How to Register Your Website with Naver Search Advisor (and Daum)

Somebody has noticed that the company does not show up on Naver, and the job has landed with you. Start with what Naver itself says: there is no registration you have missed. Naver's own FAQ states, in translation, that "no search registration procedure exists". Sites get into its results without signing up for anything.

Registering is still worth an afternoon. For a new site, Naver says it "can help", and it is the only way to see what Naver's crawler makes of your site and to ask it to come back. This guide covers Search Advisor from the account to the first crawl requests, then Daum's smaller tool.

How much of Korean search goes through Naver

It depends on who is counting, and the two usual sources disagree by a wide margin.

StatCounter counts page views that arrive from a search engine on the websites carrying its tracking code. For South Korea in September 2026, across all platforms, it reported Google at 47.97%, Naver at 43.06%, Bing at 4.83% and Daum at 0.73%. On mobile alone, Naver led with 59.56%.

InternetTrend, a Korean service that measures search referrals to Korean websites, put Naver at 66.75% and Google at 27.57% for the same month.

They sample different sites, which is the likely reason they land so far apart. You do not need to settle it: even on the lower estimate, Naver handles more than four in ten searches in Korea and leads on phones.

What registration does, and what it does not

Search Advisor is Naver's counterpart to Google Search Console. Once you have proved you own the site, you can submit a sitemap and an RSS feed, ask Naver to crawl specific URLs, and read reports on how its crawler sees the site.

It does not put pages into the index or move them up the results. Naver decides both, on the content. Registration removes avoidable obstacles and shows you what is happening; it does not change the verdict.

Two points on content:

• Language. Naver has no rule against pages that are not in Korean, so your English pages are eligible. Korean users mostly search in Korean, though, which means an English page competes for a narrow slice of queries.
• Generated and translated text. Naver's content guidelines accept AI-assisted writing when the result is useful. What they treat as spam is content produced in bulk or mechanically altered. Several hundred pages run through a translation API and published in one go sit uncomfortably close to that description.

Before you start: the Naver ID

Search Advisor runs on an ordinary Naver ID. Its terms say the service is open to anyone, free of charge, and they describe no identity check.

What matters for a company outside Korea:

• An overseas phone number works. Naver accepts non-Korean numbers when you create an ID. If you have been through Baidu's account process, which needs a mainland Chinese number, this is where Naver is far easier.
• The passport check you may have read about is something else. Naver added passport verification in June 2026 for payments in Naver Map. It has nothing to do with Search Advisor.
• The limits are generous. One ID can hold 100 sites, and one phone number can back three IDs (Naver help).
• The console is in Korean only. Browser translation is good enough to find your way around. Switch it off before you copy a token or a code snippet, because a translated string is a broken string.

A few Korean words help when you scan the menus: 사이트 (site), 소유확인 (ownership verification), 사이트맵 (sitemap), 수집 (crawling, literally "collection"), 요청 (request) and 리포트 (report).

Decide whose ID it will be before you begin. Tokens are tied to the ID that issued them, so if the person who registered the site leaves, their successor verifies again with a token of their own. A shared team login is not the way round this: Naver's general terms forbid lending or sharing an account.

Step 1: add the site and prove you own it

Sign in at searchadvisor.naver.com, open the webmaster tool (웹마스터 도구) and add the site. Enter the address exactly as your canonical URLs use it, protocol and host included, so that everything you submit later refers to the same site.

Naver then asks for proof of ownership and offers two methods.

An HTML file. Naver generates a small file for you to download. Serve it from the root of the site.

A meta tag. It goes in the <head> of the home page:

<meta name="naver-site-verification" content="your-token-here" />

There is no DNS method, unlike the TXT record that verifies a whole domain in Google Search Console. Somebody needs access to the web root or the page template.

Take the file if you have a build pipeline: it is static, and nothing in your front end will rewrite it. Take the tag if you work in a CMS. Either way, put it in the HTML the server sends, not in a tag manager.

Redirects are where the meta tag check trips up. Naver follows an ordinary HTTP 301 or 302. It does not follow a redirect done in JavaScript, a meta refresh, or one that depends on a cookie, and in those cases the check never reaches the page carrying the tag. If your root URL picks a language with a script and then sends the visitor to /en/, make that redirect server-side or use the file method, which is fetched from its own path.

Leave the file or tag in place for good. Verification renews every year, and renewal only works while the file or tag is still on the site. A redesign that clears an unexplained file out of the web root loses you the verification at the next renewal. Commit it to the repository and write down what it is for.

