per StatCounter’s June 2026 snapshot for Japan: Google 59.04%, Bing 33.1%, Yahoo! 6.23%, DuckDuckGo 0.72%, Yandex 0.33%, CocCoc 0.31% (StatCounter Global Stats). Two things to flag honestly here. First, the historical reversal: in 2010, per Fortune, Yahoo! Japan alone held ~57% of Japan’s search queries and Google ~38% — by 2026 those roles have almost completely flipped, consistent with users increasingly reaching Google directly rather than through the Yahoo! Japan wrapper. Second, Bing at roughly a third of the market is genuinely surprising and much higher than the ~5-10% Bing share Japan was historically reported at. It’s real in StatCounter’s page-view-weighted data, but I’d treat the size of the shift as noteworthy and the cause as speculation — StatCounter doesn’t publish a causal breakdown, and its methodology skews with device/browser defaults. Plausibly it reflects Windows/Edge default-browser behavior post-Copilot, but I’m not going to assert that as fact. And Yahoo! Japan remaining at ~6% of search while still being the #2 site in Japan by traffic (Similarweb, May 2026) captures the whole story: it’s a giant portal, a small standalone search engine.

The practical upshot: your Japan organic strategy is a Google strategy. There’s no separate Yahoo! Japan organic-ranking submission process — no Yahoo!-specific webmaster-guidelines portal exists, because there’s no separate index to submit to. Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. and general Google technical SEOTechnical SEO is the practice of making a site easy for search engines to crawl, render, index, and (now) be eligible for AI answers. It's the foundation that lets your content and links rank — not a ranking trick of its own. cover it.

Japanese keyword research: kanji, hiragana, katakana, romaji

This is where Japan SEO stops looking like Google SEO and starts requiring market-specific work. Japanese has four ways of renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM. words, and a single concept can legitimately be searched in more than one of them:

• Kanji (漢字) — logographic characters borrowed from Chinese. Skews formal/professional; common for established native concepts.
• Hiragana (ひらがな) — a phonetic syllabary for native words and grammar. Often softer/more informal.
• Katakana (カタカナ) — a phonetic syllabary for loanwords, foreign brand names, and tech terms. If your brand has an English name, Japanese searchers will likely type it in katakana.
• Rōmaji — Latin-alphabet transliteration. The smallest slice, but real — notably for IT/tech queries and brand names without an established katakana renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM.. (No reliable quantified share of romaji queries exists, so I’ll leave it qualitative rather than invent a number.)

There’s also a genuine tokenizationA token is the smallest unit of text (or image/audio/video) an LLM processes — roughly 4 characters, or about ¾ of an English word. A context window is the maximum number of tokens (input plus output) a model can hold at once, like its short-term memory. wrinkle underneath this: Japanese doesn’t use spaces between words, so search engines rely on dictionary-based morphological segmenters rather than splitting on whitespace. That’s part of why “exact match” keyword thinking translates poorly to Japanese — the engine is segmenting your content and the query, not matching space-delimited strings.

Why the same concept has different volumes by script

The variants are separate query stringsThe `?key=value` data tacked onto the end of a URL after a question mark — used for tracking, sessions, filtering, sorting, and search — and one of the biggest sources of duplicate URLs and wasted crawling in SEO., and keyword tools (and Google itself) treat them that way. Volumes — and sometimes intent — differ. Three examples that make the point:

• Coffee shop — the lead example, because the script tracks a real semantic distinction, not just spelling. カフェ (kafe, katakana) means a Western-style cafe; 喫茶店 (kissaten, kanji) means a traditional Japanese coffee house; コーヒーショップ (kōhī shoppu, katakana) is the literal “coffee shop.” Optimizing for one and ignoring the others can miss a genuinely different searcher.
• Mobile phone — 携帯電話 (kanji, formal/full form), 携帯 (kanji, abbreviated), and ケータイ (katakana rendering of the abbreviation) can each pull different volumes for what a Western SEO would call “one keyword.”
• Sushi — a single concept with 3+ valid renderings: すし (hiragana), スシ (katakana), 寿司 (kanji, most common). It shows the range can exceed the basic three-script framing.

An observed pattern worth stating carefully — as a pattern, not a rule: katakana tends to correlate with commercial intent for Western/international products, hiragana skews informational/native, and kanji skews formal/professional/B2B. This is industry consensus rather than a single authoritative source, so treat it as a useful heuristic, not law.

