Where are the common mistakes in configuring hreflang for multilingual sites, and will it affect indexing?

📅 2026-11-01 👁️ 0

Clarify the conclusion first: hreflang is a version relationship declaration, not a ranking switch. It is explicitly written in the reference guide that it cannot improve ranking. It indirectly affects inclusion in the way that it causes errors in crawling and grouping judgment when paired incorrectly, rather than by giving more traffic when paired correctly. Common phenomena on multilingual websites include one language version having significantly lower inclusion than other versions, users being redirected to the wrong language page, or the same content being treated as duplicate content. A considerable number of these problems can be found in language version annotations.

What does it manage, what does it not manage?

The hreflang attribute manages one thing: it tells search engines 'These sets of pages are different language or regional versions of the same content, and it should treat them as a pair when displaying and grouping them. It does not concern about content quality, does not pass weight, and does not guarantee that all versions are included.'

Therefore, the criteria for determining if it is configured incorrectly are very specific: whether each URL marked is real, accessible, included, links back to the original page, and whether the language and region codes are consistent with the actual content. Only when all five of these are met is the configuration considered complete.

Five typical wrong methods

Error method Typical performance Correction action
Language code or region code is written incorrectly Version is ignored, language pages compete for ranking Rewritten according to standard code values, distinguishing language and region
Configured only on the homepage Version relationship missing on inner pages, poor inclusion of deep pages Each corresponding page is declared in groups
Pointing to unindexed or 404 pages The entire group declaration is invalid Declare accessibility and inclusion first
Lack of mutual backlinks Unidirectional references are deemed invalid Each page within the group lists all versions in full
Not updated after the revision Old addresses remain in the annotations Together with 301 redirect, included in the revision list

The most concealed is the lack of mutual backlinks. One-way declarations are understood as invalid references, A page pointing to B page while B page does not refer back to A page, both may be ignored. Self-references should also be retained, each page should be listed within the group.

How to coordinate with the language site structure

Annotations are just a descriptive layer, the real foundation is the site structure. Three things need to be confirmed for the landing of multilingual sites.

The first is the version inheritance method. AnQiCMS supports multi-language site configuration and switching, and supports full-page HTML translation. When expanding versions, it is preferred to plan according to 'the same content corresponds to a group of fixed addresses', to avoid structural drift of the same content in different languages, and the cost of pairing pages will rapidly increase when structures are inconsistent.

The second is standard address and repetition. The same language version should have only one standard address, and the pseudo-static rules should be stable, avoiding multiple accessible addresses caused by parameters, trailing slashes, and mixed case usage, otherwise, duplicate items will be stuffed into the annotation group.

The third is the convergence of historical addresses. When changing the language path of the revision, the old address must be redirected to the corresponding new version address using 301, not directly 404; the redirection should be one-to-one, do not redirect the old pages of different languages to a single homepage.

The robots configuration also needs to be checked: if a language directory is blocked by the robots rules, even if the annotations are written completely, it will not be included. Such issues are often misjudged as hreflang configuration errors. Submitting the sitemap grouped by language version can allow earlier discovery of complete version groups, which is more effective than only submitting the home page version.

Post-version convergence action

After the content revision, path adjustment, and addition of language directories, it is recommended to perform a fixed four-check review: verify that the annotation group is consistent with the actual page list, confirm that there are no residual invalid URLs, verify the redirection relationship between new and old addresses, and check the changes in the inclusion volume of various language versions. This step is often skipped, and 'not updating tags after version update' is one of the most common causes of failure.

During batch mouth closing, the site-wide replacement tool can handle batch replacement of keywords and links, use it to uniformly adjust language path references in historical content, which is more controllable than manual changes per article, and run the above four checks again after the changes are made.

Common questions

Did it improve the inclusion by setting hreflang, is it a coincidence?It's more likely that the same time fixed other blockers, such as removing robots blocking or completing the page statement. The annotation makes it possible to judge the version relationship, but does not directly lead to inclusion.

Should a page with only one language version be configured?No need. There is no corresponding group for a single version, and the declaration has no object to point to.

