Can a foreign brand without a Japan entity register a .jp domain?

Not directly, and this is the constraint that decides the architecture question for most brands before any SEO argument is reached. Japan Registry Services, the registry that operates the .jp namespace, states that a General-use JP domain name is open to "any individual, group or organization having a permanent postal address in Japan" — the eligibility test is a real, deliverable Japanese address held by the registrant, not Japanese nationality or a Japanese company. The organizational type most brands actually want, .co.jp, is stricter still: JPRS restricts it to companies holding an official corporate registration in Japan, with non-Japanese companies registered as 外国会社 (Gaikoku Kaisha) also eligible. So a brand with no Japanese address and no Japanese registration cannot register either one on its own account. The workaround registrars sell is a trustee or local-presence service, and it carries a specific cost described in the article above: the trustee, not you, becomes the legal holder of record. Whether you need a Japanese entity for the wider business reasons is a separate question we work through in our guide to whether you need a Japan entity for ecommerce.

Does Google require a .jp domain to rank in Japan?

No. Google's own multi-regional documentation lists four URL structures — country-specific domains, subdomains on a generic domain, subdirectories on a generic domain, and URL parameters — and treats the first three as workable options with different trade-offs rather than ranking them as a single correct answer. A country-code top-level domain gives "clear geotargeting" and makes server location irrelevant, but Google's own listed drawbacks are that it is expensive, needs more infrastructure, is subject to strict ccTLD requirements, and can only target one country. Subdirectories are described as easy to set up with low maintenance; the stated downside is that users may not recognise the geotargeting from the URL alone. Google also names the other signals it reads for regional relevance — hreflang annotations, server location, and "other signals" including local addresses and phone numbers on the pages, local language and currency, and links from other local sites. A subdirectory that is genuinely in Japanese, prices in yen, and carries a Japanese contact address is sending most of those signals without a .jp domain in play at all.

Will a separate Japan site split my search authority?

This is the most common argument against a dedicated .jp site and it is worth stating precisely rather than as a slogan. Google does not publish a transferable "authority score", so the honest version of the concern is structural, not numeric: a new domain starts with no history, no accumulated links, and no established crawl pattern, and every external link, mention and citation that points at your Japanese content from that point on accrues to a property separate from your main domain rather than to the same one. A subdirectory inherits the crawl history and link profile of a domain that already exists. That is a real difference in starting position, and it is why the split is easiest to justify when the Japanese operation is expected to build its own local link profile — press, retail partners, Japanese media — rather than lean on your global domain's existing reputation. It is hardest to justify for a brand whose Japan plan is one localized catalogue and a landing page, where the new domain gets the isolation without ever generating the local signals that were supposed to pay for it.

Should I automatically redirect Japanese visitors to the Japanese version?

Google advises against it in plain terms. Its multi-regional guidance says to avoid automatically redirecting users from one language version of a site to a different language version — for example, redirecting based on what you believe the user's language to be — because such redirects can prevent users and search engines from reaching all the versions of your site. The practical failure mode is that Googlebot, which typically crawls from a US-based location, gets bounced away from your Japanese pages by the same rule that bounces a Japanese customer away from your English ones, and the pages you most want indexed for Japan are the ones the crawler cannot see. The usual alternative is a visible, dismissible banner offering the Japanese version rather than a forced redirect, leaving both the visitor and the crawler able to reach either version. This applies identically whether the Japanese content sits on a subfolder or on a separate .jp domain.

If I start with a subfolder, can I move to a .jp domain later?

Yes, and sequencing it that way is usually the lower-risk path, with two caveats worth planning for now rather than discovering later. The first is that a domain migration is a real project — every URL needs a redirect, every hreflang annotation on both sides needs rewriting, and search engines need time to reprocess the change — so the move is cheaper the earlier it happens and gets more expensive with every page and inbound link you add. The second is the eligibility gate: if the reason you are on a subfolder is that you have no Japanese address, then the later migration is gated on the same entity or trustee decision, not on your traffic reaching some threshold. Brands that intend to establish a Japanese entity eventually are in the cleanest position — run the subfolder now, register .co.jp in the company's own name once the entity exists, and migrate then, having never handed a domain to a trustee at all. The reverse sequence, migrating from a trustee-held .jp to a company-held one, means a transfer negotiated with the party that legally owns the name.

One Global Site or a Dedicated .jp Site for Japan?