The practical workflow

1. Start from the concept, not the string. List every plausible kanji, hiragana, katakana, and (where relevant) romaji rendering — for brand/foreign terms, generate multiple katakana transliterations, since more than one can exist.
2. Pull volume for each variant separately in your keyword tool (Ahrefs Keywords Explorer, etc.) rather than assuming they consolidate.
3. Compare not just volume but the SERP for each — if the ranking pages differ, the intent differs, and you may need distinct content.
4. Have a native Japanese speaker sanity-check the shortlist. Tools tell you volume; only a native reader reliably catches which variant sounds natural, spammy, or subtly wrong for your audience.

Localization vs. translation for Japan

Japan is the standard textbook case of “translation is necessary but not sufficient.” A linguistically-correct machine translation can still read as unnatural, and — worse — be non-compliant or untrustworthy. Some concrete localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. work that translation alone never surfaces:

• The Tokushoho disclosure page. Japan’s Act on Specified Commercial Transactions (特定商取引法) governs how e-commerce sites must disclose to Japanese consumers (Japanese Law Translation, the government’s official English text). In practice, Japanese e-commerce sites publish a dedicated disclosure page (conventionally titled 特定商取引法に基づく表記) linked from the footer, covering the business operator’s identity, address, and return policy. Precise wording: this is a widely-followed compliance pattern built on the Act’s advertisement-disclosure requirements and the associated ministerial ordinance — not a single statutory sentence that literally says “publish one page.” But its absence is a well-known red flag to Japanese shoppers, so it’s a real trust signal regardless of how you characterize the statute.
• Payment expectations. Convenience-store (konbini) payment and bank transfer are common expectations, alongside cards — a market convention worth accommodating, though I’d present it as widely-observed rather than pinned to one citation.
• Design conventions. Japanese homepages tend to be denser and more text-heavy than Western minimalism; that’s a real UX expectation, not a mistake to “fix.”

On the pure-SEO side, Google’s guidance here is generic international-SEO guidance — Google doesn’t publish Japan-specific ranking docs, which is itself an accurate framing point: Japan falls under the same international/multilingual/hreflangHreflang is an annotation (in HTML, HTTP headers, or XML sitemaps) that tells search engines which language and optional region a page targets, and which alternate versions exist. It only works when every page in the cluster references all the others. framework as every other market. The one directly-relevant rule for localized sites: “Localized versions of a page are only considered duplicates if the main content of the page remains untranslated” (Google Search Central) — i.e. translating only your navigation/footer while leaving the body in English can get the page treated as a duplicate. Use hreflangHreflang is an annotation (in HTML, HTTP headers, or XML sitemaps) that tells search engines which language and optional region a page targets, and which alternate versions exist. It only works when every page in the cluster references all the others. to map your ja/ja-JP variants like any other international site.

Japanese search behavior

Two habits shape Japanese SERPs. First, it’s a mobile-first, app-centric market — LINE dominates messaging, and portal apps are heavily used. (Mobile-first indexingGoogle's practice of using the mobile version of a page's content — crawled by Googlebot smartphone — for indexing and ranking. It is not a separate index and not a ranking boost. itself is historical/complete now: Google announced it November 4, 2016, began transitioning sites from late 2017, and rolled out broadly from March 2018; mobile search overtook desktop in several markets including Japan around 2015, cited as general rationale — but I’ve seen no evidence Japan was a named test market, so I won’t overstate it.) Second, formality and politeness register show up in query phrasing, which ties back to the script-choice patterns above.

Aside: the “Japanese keyword hack” — a spam attack, not a strategy

Because of the name, some people arrive at “Japanese keyword” content thinking it’s a gray-hat tactic. It isn’t. Google’s own web.dev documentation is blunt: “The Japanese keyword hack typically creates new pages with autogenerated Japanese text on your site in randomly generated directory names” (Google, web.dev). Those pages are “monetized using affiliate links to stores selling fake brand merchandise and then shown in Google Search.” A common tell: “If you’ve received a notification that someone you don’t know has verified your site in Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results., there’s a strong possibility that your site has been hacked” (same source).

The important prevention point: cleanup alone isn’t enough. Per the same web.dev article, 20% of hacked sites are hacked again within one day (web.dev) — so you have to close the entry point (usually an outdated CMSA content management system (CMS) is software that lets users create, manage, and publish digital content — like blog posts and pages — without writing raw code. WordPress, Drupal, and Joomla are the most common open-source CMS platforms./plugin), rotate passwords, and enable 2FA, not just delete the spam pages. This has been a long-running pattern; the Google Search Central Help Community has threads reporting it across at least a decade. Full step-by-step in the Playbook lens below.

