Are hreflang tags truly essential for a multilingual site?

Why does Google emphasize hreflang tags so much?

The issue is straightforward: without a clear signal, Google doesn't always guess which language version to serve to a given user. A French speaker browsing from Montreal might receive the Canadian English version while searching for French content. A German company could see its .de version displayed in the U.S. SERPs, creating a disastrous user experience.

Hreflang tags function like road signs. They tell Google: "this French URL corresponds to this English URL, which corresponds to this Spanish URL." The engine can then display the relevant version based on the browser language and the geographic location of the user.

How does this allocation actually work?

Google analyzes three elements to decide which version to show: the browser language, the geographic IP, and the hreflang signals declared in the code. If these signals are missing or contradictory, the engine makes a "best guess" based on the textual content and metadata, with a significant error rate.

The tag itself takes this form: <link rel="alternate" hreflang="fr-ca" href="https://example.com/fr-ca/" />. The language code follows the ISO 639-1 standard (two letters for the language), optionally followed by the ISO 3166-1 Alpha 2 region (two letters for the country). The language-region distinction matters: fr-FR differs from fr-CA, even though the French remains the same.

What types of sites are actually affected?

Any site offering multiple language versions of the same content falls within the scope. A French e-commerce site with an English version for export, a corporate blog available in 5 languages, a multilingual institutional site: all must implement hreflang.

The nuance lies in monolingual sites with multiple geographic targeting. A 100% English site targeting the UK, US, and Australia can use hreflang with en-GB, en-US, en-AU to refine targeting. Google then understands that the identical content is aimed at different audiences depending on geolocation.

• Mandatory reciprocity: if the FR page points to EN, the EN page must point back to FR
• Self-referencing required: each page must include a hreflang tag pointing to itself
• Cross-domain consistency: hreflang tags work across different domains (.fr, .com, .de)
• Technical validation: an error in a language code invalidates the entire linked cluster of pages
• Conflict priority: HTTP header > HTML head > XML sitemap

Does this recommendation truly reflect on-the-ground observations?

Let's be honest: correct implementation of hreflang tags remains a nightmare for 8 out of 10 sites. Technical audits reveal systematic errors: non-reciprocal tags, incorrect language codes, missing self-referencing, and conflicts between HTML method and XML sitemap.

Google claims that these tags "help with correct indexing," but the reality is more nuanced. The engine indexes all language versions regardless of hreflang tags. What hreflang controls is the display in the SERPs based on user context, not the indexing itself. [To be verified]: Google remains vague on the exact weight given to hreflang against other contradictory signals like geotargeting in Search Console or ccTLD.

In what scenarios does this approach have its limits?

Sites with partially translated content encounter a structural problem. Imagine a blog with 500 articles in French and only 50 translated into English. Should hreflang be declared only on the 50 translated pairs, or should a separate architecture be created? Google does not provide any clear guidance on this frequent scenario.

Another gray area: dialectal variants without a clear geographic boundary. Can a site offering Castilian Spanish and Latin American Spanish effectively use es-ES and es-419? Real-world feedback shows inconsistent results depending on the sector and competitive density. The 419 code (Latin America) works, but its interpretation by Google remains approximate according to markets.

What critical errors are most commonly observed?

The number one mistake: mixing language and geographic targeting in Search Console. A .fr site with hreflang fr-FR AND geographic targeting France in Search Console sends redundant signals. Worse, if the Search Console targeting says "France" but hreflang declares fr-BE (French-speaking Belgium), Google has to choose between two contradictory directives.

The second massive error concerns orphaned tags after a migration. A site transitions from /fr/ to /fr-fr/, but the old hreflang tags still point to the old URLs. Google detects 404s in the tags, considers the hreflang cluster broken, and ignores all directives. The result: a return to random "best guess".

How can you correctly implement hreflang tags without errors?

You have three methods: HTML tags in the <head>, HTTP headers, or declaration in the XML sitemap. The HTML method remains the most common and easiest to audit. Each page must contain the complete set of language variants, including itself.

Specifically, a French page /fr/produits/ with English and German equivalents will contain three tags: one pointing to /en/products/, one to /de/produkte/, and one self-reference to /fr/produits/. This self-reference is not optional: Google requires it to validate the cluster.

What technical checks should be performed after deployment?

The Search Console offers a dedicated hreflang report in Coverage > International, but it displays errors with a delay. For immediate diagnostics, use a technical crawler (Screaming Frog, OnCrawl) configured to extract and validate hreflang tags.

Check four critical points: complete reciprocity between all pages in the cluster, compliant ISO language codes (no "fr_FR" with underscores instead of dashes), absolute URLs (no relative paths), and consistency between declaration methods. If you declare hreflang in HTML and in the sitemap, both must be perfectly synchronized.

Should you use the x-default tag and how?

The hreflang="x-default" tag designates the version to display when no other matches the user's language/region. Typically, it points to a language selection page or the international English version.

Concrete example: a site offers fr-FR, en-GB, de-DE. A Japanese user does not match any of these variants. Without x-default, Google selects arbitrarily. With x-default pointing to /en/, the Japanese user systematically sees the English version. It is a fallback strategy that avoids random displays.

• Audit all hreflang tags with a technical crawler before production deployment
• Check reciprocity on a sample of 20-30 representative pages of each language version
• Test SERP display from different countries using a VPN and different browser languages
• Weekly monitor the International report in Search Console for 2 months post-deployment
• Document the chosen method (HTML vs sitemap vs HTTP header) to avoid duplicates during updates
• Synchronize hreflang tags with automatic geographic redirects if your site uses them