A scope note first. This article is about where your Japanese content lives — one domain with a Japanese section, or a separate Japanese site — and nothing else. It is not about whether you need a Japanese company to trade at all, which has its own answer in our guide to whether you need a Japan entity. It is not about what the price field says, covered in yen versus dollar pricing. And it is not about marketplaces: if your Japan plan is a Rakuten or Amazon storefront, the architecture question does not arise in this form, and the cross-border versus local store comparison is the relevant decision instead. This is for brands running or planning their own direct-to-consumer site into Japan.

The Gate That Comes Before the SEO Argument

Start here, because for a large share of brands this section ends the discussion before the trade-offs are ever reached. A .jp domain is not something you can simply buy from abroad. Japan Registry Services (JPRS), the registry that runs the namespace, publishes eligibility rules per domain type, and both of the types a commercial brand would want carry a Japan-presence requirement.

For a General-use JP domain — the plain yourbrand.jp — JPRS states that registration is open to any individual, group or organization having a permanent postal address in Japan. The registrant does not have to be a Japanese company or a Japanese national; what it has to have is a real, deliverable Japanese address, which appears as the registrant contact in the registry's public Whois record. For .co.jp, the organizational type that reads as an established Japanese business to a Japanese customer, the bar is corporate rather than postal: JPRS restricts it to companies (会社) holding an official corporate registration in Japan, and notes that non-Japanese companies registered in Japan as 外国会社 (Gaikoku Kaisha) may also apply. There is a further limit worth knowing before you plan around it: JPRS itself classifies organizational-type JP domain names — the category .co.jp belongs to — as types "in which the number of registration is limited to one per registrant." One company, one .co.jp. Registrars additionally describe the application as requiring a copy of the corporate registry extract (登記簿謄本) as proof; that procedural detail comes from registrar documentation rather than the JPRS pages we read directly, so treat it as the operational shape rather than as quoted registry policy.

The practical consequence is easy to state and frequently missed in English-language advice: a brand with no Japanese address and no Japanese registration cannot register either domain in its own name. The "just get the .jp, it's the strongest signal" recommendation assumes an eligibility you may not have. Everything below about geotargeting, speed and maintenance is downstream of this; if you fail the gate, your real choice is between a subfolder or subdomain on the domain you already own, and a .jp domain held on your behalf by somebody else.

The Trustee Route, and What You Hand Over

Registrars serving international clients sell a way around the address requirement, marketed as a trustee or local-presence service. Their own descriptions of how it works are worth reading closely, because the mechanism is not a formality. In these services the provider, its local branch, or a local agent it assigns becomes the legal holder and administrative contact of the domain. That much is common to both domain types: the vendor pages for general .jp and for .co.jp use the same wording for who holds the name.

Below that, the terms diverge, and the difference matters. For general .jp, the vendors we read state that the domains may be hosted on your own name servers — you keep DNS. For .co.jp, one such registrar states plainly that the agent is liable under Japanese law for infringement associated with the domain, and that consequently the agent insists on retaining control of DNS and will not set that requirement aside under any circumstances. The same .co.jp page frames the arrangement as a preliminary solution, recommending that a brand building a long-term Japanese presence register a local company and have the domain transferred to it for a fee. So the version of this service that gets you the domain type Japanese customers read as an established local business is also the version where you do not run your own DNS.

These are commercial pages from vendors selling the service, not registry policy, so read them as vendor terms rather than as law — but note which direction the bias runs. A vendor's own marketing page has every incentive to make its service sound frictionless, and it is telling you anyway that you will not be the legal holder and will not control DNS. That is not a detail buried by a critic; it is the seller's own description.

If that arrangement sounds familiar, it should. It is the same structural problem as an agency holding your marketplace account, which we work through in switching Japan EC agencies and who owns your store. The asset your Japanese customers, your printed packaging and your accumulated search history all point at is registered to a third party, and recovering it later is a negotiation rather than a settings change. A .jp domain acquired this way is a real domain with a real geotargeting benefit, and it is also a dependency on a counterparty you have not otherwise vetted. Whether that trade is worth making depends entirely on how central Japan is to your plan — which is exactly the question the rest of this article is built to answer.

Deciding this before you commit a domain, a theme and a year of content to it? Tell us what your current site runs on and where Japan sits in your plan, and we'll tell you which structure we'd build on.

Talk to LAUNOVA

What Google Actually Says About the Three Structures

Google's own multi-regional documentation is more measured than the confident rankings that circulate in SEO blog posts. It lays out four URL structures and gives each a pros-and-cons entry rather than a verdict.

