Skip to main content
SlapMyWeb
Mobile & UX

Mobile Redirect

Sending mobile users to a separate URL. Mostly a legacy pattern; responsive design supersedes it.

A mobile redirect sends phone visitors to a separate mobile URL, typically an m-dot subdomain. It is a legacy pattern: with mobile-first indexing the mobile version is what gets indexed, and maintaining two URL sets splits signals and invites faulty redirects. Responsive design on one URL is the current recommendation.

Definition

A mobile redirect sends phone users to a different URL from desktop users — historically an m. subdomain. Under mobile-first indexing the mobile URL is the one Google indexes, so the separate-mobile-site pattern is now a liability rather than a solution.

Responsive design, where one URL serves every device and CSS handles the layout, is the recommended configuration and removes the whole class of problem.

Why It Matters

Two failure modes are common and both are expensive. Faulty redirects send every mobile visitor to the homepage instead of the equivalent page, so a deep link from search lands nowhere useful. And signal splitting means the desktop and mobile URLs each accumulate their own links unless rel="canonical" and rel="alternate" are paired correctly in both directions.

Example

If a separate mobile site cannot be retired yet, the mobile page needs <link rel="canonical" href="https://example.com/page"> and the desktop page needs <link rel="alternate" media="only screen and (max-width: 640px)" href="https://m.example.com/page">. Every mobile URL must map to its own desktop equivalent, never to the homepage.

Why separate mobile URLs stopped making sense

The m-dot pattern solved a real problem. In 2010 a desktop layout on a phone was unusable, CSS media queries were young, and phones were slow enough that serving a stripped-down document was the only way to make a site work. Splitting the site in two was the pragmatic answer.

Three things removed the reason for it. Responsive CSS matured, so one document can serve every viewport. Phones got fast enough that the performance argument for a separate lightweight site mostly evaporated. And in 2018 Google began moving sites to mobile-first indexing, completed for all sites by 2024 — meaning the mobile version is the one indexed and ranked, not a secondary copy of a desktop original.

That last change inverted the pattern's logic. A separate mobile site used to be a lighter alternative to the real one. Now it is the real one as far as search is concerned, while usually carrying less content, fewer internal links and thinner metadata than the desktop version it was derived from.

The two ways it goes wrong

Faulty redirects. The correct behaviour is that every mobile URL maps to the equivalent desktop page — m.example.com/blue-widget to example.com/blue-widget. What is implemented in practice, often as a single catch-all rule, is that everything redirects to the mobile homepage. Google names this specifically as a mobile usability problem, and it is worse than it sounds: every deep link from search lands on the homepage, so the visitor arrives having asked a specific question and is handed the front door.

Split signals. Two URLs for one piece of content accumulate links, shares and engagement separately unless they are explicitly paired. The pairing is a rel="canonical" on the mobile page pointing at the desktop URL, plus a rel="alternate" on the desktop page pointing back. Get either direction wrong and the two versions compete rather than combine.

A worked example

Say example.com/pricing has a mobile counterpart. The desktop page declares the alternate:

<link rel="alternate" media="only screen and (max-width: 640px)"
      href="https://m.example.com/pricing">

And the mobile page declares the canonical back:

<link rel="canonical" href="https://example.com/pricing">

Both are required. The canonical alone tells Google which version to index but leaves it to discover the pairing; the alternate alone announces a mobile version without consolidating the signals.

Then check the redirect actually maps path to path:

curl -sI -A "Mozilla/5.0 (iPhone; CPU iPhone OS 17_0 like Mac OS X)" \
  https://example.com/pricing | grep -i location

location: https://m.example.com/pricing     # correct
location: https://m.example.com/            # the failure mode

Migrating off it

The end state is one responsive URL. The migration is a standard site move and the order matters.

Build the responsive version at the desktop URLs first and verify it on real devices, because you cannot retire the mobile site until the single URL genuinely serves phones well. Then remove the user-agent redirect, so both URL sets serve the same responsive page. Then 301 each mobile URL to its desktop equivalent — path to path, never to the homepage — and remove the canonical and alternate pairing, which now has nothing to pair. Keep the redirects permanently: links to m-dot URLs will exist for years.

Expect a settling period. Google has to recrawl both URL sets and consolidate them, which takes weeks rather than days on a large site.

How SlapMyWeb checks this

Every SlapMyWeb audit requests the page with a mobile user agent as well as a desktop one and compares what comes back. A redirect that fires only for the mobile request is reported with its destination, so a rule sending every phone visitor to the homepage is visible immediately rather than after someone notices the traffic. The audit also checks that a mobile URL declares a canonical pointing at its desktop equivalent, and that the desktop page declares the matching rel="alternate". You can reproduce the redirect half yourself with the HTTP header checker, and generate corrected rules with the redirect generator.

Know the term.
Check your own site.

A free audit tells you whether this is currently costing you score points — and exactly what to change.

Run a free audit
Free foreverNo signupResults in 30s