AI summary

A condensed take on the Advanced version:

• Japan is a Google market, not a non-Google one. Yahoo! Japan has run on Google’s search/ad technology since the July 27, 2010 deal; the 2024 Japan Fair Trade Commission filing confirms it was operative through Oct. 31, 2022, and it was still cited in 2025 US antitrust litigation. Separately, LINE Yahoo’s own May 2024 earnings disclosure confirmed a 2-year contract extension keeping Google-powered search running through at least March 2027, after LINE Yahoo shelved a considered switch to Naver’s search technology over unresolved data- security concerns. Most “Yahoo! Japan SEOJapan SEO is the practice of optimizing sites to rank in Japan — a market where Google leads and Yahoo! Japan has run on Google's search technology since 2010. The distinctive work is Japanese-language keyword research across kanji, hiragana, katakana, and romaji, plus genuine localization rather than translation.” is just Google SEO — no separate organic submission process exists.
• Not the US Yahoo. US yahoo.com runs on Bing; Yahoo! Japan is a separate entity, today a brand of LY Corporation (half-owned via a SoftBank/Naver JV).
• Market share (StatCounter, June 2026): Google ~59%, Bing ~33% (surprisingly high — report the number, not an unverified cause), Yahoo! ~6%. A near-reversal from 2010 (Yahoo! ~57%, Google ~38%).
• Keyword research spans four scripts: kanji, hiragana, katakana, romaji. Variants are separate query stringsThe `?key=value` data tacked onto the end of a URL after a question mark — used for tracking, sessions, filtering, sorting, and search — and one of the biggest sources of duplicate URLs and wasted crawling in SEO. with different volumes and sometimes different intent (カフェ = Western cafe vs. 喫茶店 = traditional coffee house). Katakana ≈ commercial / loanwords / brand names; hiragana ≈ informational; kanji ≈ formal (pattern, not rule). Japanese has no word spaces, so engines segment morphologically.
• LocalizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. > translation: the Tokushoho disclosure page (compliance pattern, not a literal statutory “page” requirement), konbini/bank-transfer payment, denser page design. Google’s guidance is generic international SEOInternational SEO is the practice of optimizing a site so search engines understand which countries and/or languages it targets, and serve the right version to each user. It spans URL structure, hreflang, and on-page localization. — Japan has no special ranking docs; use hreflangHreflang is an annotation (in HTML, HTTP headers, or XML sitemaps) that tells search engines which language and optional region a page targets, and which alternate versions exist. It only works when every page in the cluster references all the others. normally.
• “Japanese keyword hack” = a site-compromise spam attack, not a tactic. web.dev documents it; 20% of hacked sites are re-hacked within a day, so fix the entry point.

Official documentation

Primary-source documentation relevant to Japan SEOJapan SEO is the practice of optimizing sites to rank in Japan — a market where Google leads and Yahoo! Japan has run on Google's search technology since 2010. The distinctive work is Japanese-language keyword research across kanji, hiragana, katakana, and romaji, plus genuine localization rather than translation.. Google is the operative engine (including for Yahoo! Japan’s organic results), so Google’s international/multilingual docs are the closest thing to “official Japan guidance” — Google publishes no Japan-specific ranking documentation.

Google — international / multilingual

• Managing multi-regional and multilingual sites — the framework Japan falls under (there is no Japan-specific version).
• Tell Google about localized versions of your page — the localized-versions / duplicate rule for translated pages.
• How Google Search handles multilingual searches (Sept. 2023) — how Google auto-detects which language(s) to serve. Background only; confirm the article body against the live page — automated fetches returned only the blog indexStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed..

Google — the “Japanese keyword hack” (web.dev)

• Fix the Japanese keyword hack — Google’s own detection-and-cleanup guide for the spam attack.

Japanese government (localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native./compliance)

• Act on Specified Commercial Transactions (Tokushoho) — official English translation — the legal basis for e-commerce disclosure requirements in Japan.

Regulatory (the Google/Yahoo! Japan arrangement)

• JFTC — Approval of the Commitment Plan submitted by Google LLC (April 22, 2024) — the dated, quasi-official confirmation the search-technology arrangement was operative through Oct. 2022.

