Does hreflang really help with ranking, or is it just for switching URLs?

Do hreflang tags influence organic rankings?

No, and this is where many go wrong. Google does not consider hreflang to decide which page will rank or at what position. Each language or geographic version (e.g., /fr/, /en-us/, /en-gb/) is evaluated independently by the ranking algorithm. If your /fr/ page has a weak backlink profile, duplicate content, or poor UX, it will rank poorly regardless of how perfect your hreflang is.

Hreflang comes into play after the ranking phase, solely to replace the URL displayed in search results with the one that best matches the user's language or country. If your /en/ ranks in position 3 for a query and a French user conducts the same search, Google may substitute /fr/ at the same position, provided the hreflang is correctly implemented and reciprocal.

What does “distinct pages in indexing” really mean?

This means that each language version must have its own ranking signals. There's no automatic “boost” because the English version performs well. If /fr/ has no French-speaking backlinks, little unique content, or a tight crawl budget, it will remain invisible even with perfect hreflang. Google crawls, indexes, and evaluates each of these URLs separately, as if they were distinct sites.

This is particularly evident for international e-commerce sites where certain versions (often /en-us/ or /uk/) accumulate 80% of external backlinks. Secondary versions need to be actively promoted, with local link building, culturally adapted content, and specific on-page optimization. Hreflang only performs the final step: showing the right entrance to the user once all doors have been built and optimized.

Why do so many SEO audits overrate the importance of hreflang?

Because a misconfigured hreflang is spectacular in the tools: cascading errors, red annotations in Search Console, alarming messages in crawlers. However, these alerts often mask the real problem: under-optimized language versions that rank poorly even without hreflang bugs. A site can have perfect hreflang and no visibility in Germany or Spain because the /de/ and /es/ pages lack local authority and relevance.

Conversely, a site with broken hreflang but very well-optimized language versions will continue to rank. The impact will be on UX (incorrect language displayed) and qualified traffic (visitors landing on the wrong version), not on gross ranking. It’s a plumbing issue, not an engine issue. Therefore, full audits should prioritize evaluating: localized content, backlinks by version, allocated crawl budget, engagement signals by language.

• Each language version ranks independently: no automatic transfer of authority between /fr/ and /en/.
• Hreflang does not boost the ranking, it only switches the displayed URL once ranking is established.
• A hreflang error impacts UX and qualified traffic, not directly organic positions.
• Prioritize on-page and off-page optimization for each version before focusing on hreflang.
• Audit tools often over-sell the severity of hreflang errors at the expense of real ranking signals.

Does this statement align with what is observed in the field?

Yes, largely. A/B testing on multilingual sites shows that fixing broken hreflang does not cause positions to jump. What changes is the click-through rate and bounce rate, as visitors finally land on the correct language. However, when targeted link building is invested in /it/ or /es/ and content is adapted with local phrases, then the positions genuinely rise.

But be careful: Google refers to “language and geographic versions” as if they are interchangeable, which is not always the case. [To be verified] A site with identical content in /fr-fr/ and /fr-ca/ may see Google ignore the hreflang annotation if the content is strictly duplicated. The statement overlooks the case of duplicate content across versions, which remains a blind spot for many international sites.

What nuances should be added to this announcement by Mueller?

The first nuance: hreflang can indirectly affect ranking through UX signals. If a German user lands on the English version because your hreflang is broken, they will likely bounce quickly. Google registers this dissatisfaction signal. Repeated at a large scale, it can degrade the organic performance of the English version itself. Thus, saying “hreflang does not impact ranking” is technically true but somewhat short.

The second nuance: the implementation of hreflang often reveals more severe structural problems. When checking annotations, we find that certain language versions are accidentally set to noindex, blocked by robots.txt, or never crawled. Hreflang then becomes a symptom of a shaky international architecture, not the root cause of the problem.

In what cases does this rule not fully apply?

Special case: sites with exact content duplication between closely related language versions (e.g., /es/ for Spain and /es-mx/ for Mexico with the same text). Google may arbitrarily choose which version to index, even with a clean hreflang. Here, the annotation is not enough: distinct content or at least localized blocks (currencies, legal mentions, local customer testimonials) are needed.

Another edge case: sites hosted on ccTLDs (.fr, .de, .es) where Google already applies a strong geographic bias through the domain. On a .fr, hreflang has less impact than on a .com/fr/ because the TLD already sends a targeting signal. Mueller's statement is valid, but the relative importance of hreflang varies depending on the chosen domain architecture.

What should you do to optimize a multilingual site?

First, audit each language version as a standalone micro-site. Check the backlink profile, the number of indexed pages, the crawl budget consumed, and the Core Web Vitals by version. Many international sites have 90% of their authority concentrated on /en/ while the /fr/, /de/, /es/ versions are nearly invisible. Hreflang won’t change anything as long as these versions lack their own momentum.

Next, implement hreflang in a bidirectional and comprehensive manner. Each URL must point to all its language alternatives AND to itself with an x-default for the default version. Test with Search Console (International Targeting report) and a crawler that can interpret hreflang. The most frequent errors: non-reciprocal annotations, invalid language codes (fr-FR instead of fr), hreflang tags in the body instead of the head.

What critical errors should be absolutely avoided?

Never treat hreflang as a miracle solution for cannibalization. If you have duplicate content between /fr-fr/ and /fr-be/, Google may ignore your annotations and arbitrarily choose which version to index. Hreflang is not a duplicate management tool; it's a targeting indicator. If the content is identical, differentiate it (localized block, local customer testimonials, specific legal mentions).

Another common mistake: forgetting cross-language canonical tags. If each language version has its own canonical pointing to itself, it aligns with Mueller's statement (distinct pages). But if you put a canonical from /fr/ to /en/, you completely undermine the indexing of /fr/. The canonical must remain intra-language, while hreflang manages cross-language interactions.

How can you check if your hreflang configuration is working properly?

Use a crawler like Screaming Frog or Oncrawl with the hreflang validation option. Export all annotations and check: reciprocity (if A points to B, B must point to A), completeness (all versions are mentioned), correct ISO syntax (language-COUNTRY in lowercase for the language, uppercase for the country). Then crawl from a localized IP or with a mobile user-agent to confirm that Google is showing the correct version.

In Search Console, check the “International Targeting” report under Settings. Google lists all detected hreflang errors (missing return URL, malformed tag, incorrect language code). Fix them in waves: first the blocking errors (no reciprocity), then the warnings (missing x-default). Resubmit the XML sitemaps enriched with xhtml:link tags to speed up processing.

• Audit each language version separately: backlinks, indexing, Core Web Vitals, allocated crawl budget.
• Implement hreflang bidirectionally with x-default for the default version.
• Truly differentiate the content between closely related geographic versions to prevent Google from ignoring the annotations.
• Verify canonical/hreflang consistency: intra-language canonical only, inter-language hreflang.
• Test with Search Console and a specialized crawler to validate syntax and reciprocity.
• Prioritize local link building on underperforming versions instead of over-optimizing hreflang.