Country-specific domain (example.jp): pros are clear geotargeting, server location becoming irrelevant, and easy separation of sites. Cons, in Google's own list, are that it is expensive, requires more infrastructure, has strict ccTLD requirements, and can target only a single country.
Subdomain on a generic domain (jp.example.com): easy to set up, allows different server locations, easy to separate sites. The stated con is that users may not recognise the geotargeting from the URL alone.
Subdirectory on a generic domain (example.com/jp/): easy to set up, low maintenance because it shares a host. Cons are the same recognition problem, a single server location, and that separating sites is harder.
URL parameters (example.com?loc=jp): Google does not recommend this at all.

Note what the ccTLD row actually says: "strict ccTLD requirements" is Google naming, in one clause, the entire eligibility problem the previous two sections spent a page on. And note what the subdirectory row does not say — there is no claim that a subdirectory cannot rank in Japan.

More useful than the table is Google's list of the signals it reads to decide which country a page is aimed at: country-code top-level domains; hreflang annotations, whether in tags, headers or sitemaps; server location via the server's IP address; and "other signals" that Google enumerates as local addresses and phone numbers on the pages, the use of local language and currency, links from other local sites, and signals from your Business Profile. Google is explicit that server location is not definitive, since sites use CDNs or host where infrastructure is better.

Read that list as a checklist and the picture changes. A ccTLD is one signal out of several. A subdirectory written in Japanese, priced in yen, carrying a Japanese contact address and phone number, and linked to by Japanese sites is sending nearly every other signal on the list. The brand that skips the .jp domain and does those four things is in a materially different position from the brand that skips the .jp domain and publishes machine-translated pages under /jp/.

The Lever Google Removed in 2022

There used to be a manual override. Search Console's International Targeting report let a site owner set a site-wide country target, which was the standard way to point a generic domain or a subdirectory at a specific market. Google discontinued it. Its own help documentation states that the ability to target search results to specific countries using Search Console country targeting "was determined to have little value for the ecosystem, and is no longer supported," while confirming that Google will continue to support and use hreflang tags on pages. Industry coverage at the time dated the removal to September 2022; the Google page we read did not itself state the date, so treat the month as reported rather than officially confirmed.

This matters more than it first appears for anyone choosing a subdirectory. There is now no site-wide switch in Search Console that tells Google your /jp/ section is for Japan. The declaration has to come from the signals themselves — and the one you fully control is hreflang.

Google's requirements there are specific and unforgiving in one respect. Annotations can be delivered three ways: HTML link elements in the page head, HTTP Link headers, or xhtml:link entries in an XML sitemap. They must be bidirectional: if page X links to page Y, page Y must link back to page X, and Google states that where this is not the case, the annotations "may be ignored or not interpreted correctly." A one-way declaration is not a weak signal — it is potentially no signal. There is also the x-default value for visitors whose settings match no version you publish.

The operational risk sits with the two-site setup rather than the one-site setup. When your Japanese pages live in a subdirectory of the same codebase, the return links are generated by the same template and stay in sync by construction. When they live on a separate .jp site — different codebase, possibly a different platform, possibly a different vendor — keeping every English page's annotation pointed at the right Japanese URL and every Japanese page's annotation pointed back becomes a maintenance obligation that breaks quietly. Platform automation helps if you stay inside one platform: Shopify generates hreflang annotations automatically for international domains, subdomains and subfolders configured through its markets feature, and updates them when the configuration changes. Split your Japanese storefront onto a second Shopify store and that automation stops spanning the two properties, because they are no longer one store's market configuration.

Speed Is a Hosting Decision, Not a Domain Decision

"A Japanese site loads faster for Japanese customers" is the most-repeated argument for the split, and it conflates two things that are separable. Latency comes from the physical distance between the visitor and the server that answers them, plus whatever the site does after the first byte. The domain in the address bar does not determine either. A .jp domain served from a single origin in Virginia is slow in Tokyo; a .com subdirectory served from a CDN with Japanese edge nodes is not.

Google's own wording supports treating these separately: server location is described as a signal that is often physically near your users, while explicitly noting it is not definitive precisely because sites use content delivery networks. If you are on a hosted commerce platform, its CDN is already answering Japanese requests from a nearby edge, and buying a domain does not change that path.