Yahoo! Japan

• No standalone Yahoo!-Japan organic-SEO / webmaster-guidelines portal exists — because organic ranking is Google’s algorithm. Portal properties (Shopping, Auctions) have their own merchant onboarding, but that’s e-commerce feed integration, not organic-ranking SEO.

Quotes from the source

On-the-record statements. Each link deep-links to the quoted passage on the source page.

The 2010 Google / Yahoo! Japan deal

• “Yahoo search will be powered by Google in the future” and “users of the service will be served both paid and algorithmic search results generated by Google technology.” — TechCrunch, July 27, 2010. Jump to quote
• “We had many discussions with Yahoo Inc. and in the end, they recognized that this was the right strategy for us in Japan.” — Masahiro Inoue, President/CEO, Yahoo! Japan Corporation, quoted in Fortune, July 27, 2010. Jump to quote

(Google’s own July 27, 2010 announcement — Daniel Alegre on the Google Japan Blog — is Japanese-language only, so it’s paraphrased in the Advanced tab rather than quoted as English; do not treat any English renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM. of it as a verbatim Google quote.)

The arrangement’s persistence — JFTC

• Google LLC “restricted the provision of technologies for search engines and search advertising to Yahoo” from September 2, 2015 through October 31, 2022; and approval of the commitment plan “does not mean that Google LLC’s conduct violated the Act.” — Japan Fair Trade Commission, April 22, 2024. Jump to quote

The arrangement’s persistence — the commercial contract (2024 extension)

• “Yahoo! JAPANの検索サービスは、少なくとも2027年3月末まではグーグルから技術供与される” (“Yahoo! JAPAN’s search service will continue to receive technology from Google through at least the end of March 2027”) — Web担当者Forum, citing Nikkei, May 17, 2024. Jump to quote

Google on localized-page duplication

• “Localized versions of a page are only considered duplicates if the main content of the page remains untranslated.” — Google Search Central. Jump to quote

The “Japanese keyword hack” (web.dev)

• “The Japanese keyword hack typically creates new pages with autogenerated Japanese text on your site in randomly generated directory names.” — Google, web.dev. Jump to quote
• “These pages are monetized using affiliate links to stores selling fake brand merchandise and then shown in Google Search.” — same source. Jump to quote
• “If you’ve received a notification that someone you don’t know has verified your site in Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results., there’s a strong possibility that your site has been hacked.” — same source. Jump to quote

How should you approach SEO for Japan?

Answer a couple of questions and you’ll land on a recommendation. Every branch comes from the landscape and mechanics above.

SOP: Japanese multi-script keyword research

A repeatable procedure for building a Japanese keyword map that doesn’t miss script variants. Run it per topic cluster.

1. List the seed concepts in English for the cluster you’re targeting.
2. Generate every script renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM. for each concept: the kanji form, the hiragana form, the katakana form, and (for brand/foreign/tech terms) the romaji form. For foreign brand names, generate multiple katakana transliterations — more than one spelling can circulate.
3. Have a native Japanese speaker prune obviously-wrong or unnatural renderingsTurning HTML, CSS, and JavaScript into the final visual page and DOM. before you spend tool credits — but keep any variant that’s plausibly a real query even if it “feels” secondary.
4. Pull search volume for each variant separately in your keyword tool (Ahrefs Keywords Explorer, etc.). Do not assume the tool consolidates variants.
5. Eyeball the SERP for each variant. If two variants return clearly different ranking pages, they have different intent — note them as separate targets, not duplicates (the カフェ vs. 喫茶店 case).
6. Tag each variant by likely intent using the script heuristic (katakana ≈ commercial/loanword, hiragana ≈ informational, kanji ≈ formal) — as a hypothesis to validate against the SERP, not a fixed rule.
7. Map variants to pages. Same-intent variants can share a page (as natural on-page variations); different-intent variants may need their own pages.
8. Record volumes with a date and re-pull periodically — Japanese script-variant volumes shift, and you don’t want to reuse stale numbers.

