Is it really necessary to have a separate sitemap for mobile-first indexing?

Why is this clarification on mobile sitemaps needed now?

Since the gradual rollout of mobile-first indexing, confusion persists among practitioners: should two separate sitemaps be submitted, one for desktop and one for mobile? Mueller's statement provides a clear answer. One sitemap containing the desktop URLs is sufficient, even though Google primarily crawls and indexes your mobile pages.

This approach is based on a simple but often misunderstood principle: the canonical URL is that of the desktop in your Search Console and indexing reports. Mobile users are then redirected via user-agent detection to m.example.com or example.com/mobile if your architecture allows it. Google treats this redirection as transparent for indexing.

How does Google technically manage this dual version?

The crawler now always uses a mobile user-agent (Googlebot smartphone) to discover and assess your content. When it encounters a desktop URL in your sitemap, it crawls it with its mobile agent and automatically follows the 302 redirect to the mobile version if it exists.

The mobile version's content then becomes the basis for ranking assessment, but the URL displayed in the SERPs remains that of the desktop. This dissociation between the displayed canonical URL and the content actually indexed causes frequent misunderstandings in technical audits. Many third-party tools only capture the desktop version, creating a discrepancy with what Google actually sees.

What are the implications for responsive sites versus separate mobile?

For a responsive site (one URL serving both desktop and mobile), this directive does not change anything: you continue to submit your normal URLs in your sitemap. Google crawls with a mobile user-agent and directly obtains the right content via adaptive CSS.

For a separate mobile architecture (m.example.com or a /mobile/ subdirectory), the instruction becomes critical: do not create a second sitemap with the m.example.com URLs. Only submit the sitemap containing the www.example.com URLs. Google will follow the user-agent based 302 redirects to reach the mobile versions. This approach avoids duplication in Search Console and simplifies performance tracking.

• One XML sitemap containing the desktop/canonical URLs is sufficient for all types of architecture
• User-agent redirects to mobile versions are automatically followed by Googlebot mobile
• The desktop URL remains canonical in Search Console even if mobile content is indexed
• Responsive sites do not have any specific actions to take
• m.domain.com architectures should ensure that the user-agent 302 redirects work correctly

Is this directive consistent with field observations?

Overall yes, but with important nuances depending on the chosen mobile architecture. Responsive sites have never faced major issues since they serve a single URL. However, configurations with a separate mobile domain (m.example.com) sometimes show discrepancies between what Google claims and what is observed in the logs.

Specifically, some separate mobile sites find that Googlebot crawls the m.example.com URLs directly without using the redirect from www.example.com, especially if these mobile URLs are directly linked from external sources. [To verify]: Does Google really always follow the desktop → redirect → mobile path, or does it also directly index mobile URLs discovered in natural crawling? Server logs suggest the latter hypothesis in many cases.

What practical risks does this approach pose?

The main risk concerns the consistency between desktop and mobile versions. If your mobile version has severely reduced content (a common practice a few years ago to lighten page weight), Google will index this reduced content while displaying the desktop URL in the SERPs. The result: loss of rankings on queries where the richer desktop content performed better.

Another critical point: Search Console data can create confusion. Reports display desktop URLs, but performance reflects the indexed mobile content. When you analyze a traffic drop on a desktop URL in GSC, you must always check what Google actually sees by inspecting it with the mobile mode URL inspection tool. The mismatch between the reported URL and the evaluated content is not intuitive.

In what cases might this rule not apply optimally?

Sites with dedicated mobile applications (App Indexing) find themselves in a gray area. If you have a referable iOS/Android app and a mobile web version, the logic of a single sitemap becomes less clear. Google then recommends specific annotations (app-indexing tags) but does not always clarify how to articulate web sitemaps and app content.

AMP architectures represent another edge case. If you serve AMP pages on mobile (via redirection or dynamic serving), should they be mentioned in the sitemap? The official doctrine says no for standalone AMP (Google discovers them via the rel=amphtml tag), but some SEOs observe better performance by explicitly including them. [To verify] with A/B tests on your own index.

What should you concretely check on your site?

First action: audit your current sitemap to confirm that it only contains canonical desktop URLs (or unique URLs if you are responsive). Open your sitemap.xml file and scan: no m.example.com or /mobile/ URLs should appear if you have a separate architecture. These URLs should only be discovered via user-agent redirects.

Second critical verification: test your mobile redirects with various user-agents. Use curl or a tool like Screaming Frog configured in mobile mode to simulate Googlebot smartphone. The HTTP response must be a 302 (temporary) to the mobile version, never a 301. A 301 would signal to Google that the mobile version is permanent, which can create canonicalization conflicts.

What mistakes should you absolutely avoid?

The most common mistake: creating two separate Search Console properties (one for www.example.com, one for m.example.com) and submitting a sitemap for each. This practice fragments your data and can send conflicting signals to Google about which version to prioritize. Consolidate everything in the property representing your main domain.

Another pitfall: neglecting content parity between desktop and mobile while thinking that "Google indexes the desktop URL so it doesn’t matter." Incorrect. Google evaluates and ranks according to mobile content. If your mobile version hides text, reduces images, or omits entire sections, these elements do not exist for ranking. Always check with the mobile mode URL inspection tool what Googlebot actually sees.

How to monitor that everything is functioning correctly?

Implement a regular server log monitoring filtering specifically for Googlebot smartphone requests. You should observe that the bot crawls your desktop URLs and then follows the 302 redirects to mobile. If you see massive direct crawling of m.example.com URLs without passing through www.example.com, it indicates that Google has discovered these URLs through other means (external links, indexing history).

Also use the coverage report in Search Console to identify any mobile URLs indexed directly despite your precautions. If m.example.com URLs appear as indexed while they should be transparent, investigate: check canonical tags, redirects, and internal links that may point directly to those mobile URLs.

• Ensure your XML sitemap contains only canonical desktop URLs (or unique ones for responsive)
• Test the user-agent redirects with curl or Screaming Frog in mobile mode (they should return 302)
• Consolidate into a single Search Console property representing your main domain
• Audit content parity between desktop and mobile versions using the URL inspection tool
• Monitor server logs to confirm Googlebot smartphone’s crawling behavior
• Eliminate any internal links pointing directly to m.example.com URLs from your desktop pages