The number worth holding onto is the target, not a comparison. Google's Core Web Vitals guidance sets Largest Contentful Paint at 2.5 seconds or less, measured at the 75th percentile of page loads and segmented across mobile and desktop. That is the bar for Japanese visitors regardless of which structure you pick, and it is a bar you clear or miss on hosting, image weight, and third-party scripts. We are deliberately not quoting a millisecond figure for Tokyo-to-US round trips: the honest answer depends on your host, your CDN's edge map and your page weight, and a made-up number would be worse than none. Measure your own site from Japan before treating speed as an argument for either option.

The Recurring Bill Each Option Signs You Up For

The costs that decide this in practice are the ones that repeat, not the domain registration fee. Running a separate Japanese site means a second content system to update whenever a product, a price or a policy changes, and the failure mode is not dramatic — it is a Japanese site that is three product launches behind, because the Japanese update is a separate task somebody has to remember. It means a second set of legal and policy pages, since Japan's own disclosure requirements apply to the storefront Japanese customers actually see. It means a second platform subscription, a second set of paid apps, and a second analytics property whose numbers have to be reconciled with the first.

A subdirectory converts most of that into a translation workflow inside one system. Every page has an owner, every deployment covers both languages, and the hreflang wiring is generated rather than maintained. What you give up is the ability to run the Japanese site on different infrastructure or hand it wholesale to a Japanese team, and — if your product catalogue diverges sharply by market — the freedom to let the Japanese site have a genuinely different structure rather than a translated version of the same one.

That last point is the strongest honest case for the split, and it is worth stating fairly. If your Japanese assortment, pricing architecture and merchandising are going to diverge substantially from your home market, and if a Japanese team will own the site day to day, a separate property stops being an SEO play and becomes an organisational one. Under those conditions the maintenance argument reverses: keeping two genuinely different businesses inside one template is its own recurring cost. Most brands entering Japan are not in that position at the point where they are asking this question, but some are, and the ones that are should not be talked out of it by an authority argument.

A Decision Rule

In order, because each question narrows what the next one has to answer.

Do you have, or will you shortly have, a Japanese corporate registration or a permanent Japanese address you control? If yes, .co.jp or a general .jp registered in your own name is genuinely available and the rest of the trade-offs apply normally. If no, the real comparison is subfolder versus trustee-held domain, and question 2 decides it.
Are you willing to have a third party as the legal holder of your Japanese domain? For a brand whose Japan operation is one localized catalogue, that dependency buys a geotargeting signal you can substitute for with language, currency, a local address and local links. For a brand making Japan a core market, the answer is usually to fix the entity question rather than to accept the trustee.
Will the Japanese site build its own local link profile? Japanese press, retail partners, local media and industry directories linking to a separate domain make that domain's isolation pay for itself over time. No local link-building plan means the split gets the isolation and none of the compensation.
Who updates the Japanese site next Tuesday? If the answer is "the same person who updates everything else, in the same admin," a subdirectory matches how the work actually happens. If it is a Japanese team with its own roadmap, a separate property matches your org chart and the maintenance argument weakens.
Can you keep bidirectional hreflang correct across two systems? If nobody owns that, the split will silently lose the signal it was supposed to buy, and you are better off with the structure where correctness is generated rather than maintained.

The default for a brand at the start of Japan entry, on this reading, is a Japanese section on the domain you already own — genuinely in Japanese, priced in yen, with a Japanese contact address — with the .jp domain deferred until either a Japanese entity exists or Japan has earned its own team. That is a sequencing recommendation, not a claim that dedicated domains do not work.

What LAUNOVA Does — and Does Not Do

LAUNOVA runs Japanese-language ecommerce operations for overseas brands. On the site side that includes Shopify Japan localization — Japanese product copy, checkout and payment-method coverage, and the storefront configuration that makes a Japanese section read as a Japanese store rather than a translated foreign one — and, as part of Japan market entry work, thinking through this architecture decision before a domain and a year of content are committed to it. We scope and price that against what you actually hand over rather than a fixed package, so there is no rate card to quote here.

What we do not do: we are not a domain registrar and we do not act as a trustee or local-presence holder for anyone's .jp domain. We do not advise on whether to incorporate in Japan, which is a question for a Japanese lawyer or 司法書士, and we will route you to one rather than improvise. We do not provide legal opinions on Japanese disclosure requirements for your specific storefront. And we have a visible interest in this topic worth naming: we sell localization and operations work, which is easier to deliver inside one well-structured site than across two diverging ones — so read our default recommendation knowing that. If the four questions above point you toward a separate Japanese property, we will say so.

