Should you really separate mobile and desktop URLs to avoid duplicate content?

Why is Google still discussing separate URLs in the mobile-first era?

This statement may come as a surprise given that responsive design has been the standard for years. Google continues to document this approach because it is still used by many legacy sites, especially in e-commerce and media.

Separate URLs (m.example.com vs www.example.com) allow for completely different HTML code to be served depending on the device. The browser’s user agent triggers server-side redirection or distinct DNS routing.

How do we technically avoid duplicate content with two versions?

The risk of duplication is real: two different URLs displaying the same content is precisely the definition of duplicate content. Google may index both versions and dilute PageRank.

The technical solution relies on the correct implementation of link rel canonical and alternate tags. The desktop version must point to the mobile version with rel=alternate, and vice versa with rel=canonical. Without this configuration, you create an indexing nightmare.

Which sites still use this architecture today?

This approach is primarily found on legacy platforms where a complete redesign to responsive would be too costly. Giants like Amazon or eBay have long maintained separate mobile subdomains.

Some use cases still justify this separation: sites with radically different functionalities between mobile and desktop, progressive web applications deployed on a dedicated subdomain, or extreme performance optimizations requiring distinct technical stacks.

• Separate URLs = two distinct HTML versions served based on the detected user agent
• Requires strict bidirectional canonical/alternate tags between versions
• Historical architecture that persists on legacy sites unable to migrate to responsive
• Responsive design remains the default recommended approach for any new project
• Mobile-first indexing prioritizes the mobile version, regardless of the chosen architecture

Is this recommendation still relevant or outdated?

Let’s be honest: Google documents a practice it doesn't really recommend anymore. Responsive design with a single URL has been the preferred method since the introduction of mobile-first indexing. This statement is more about legacy support than an encouragement to create new sites with separate URLs.

The problem is that Google maintains this documentation because thousands of sites continue to use this architecture. Completely abandoning these guidelines would create a documentation gap. But don't be fooled: if you’re launching a project today, starting with separate URLs would be a strategic mistake.

What real risks do sites with separate mobile URLs face?

The maintenance complexity is the first pitfall. Every content change must be duplicated across two versions, every technical test must be conducted twice, and each bug may manifest differently depending on the platform.

Configuration errors with canonical/alternate tags are extremely common. I have audited dozens of sites where bidirectional annotations were missing or pointed to the wrong URLs. The result: partial indexing, version cannibalization, and loss of rankings. [To be verified] Google claims to handle these cases gracefully, but the reality shows that annotation errors cause measurable visibility losses.

In what cases is this architecture still justifiable?

Three scenarios still legitimize separate URLs. First case: you manage a massive legacy site where migrating to responsive would require 18 months of development and pose an unacceptable business risk. It’s better to properly maintain the existing setup than hastily execute a poorly conducted redesign.

Second case: you are developing radically different user experiences between mobile and desktop, to the point where shared code becomes counterproductive. Some complex web applications fall into this category. Third case: you are optimizing for markets where bandwidth constraints justify ultra-light versions served on a dedicated subdomain with optimized CDN.

How can you audit whether your separate URL implementation is correct?

First check: inspect the HTML source code of both versions. The desktop page should contain a link rel="alternate" media="only screen and (max-width: 640px)" href="https://m.example.com/page" pointing to the mobile version. The mobile page should contain link rel="canonical" href="https://www.example.com/page" pointing to the desktop.

Use Search Console to verify which version Google is actually indexing. Compare the number of indexed pages with your sitemap. A significant discrepancy often reveals annotation issues. Test with the URL inspection tool to ensure that Googlebot mobile correctly discovers the alternate annotations.

What technical errors break indexing with separate URLs?

The classic error: temporary 302 redirects instead of serving the content directly based on the user agent. If you consistently redirect desktop users to www and mobile users to m, Google may interpret this as cloaking or ignore the annotations.

Another pitfall: annotations pointing to 404 or 301 URLs. If your mobile version points to a desktop URL that has changed without updating the annotation, Google gets lost. Also, check that your separate XML sitemaps (one for desktop, one for mobile) declare the correct annotations in each entry.

Should you migrate to responsive or maintain separate URLs?

If your site generates a stable traffic and the current implementation is working correctly, don’t break what’s working. A poorly prepared migration to responsive can destroy hard-won rankings. Evaluate the ROI: how much does it cost to maintain both versions versus the cost of a complete redesign?

On the other hand, if you notice recurring indexing issues, unexplained ranking drops, or maintenance complexity that slows down your team, plan the migration. Responsive design radically simplifies technical infrastructure, eliminates duplicate content risks, and makes A/B testing easier. Plan for 3 to 6 months for a clean migration with thorough testing.

• Check the presence and accuracy of rel=canonical and rel=alternate tags on all pages
• Audit the HTTP codes of annotated URLs (none should return 404, 301, or 302)
• Compare the number of indexed pages between desktop and mobile versions in Search Console
• Test with Googlebot mobile to confirm that the annotations are detected and followed
• Ensure that the XML sitemaps correctly declare the bidirectional annotations
• Monitor Core Web Vitals separately on both versions to detect performance discrepancies