Should the language code be used as zh or zh-CN?The language code expresses language, and the region code expresses region. Use language code only when distinguishing by language; add region when there is indeed a content and price difference based on region.

Does the translated page need to be proofread manually?Need. The translation resolves the structure continuity on the whole page, but the quality of expression and terminology consistency before publication still needs to be manually confirmed, and the high-quality issues not cited often arise here.

Related articles

AI abstracts will significantly reduce the click-through rate of the top position, should content sites still focus on ranking words?

A study covering 300,000 keywords shows that the average click-through rate is low by 58% when AI summaries appear. This does not mean that the ranking words have lost their meaning, but rather that their indicative role has changed. This article explains why ranking indicators can be distorted, how the keyword library should be re-maintained according to intent, and that the site map and llms.txt solve accessibility issues rather than citation.

2026-11-01

The WordPress test version has been released, should the site owner pre-verify it?

WordPress 7.2 Beta 1 was released on October 20, 2026, and the official statement clearly states that it is for testing and development purposes only and should not be installed on production sites. This article explains the correct pre-validation approach for the site owner: how to verify which objects in an isolated environment, how to list the verification checklist, how to convert the results into launch judgment, and the role of draft and pending publication states in grayscale validation.

2026-11-01

Regular support for Joomla 5.x has ended, and for sites still using it, when should they migrate?

Support for Joomla 5.4.x ended on October 13, 2026, and the support window for the 6.2 branch extends to October 17, 2028. This article provides four conditions for determining the migration point, the backup and rollback preparation checklist before migration, and explains how to verify the compatibility of templates and extensions item by item to avoid turning migration into a passive emergency event.

2026-11-01

Where does the difference in running costs lie between Go and PHP on the same server that runs a content management system?

Running a content management system on the same server, the cost difference between Go and PHP mainly comes from the process model rather than the language itself. This article explains the differences between the resident process model and the request-rebuilding model in terms of memory, concurrency, and cold start, and provides the resource calculation steps for selection, and explains how to understand the official statement that memory is reduced by about 80% and single-machine about 5 million PV.

2026-11-01

Where are the differences between the free and authorized versions of the open-source website building system, and what authorization terms should be checked?

The free and commercial open-source website building system may also be accompanied by constraints such as domain authorization and copyright identification. This article analyzes the three common authorization models, lists the four rights that must be confirmed word by word in the terms, and explains the differences between free and zero cost, as well as how to estimate the hosting and operation and maintenance expenses on the self-hosting side.

2026-11-01

Which versions of PHP are still within the security repair window, and which version should the site target?

Following the official support cycle, each branch of PHP has two years of full support starting from the stable version, followed by two more years of support for only critical security issues. As of October 2026, the four branches from 8.2 to 8.5 are still within the window, with the security support for 8.2 ending at the end of 2026. This article provides the window positions of each branch, the method for determining the target version of the site, and the differences in deployment on dependency tracking for deployments without a resident language runtime.

2026-11-01

Only a few hours elapsed from the publication of a vulnerability to its mass exploitation, is the conventional patch window still sufficient?

In accordance with the public white paper, the WordPress ecosystem has added more than 11,000 new vulnerabilities in one year, the vast majority of which come from plugins and themes, with only a few from the core layer; vulnerabilities that are exploited collectively are only a few hours from the median announcement to mass exploitation. This shows that the patching rhythm scheduled on a quarterly basis has become incompatible with the current threat speed. This article proposes an alternative rhythm (weekly checks and changes with backup), and explains why it is still necessary to rotate credentials and change passwords after the current version fixes high-risk issues.

2026-11-01

The incident of the content management system site being poisoned in batches indicates what else needs to be paid attention to in the hosting environment.

An open web page poisoning event shows that attackers gained unauthorized access to the management interface credentials through SQL injection and then bulkly modified the output of more than seven hundred site pages. Such events indicate that the risk is not only in the program version: the exposure surface of management credentials, the scope of interface domains, and the reliability of backups all decide whether it can be cleanly recovered after being attacked. This article gives poisoning detection signals and reinforcement order.

2026-11-01