If you are working through this now, the most useful things to tell us are what platform your current site runs on, whether you already hold or can obtain a Japanese address, and whether Japan will have its own team. Get in touch and we will tell you which structure we would build on and why.

Do You Need a Japan Entity?

The entity question that gates .co.jp eligibility, worked through on its own terms.

Should a Japan Storefront Price in Yen?

Local currency is one of the geotargeting signals this article counts on — here is what it costs to skip.

Switching Japan EC Agencies

The same ownership problem as a trustee-held domain, in the account layer.

Sources

• JP domain name types and eligibility — General-use JP open to "any individual, group or organization having a permanent postal address in Japan"; CO.JP restricted to companies with official corporate registration in Japan, with non-Japanese companies registered as "Gaikoku Kaisha" also eligible. First-party, Japan Registry Services (jprs.co.jp/en/jpdomain.html and jprs.co.jp/en/regist.html), both retrieved directly for this article.
• The one-registration-per-registrant limit on organizational-type JP domain names, the category containing CO.JP — "Domain name types in which the number of registration is limited to one per registrant." First-party, Japan Registry Services (jprs.co.jp/en/jpdomain.html).
• The 登記簿謄本 (corporate registry extract) documentation requirement for a CO.JP application. Secondary, registrar documentation (Japan-specialist registrars and Web Solutions), cross-checked against each other. The JPRS pages we read do not state this, so it is presented as the operational shape of the process rather than as quoted registry policy; confirm with your registrar.
• Trustee / local-presence services for JP domains — the provider, its local branch, or its assigned local agent becomes the legal holder and administrative contact. This wording appears on both the general .jp and the .co.jp vendor pages. The DNS-control and transfer terms are .co.jp-specific: on the .co.jp pages the agent is described as liable under Japanese law for infringement associated with the domain and therefore insisting on DNS control without exception, with the arrangement framed as a preliminary solution pending a local company and transfer for a fee; the general .jp pages from the same vendors state instead that the domains may be hosted on the customer's own name servers. Vendor commercial pages (Web Solutions, EuropeID), not registry policy. These are sellers describing their own product; we read the disclosure as credible precisely because it runs against the seller's interest, but the terms of any specific service must be checked in that provider's contract.
• URL structures for multi-regional sites — the four options (country-specific domain, subdomain, subdirectory, URL parameters) with Google's own stated pros and cons, including "strict ccTLD requirements" and single-country targeting as ccTLD drawbacks. First-party, Google Search Central, "Managing multi-regional and multilingual sites" (developers.google.com), retrieved directly.
• Signals Google uses for regional relevance — ccTLDs, hreflang annotations in tags/headers/sitemaps, server location by IP, and other signals including local addresses and phone numbers, local language and currency, links from local sites, and Business Profile signals; server location explicitly described as not definitive because of CDNs. First-party, same Google Search Central page.
• Guidance against automatic language-based redirection — "Avoid automatically redirecting users from one language version of a site to a different language version," on the grounds that it can prevent users and search engines from viewing all versions. First-party, same Google Search Central page.
• Search Console International Targeting deprecation — country targeting "was determined to have little value for the ecosystem, and is no longer supported"; Google continues to support and use hreflang tags. First-party, Google Search Console Help (support.google.com/webmasters/answer/12474899). The September 2022 removal date is secondary (Search Engine Land, Search Engine Roundtable) — the Google page we read does not state a date.
• hreflang implementation — three delivery methods (HTML tags, HTTP headers, sitemap entries), the bidirectional return-link requirement, the statement that annotations "may be ignored or not interpreted correctly" without it, and the x-default value. First-party, Google Search Central, "Tell Google about localized versions of your page" (developers.google.com), retrieved directly.
• Largest Contentful Paint threshold of 2.5 seconds or less, measured at the 75th percentile of page loads segmented across mobile and desktop. First-party, web.dev LCP documentation (Google).
• Automatic hreflang generation for international domains, subdomains and subfolders configured through Shopify's markets feature, updating when configuration changes. First-party, Shopify Help Center international SEO and international domains documentation — vendor product documentation describing that vendor's own behaviour, not independent verification.
• No latency or load-time comparison figure is quoted in this article. We did not run or locate a methodologically sound Japan-versus-overseas measurement we could stand behind, and a fabricated millisecond number would be worse than none. The mechanism (distance to the answering edge, not the domain suffix) is stated instead. Nothing in this article is legal or tax advice, and no case, client or figure here is a LAUNOVA number.

One Global Site or a Dedicated .jp Site? The Japan Architecture Decision