Step 2: make sure Yeti can get in

Naver's crawler is called Yeti, and that is the name it answers to in robots.txt. For a site hosted outside Korea, the problem Naver's documentation calls out is a firewall that blocks it.

robots.txt

Read your robots.txt as Yeti would. A common boilerplate lets in a named list of crawlers and shuts out everyone else:

User-agent: Googlebot
Allow: /

User-agent: Bingbot
Allow: /

User-agent: *
Disallow: /

Yeti is not named, so it falls into the last group and is shut out of the whole site. Give it a group of its own:

User-agent: Yeti
Allow: /

The status code of the file matters as much as what it says. Naver's robots.txt guide spells it out:

• A 5xx error on robots.txt blocks everything. Yeti reads a server error as "crawl nothing". A WAF that answers bots with a 503, or an origin that fails under load, can switch Naver off without anyone noticing.
• A 4xx, or no file at all, allows everything. That matches the robots.txt standard (RFC 9309), and it is the harmless failure.
• Redirects are followed, up to five. A robots.txt that hops from http to https and then to www is fine. A longer chain is not.

The firewall

Geo-blocking rules, bot-management settings and rate limits on unfamiliar crawlers can each stop Yeti, and none of them shows up when you test from your own browser.

Naver publishes the IP addresses its crawler uses at searchadvisor.naver.com/doc/naverbot.json. Allowlist that list in your WAF or CDN rather than trusting the user agent string, which any scraper can copy. Then read your logs: requests from those addresses that come back with a 403 or a 503 are your answer.

A quick test from your own machine catches rules keyed on the user agent:

curl -sI -A "Yeti" https://example.com/robots.txt
curl -sI -A "Yeti" https://example.com/

It cannot catch rules keyed on IP address or country, because the request still comes from you. Check those in the firewall configuration.

Step 3: submit the sitemap and the RSS feed

Submit the full URL of your sitemap in the console. Naver sets three conditions:

• Same domain. A sitemap served from a separate storage bucket or CDN hostname does not qualify; generate it on the site itself.
• Under 50,000 URLs, the same ceiling as the sitemaps.org protocol.
• Under 10MB. That is much tighter than the protocol's 50MB, so a large sitemap that Google accepts can still be too big for Naver.

If the site publishes on a schedule (a blog, a newsroom, release notes), submit the RSS feed as well. The sitemap is the full inventory; the feed is the short list of what is new.

Naver also accepts IndexNow, the push protocol Microsoft Bing and Yandex launched in 2021, for telling it that a URL has changed. Naver gives no indexing guarantee for IndexNow submissions.

Step 4: crawl requests

A crawl request asks Naver to fetch one specific URL. It is the nearest thing to the "Request indexing" button in Google Search Console, and like that button it is rationed: Naver applies a daily cap per site and does not publish the number. Figures in third-party guides do not come from Naver; treat whatever the console allows on the day as the limit.

Each request takes "at least a day, up to several weeks", and none is guaranteed. Spend them where they count:

1. the home page
2. your main product or service pages
3. any page written for Korean visitors
4. pages that are new or have changed substantially

Leave the long tail to the sitemap. That is what it is for.

Daum: the smaller registration

Daum is a much smaller search engine: 0.73% in StatCounter's September figures. Since 27 July 2026 it has been run by 주식회사 다음, the company previously called AXZ, which Upstage acquired.

Its webmaster tool is live and still marked Beta, and it works differently from Naver's. There is no account: you log in with the site's URL and a PIN. Ownership is proved through robots.txt, where you add an authentication key that Daum gives you. And Daum says plainly that crawling "can take several months".

One content rule matters more on Daum than on Naver. Daum restricts documents produced with a translation tool. If your Korean pages came out of machine translation, do not expect Daum to treat them like original writing.

Daum also runs a separate listing service, 검색등록 (search registration), which is not the webmaster tool. It asks business sites to show a Korean business registration number and a mail-order sales registration number, which a company without a Korean entity will not have.

Two weeks later: what to check

Two weeks is a first look, not a verdict, and against Daum's own "several months" it is barely a start.

• Verification is still valid in both tools. A deploy since registration may have removed the file, the tag or the Daum key.
• The sitemap has been read, and Naver reports no errors against it.
• The crawl requests: which have been processed and which are still waiting.
• Yeti in your server logs, from addresses on Naver's published list, receiving 200s rather than 403s or 503s.
• robots.txt still returns a 200, still has the Yeti group and the Daum key, and has not picked up a new wildcard block.

If you would rather hand it over

Our Naver and Daum registration service does the work in this guide at a fixed price per domain: verification in both tools, the sitemap and RSS, the robots.txt and firewall check for Yeti, crawl requests for your key pages within Naver's daily limit, and a written report at two weeks. Naver's terms forbid sharing an account, so we register under our own Naver ID. You add the verification file or tag and the Daum key to your site, and what you get back is the report, not a login.

If the real question is a Korean version of the site rather than its registration, book a call.