Why does Google still require rel alternate and rel canonical for mobile and desktop versions?

Does this recommendation still apply to all websites?

No. Google's directive specifically targets sites with distinct URLs for mobile and desktop (m-dot configuration like m.example.com vs www.example.com). This architecture was common a decade ago but has lost ground to responsive design.

With responsive design, a single URL serves the same content tailored for all screens. There’s no need for rel alternate or rel canonical: the duplication issue does not arise. If your site is responsive, this directive simply does not apply to you.

What is the exact role of these tags in a separate mobile architecture?

The rel alternate placed on the desktop page tells Google, "Hey, there’s a mobile version at this URL." It’s a discovery signal. The crawler understands it should also explore the m-dot version.

The rel canonical on the mobile page says, "This mobile page is a variant of the desktop page, which remains the main version." This prevents Google from considering the two versions as duplicate content and diluting the ranking between the two URLs.

Why does Google continue to promote this configuration?

Because millions of pages still use this architecture. Google cannot ignore this reality. Some large historical sites (e-commerce, media) have never migrated to responsive design for technical or budgetary reasons.

Therefore, the search engine must maintain documentation compatible with these legacy configurations. But let’s be clear: this is not a universal recommendation; it’s a guideline for a specific and declining scenario.

• rel alternate: signals the existence of the mobile version from the desktop page
• rel canonical: consolidates SEO signals toward the desktop version from the mobile
• This configuration only applies to architectures with distinct URLs (m-dot or mobile subdomains)
• Responsive sites need neither of these
• Google continues to document this approach to maintain compatibility with existing legacy sites

Is this directive consistent with the evolution of mobile-first indexing?

Technically yes, but it’s misleading. Since switching to mobile-first indexing, Google prioritizes crawling and indexing the mobile version of your pages. In an m-dot configuration with rel canonical pointing to desktop, you create a paradox: the mobile version says, "I am not the main version," yet Google indexes it as such.

In practice, this works because Google has learned to manage this contradiction. But it is far from elegant. [To verify]: some observe ranking fluctuations when both versions are not perfectly synchronized in content or structured data.

What concrete risks arise from a poor implementation?

I have seen sites lose 30 to 40% visibility due to errors in these tags. The classic case: misconfigured rel canonical pointing to the wrong URL, or worse, absent on some mobile pages. Google ends up with two indexable versions and chooses... randomly.

Another trap: thin mobile content. If your m-dot version hides content (closed accordions, truncated sections), and Google indexes this version through mobile-first, you lose relevance signals. The rel canonical does not compensate for insufficient content.

In what cases does this configuration remain relevant?

Honestly? Very few. If you are launching a site today, opting for an m-dot architecture is a strategic mistake. The only defensible scenario is a massive historical site where migrating to responsive would cost millions and involve significant technical risks.

Even then, the trend favors dynamic serving (same URL, different HTML based on user-agent) rather than separate URLs. It's simpler to maintain and less prone to markup errors. If you're still on m-dot, plan your transition rather than optimize this setup.

What concrete actions should you take if your site uses distinct mobile URLs?

First, thoroughly audit the consistency of the tags. Each desktop page must have a rel alternate pointing to its exact mobile equivalent. Each mobile page must have a rel canonical pointing to the corresponding desktop page. One mismatch, and Google may ignore the entire pair.

Use the Search Console to verify that Google correctly detects the annotations. Check the coverage reports: if mobile pages appear as indexed but have a canonical pointing to desktop, it means Google hasn't followed your directive. This happens when the content differs too much between the two versions.

What errors should you absolutely avoid?

Number one error: poorly managed pagination. If your desktop page /products?page=2 has a rel alternate to /m/products without parameters, you create a mismatch. The parameters must be consistent between the two versions or explicitly managed in the tags.

Second common error: relative rel canonical instead of absolute. On the mobile version, always use the full URL with protocol and domain. A relative canonical can be misinterpreted if the mobile domain is a subdomain.

How can you tell if your implementation is correct?

Test it with a full crawl of your site (Screaming Frog, Oncrawl). Export all desktop URLs with their rel alternate, all mobile ones with their canonical. Cross-reference the data: each pair should be bidirectional and symmetrical.

Also check that Google Mobile-Friendly Test validates your mobile pages. If a page is flagged as non-mobile-friendly despite having a rel alternate, Google may deprioritize that signal. Technical consistency is critical.

• Ensure each desktop page has a rel alternate to its exact mobile equivalent
• Confirm that each mobile page has an absolute rel canonical to the desktop version
• Audit pagination and URL parameters to ensure symmetry
• Monitor Search Console reports (coverage, mobile usability) for inconsistencies
• Crawl the site to identify orphan URLs or broken matches
• Test a representative sample with Google’s Mobile-Friendly Test