Japan SEO myths and mistakes to avoid

The claims that come up most often — several are widely repeated and wrong:

• “You need to submit your site to Yahoo! Japan separately, like the old Yahoo Directory.” No. Yahoo! Japan’s organic results have come from Google’s algorithm since the July 27, 2010 deal (reconfirmed operative through Oct. 2022 by the 2024 JFTC filing, and under a commercial contract LINE Yahoo extended through at least March 2027). (Do-instead: run one Google program — Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results., standard technical SEOTechnical SEO is the practice of making a site easy for search engines to crawl, render, index, and (now) be eligible for AI answers. It's the foundation that lets your content and links rank — not a ranking trick of its own. — and you’ve covered Yahoo! Japan’s organic results. Portal features like Shopping have separate merchant onboarding, but that’s not organic-ranking SEO.)
• “Yahoo! Japan is the same as US Yahoo, so it uses Bing.” No. US yahoo.com runs on Bing; Yahoo! Japan is a separate entity (now a brand of LY Corporation) running on Google. (Do-instead: don’t apply US-Yahoo/Bing assumptions to Japan — treat Japan organic as Google.)
• “Kanji, hiragana, and katakana versions of a keyword are just stylistic variants — pick one.” No — they’re separate query stringsThe `?key=value` data tacked onto the end of a URL after a question mark — used for tracking, sessions, filtering, sorting, and search — and one of the biggest sources of duplicate URLs and wasted crawling in SEO. with different volumes and sometimes different meaning (カフェ = Western cafe vs. 喫茶店 = traditional coffee house). (Do-instead: research each variant separately and compare SERPs before deciding which to target or whether you need distinct pages.)
• “A good translation is the same as localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native..” No — translation converts language; localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. adapts payment methods, legal disclosures (the Tokushoho page), design conventions, and trust signals. (Do-instead: localize with native speakers and cover the compliance/trust layer; a literal translation can be grammatically correct yet non-compliant and off-putting.)
• “The Japanese keyword hack is a legitimate/gray-hat SEO tactic.” No — it’s a site-compromise spam attack documented by Google’s own web.dev. (Do-instead: if you see auto-generated Japanese spam pages on your site, treat it as a security incident — see the Playbook.)
• “Romaji is how most Japanese people search.” No — native-script (kanji / hiragana / katakana) queries dominate; romaji is a minority behavior, mainly for IT/tech terms and brand names without a settled katakana form. (Do-instead: cover romaji where it’s genuinely relevant, but don’t build your keyword map around it, and don’t invent a romaji-share percentage.)
• “Bing is negligible in Japan, so ignore it.” Stale. StatCounter’s June 2026 data puts Bing around a third of Japan’s search market. (Do-instead: since Japan organic is a Google program anyway, this mostly doesn’t change your SEO work — but don’t repeat the old “~5% Bing” figure, and treat the cause of Bing’s rise as unconfirmed.)

Incident playbook: the “Japanese keyword hack”

Linear runbook for when auto-generated Japanese spam pages appear on your site (a site compromise, not an SEO problem). Follow in order.

1. Confirm the symptom. Run a site:yourdomain.com search and look for pages of Japanese text in randomly-named directories you didn’t create. Check Google Search Console’s Security IssuesA Google Search Console report that flags signs your site was hacked or hosts content that could harm visitors — phishing, malware, or unwanted software. It's a user-safety flag, not a spam-policy penalty, and you clear it by fixing the problem and requesting a security review. and IndexingStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. reports.
2. Check for rogue GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. verifications. In Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance., review verified owners. An unknown owner is a strong signal the site was hacked (per web.dev) — remove any you don’t recognize.
3. Take a backup / snapshot of the current (compromised) state before you change anything, for forensics.
4. Find and close the entry point. Usually an outdated CMSA content management system (CMS) is software that lets users create, manage, and publish digital content — like blog posts and pages — without writing raw code. WordPress, Drupal, and Joomla are the most common open-source CMS platforms. core, theme, or plugin, or a compromised credential. Update everything; review recently-modified files and any unfamiliar admin users.
5. Remove the injected content. Delete the spam pages/directories and any injected redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't., cloaking logic, or malicious scripts. Confirm nothing legitimate depends on them.
6. Rotate all credentials and enable 2FA. CMS admins, hosting/SFTP, database. This is the step people skip — and 20% of hacked sites are re-hacked within one day per web.dev, almost always because the door was left open.
7. Request a review in Search Console if a manual action / security issue was flagged, once the site is clean.
8. Monitor site: results and GSC for recurrence over the following weeks, and keep the CMS/plugins patched going forward.

Multi-script keyword examples

Concrete before/after cases showing why single-script keyword research misses traffic and intent in Japan.

1. Coffee shop — script tracks meaning, not just spelling

• Before: Target only カフェ (kafe), assuming it means “coffee shop.”
• After: Split into カフェ (Western-style cafe), 喫茶店 (kissaten, traditional Japanese coffee house), and コーヒーショップ (literal “coffee shop”). The 喫茶店 searcher wants something genuinely different — a distinct page/section can win a visitor カフェ optimization would never reach.

2. Mobile phone — abbreviation and script split the volume

• Before: Target the full formal kanji 携帯電話 only.
• After: Also cover 携帯 (abbreviated kanji) and ケータイ (katakana of the abbreviation). Each pulls its own volume for what looks like “one keyword” in English.

3. Sushi — a single concept, 3+ valid renderingsTurning HTML, CSS, and JavaScript into the final visual page and DOM.

• Before: Optimize for 寿司 (kanji) alone.
• After: Note すし (hiragana) and スシ (katakana) as real query variants too. Even where they share intent, mapping them shows the true topical footprint and can inform natural on-page usage.

4. A foreign brand name entering Japan

• Before: Assume searchers type the brand in its English (romaji) form.
• After: Research the katakana transliteration(s) of the brand — that’s how most Japanese searchers will type a foreign name — and check whether more than one katakana spelling is in circulation, so you don’t cede the dominant variant.

AI prompts for Japan SEO

Ready-to-copy prompts. Always have a native Japanese speaker verify the output — AI is a drafting aid for script variants, not a substitute for native review.

Generate script variants for keyword research

I'm doing Japanese keyword research. For each English concept below, list its
common Japanese renderings in: (1) kanji, (2) hiragana, (3) katakana, and
(4) romaji. For loanwords/brand names, include multiple katakana spellings if more
than one is commonly used. For each variant, note whether it likely signals
commercial, informational, or formal/professional intent, and flag any variants that
carry a genuinely different meaning (not just a spelling difference).

Concepts: [list your concepts]

Sanity-check localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. vs. literal translation

Here is an English marketing sentence and its proposed Japanese translation. Tell me
whether the Japanese reads as natural, native-quality copy or as a literal/machine
translation, and rewrite it to sound natural to a Japanese consumer while preserving
meaning. Flag any cultural, formality-register, or trust issues.

English: [sentence]
Proposed Japanese: [translation]

Draft a Tokushoho page checklist

I run an e-commerce site selling to Japanese consumers. List the fields a standard
Japanese "特定商取引法に基づく表記" (Act on Specified Commercial Transactions)
disclosure page conventionally includes. Present it as a checklist. Note that this is
general information, not legal advice, and that I should confirm requirements with a
Japan-qualified professional.

Audit a suspected Japanese keyword hack

My site may have the "Japanese keyword hack." Give me a step-by-step detection and
cleanup checklist: how to confirm it via site: search and Google Search Console, how
to find and close the likely entry point (outdated CMS/plugins, rogue GSC owners),
how to remove injected pages, and the prevention steps (credential rotation, 2FA,
updates) that stop re-infection.

Japan SEO setup checklist

The Japan-specific work, in order — most of it keyword-research and localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native., not a separate engine:

• Confirm you’re treating Japan organic as a Google program (Yahoo! Japan runs on Google — no separate organic tooling or submission).
• Verify the site in Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results.; run standard Google technical SEOTechnical SEO is the practice of making a site easy for search engines to crawl, render, index, and (now) be eligible for AI answers. It's the foundation that lets your content and links rank — not a ranking trick of its own..
• Map language/region with hreflangHreflang is an annotation (in HTML, HTTP headers, or XML sitemaps) that tells search engines which language and optional region a page targets, and which alternate versions exist. It only works when every page in the cluster references all the others. (ja / ja-JP) like any international site.
• Ensure the main content is fully translated/localized, not just the template — a body left in English can be treated as a duplicate.
• Do multi-script keyword research: kanji, hiragana, katakana, and (where relevant) romaji, each variant pulled separately.
• For foreign/brand terms, research the katakana renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM.(s) specifically.
• Compare SERPs per variant to catch different-intent variants (e.g. カフェ vs. 喫茶店) and map them to the right pages.
• Have a native Japanese speaker review copy and the keyword shortlist.
• For e-commerce, publish the Tokushoho disclosure page (特定商取引法に基づく表記) and offer expected payment methods (konbini, bank transfer).
• Don’t “fix” dense Japanese page conventions into Western minimalism if the market expects them.
• Ignore the temptation to build a separate Yahoo! Japan program or a romaji-first keyword map.
• Monitor site: results and GSC for any sign of the Japanese keyword hack spam injection.

Resources worth your time

My related writing

• The Beginner’s Guide to Technical SEO — the technical foundation a Japan program still sits on, since Japan organic is a Google program.
• How to Do Keyword Research for SEO — the general workflow the multi-script Japanese process extends. (This is Japan’s first dedicated piece from me — I don’t have prior Japan/Yahoo-Japan-specific writing to point you to.)

From around the industry

• Yahoo Japan To Use Google Search (And Not Bing) In The Future (TechCrunch) — the contemporaneous account of the 2010 deal and the NetRatings query data behind the combined-share estimate.
• Yahoo Japan chooses Google for search and ads (Fortune) — Masahiro Inoue’s on-record quotes and the 2010 market-share figures (Yahoo! Japan ~57%, Google ~38%).
• Approval of the Commitment Plan submitted by Google LLC (Japan Fair Trade Commission) — the dated regulatory confirmation the arrangement was operative through Oct. 2022.
• Yahoo! JAPANが検索サービス乗り換えを断念、グーグルとの提携を2年延長 (Web担当者Forum, citing Nikkei, May 17, 2024) — the more current commercial-contract fact: LINE Yahoo extended its Google search-technology supply contract by 2 years, guaranteeing Google-powered search through at least March 2027, after abandoning a considered switch to Naver’s search technology.
• Fix the Japanese keyword hack (Google, web.dev) — the authoritative detection-and-cleanup guide for the spam attack.
• Localization SEO That Works: 10 Real-World Lessons from 14 Markets (Ahrefs, Mateusz Makosiewicz) — includes Ahrefs’ own experience treating Japan as an early localizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native. market, a real-world proof point that Japan traffic patterns are multi-script and non-obvious.
• Are You Choosing the Right Keyword for your Japanese SEO? (Oban International) — a worked kanji/katakana volume example (treat the specific undated figures as illustrative; pull fresh numbers from your own tool).
• Search Engine Market Share Japan (StatCounter Global Stats) — the current, dated market-share figures (re-pull before relying on them; the split drifts month to month).

Test yourself: Japan SEO

Five quick questions on how search actually works in Japan. Pick an answer for each, then check.

Japan SEO

Japan SEO is the practice of optimizing sites to rank in Japan — a market where Google leads and Yahoo! Japan has run on Google's search technology since 2010. The distinctive work is Japanese-language keyword research across kanji, hiragana, katakana, and romaji, plus genuine localization rather than translation.

Related: Naver SEO

Japan SEO

Japan SEO is optimizing a site to rank in Japan. The surprise for anyone arriving with an “SEO for other search engines” mental model is that Japan is not a non-Google market: Google leads organic search there, and Yahoo! Japan — still one of the country’s largest consumer portals — has run on Google’s search and ad technology since a July 2010 deal. So “optimizing for Yahoo! Japan” in the ranking sense is largely optimizing for Google; the real differences sit on the portal/UX and paid layers, not the organic algorithm.

That reframes what “Japan SEO” actually is. Two things make it its own discipline, neither of them a separate engine:

• The writing system. Japanese mixes kanji (漢字), hiragana (ひらがな), katakana (カタカナ), and rōmaji (Latin-script transliteration). A single concept can be searched in several valid scripts, and those variants don’t always behave as synonyms — search volume, intent, and even the pages that rank can differ by script. That makes Japanese keyword research fundamentally different from Latin-script markets.
• LocalizationLocalization is adapting content for a specific target market — not just translating the words, but adjusting currency, formats, idioms, cultural references, local search terms, and trust signals so the experience feels native., not translation. Japan is the textbook case of “translation is necessary but not sufficient.” Currency, payment methods (konbini, bank transfer), legal disclosures (the Tokushoho page), politeness register, and page-design conventions all have to be adapted, not just linguistically converted.

A separate, easily-confused topic is the “Japanese keyword hack” — a spam/security attack that injects auto-generated Japanese pages into a compromised site. It shares the name but is a hacking incident, not a keyword-research practice. Corporately, “Yahoo! Japan” today is a brand operated by LY Corporation (renamed from Z Holdings in 2023), not an independent company — and it’s unrelated to the US Yahoo, which runs on Bing.

Related: Naver SEO

Changelog