Google Draws a Line on Mobile: Misconfigured Sites Will Rank Lower

Every phone user knows the moment, and now it has a price tag. You tap a search result for a specific page (the program, the article, the person) and land instead on a homepage that discarded your destination. Courtesy of a mobile redirect that recognizes your device but not your intent.

Google announced on June 11 that this pattern, along with its siblings, will begin costing smartphone rankings. Sites with faulty mobile redirects, mobile-only error pages, and media that phones cannot play should expect to drop in smartphone search results.

The announcement is typically specific about the failures and typically clear about the timing. Changes roll out ‘in the near future.’ By the pattern we have documented for seven years, that means this is the notice period. The audit list is published. The organizations that treat announcements as project calendars have their assignment. Here it is, unpacked, with the architectural moral that has been building since 2011 attached.

The itemized failures, and where they live

• Faulty redirects, the headline offense. The desktop page exists. The mobile detection fires. And the redirect sends every phone, regardless of requested URL, to the m-dot homepage. The visitor’s intent is discarded at the door. Google’s guidance is blunt. If a mobile equivalent of the requested page does not exist, serving the desktop page beats hijacking the visit. Note where this failure structurally lives: in the redirect plumbing of separate mobile sites. Responsive architectures cannot commit it, because there is nothing to redirect.

• Mobile-only errors. Pages that work on desktop but greet phones with errors. Misconfigured detection, m-dot sections that lag the main site’s content, the parity drift our m-dot analysis predicted, or device targeting that excludes some phones entirely. The desktop team never sees these failures, which is why they persist. The smartphone crawler now sees them for you, with consequences.

• Unplayable media. Video delivered in formats phones cannot render. In 2013 that means, overwhelmingly, Flash-embedded players still serving the audience our migration-triage coverage addressed in 2010. Three years of notice on that one. The enforcement is now search-visible.

The audit, runnable this month

The verification is mechanical and worth doing URL by URL rather than by assumption.

Take your top fifty organic landing pages from analytics. Request each from an actual phone, plus a throttled emulator for breadth. Confirm the URL you asked for is the content you received. Deep links must land deep.

Then the error sweep. Webmaster tools’ crawl-error reporting now separates smartphone errors, and that report is precisely this announcement’s dashboard. Read it, fix what it names, watch it trend.

Then media. Every page carrying video gets the phone test, and every Flash-only player found gets the HTML5-with-fallback treatment that has been standard advice here since the Thoughts on Flash triage.

For m-dot operators, one more. Walk the redirect map itself, page class by page class, because the mapping table nobody has reviewed since launch is where these failures breed.

Budget: a day or two for most institutional sites. Against ranking exposure on the fastest-growing device class, that is the cheapest insurance this quarter offers.

The running tally, updated again

Regular readers will have anticipated this section. The failure classes Google just itemized are not random UX sins. With the exception of unplayable media, they are the specific failure modes of separate-mobile-site architecture. Redirect plumbing. Parity drift. Configuration surfaces that desktop teams cannot see.

Our 2011 framework gave the m-dot its honest lane, sharply transactional experiences and deliberate bridges, and that lane remains. But the tally keeps filling on one side. Dual maintenance forever. Parity decay. And now itemized ranking exposure for the plumbing’s signature failures, against responsive’s structural immunity to the whole category.

When the fall’s redesign conversations arrive, this announcement belongs in the deck, not as fear but as math. One architecture requires perpetual vigilance against penalties the other cannot incur. Vigilance has a staffing cost. Immunity has a build cost, paid once. Price both honestly and the conversation usually ends the same way ours have been ending since the Boston Globe shipped.

The tablet wrinkle, before it becomes next year’s item

One configuration class deserves an early paragraph: tablet handling. It sits next to this announcement and will, on current trends, join it.

Sites running device detection now face a three-way sort: desktop, tablet, phone. The m-dot setups that treat tablets as phones serve a stripped experience to screens that wanted the full one. Those treating tablets as desktops serve mouse-era interaction to fingers. The iPad-era scramble our 2010 coverage predicted, still unresolved wherever detection rules calcified.

Google’s current announcement addresses smartphone results specifically. But the logic generalizes: serve the experience the device can actually use, or lose visibility on it. The audit above should log tablet behavior while it is in there.

The responsive resolution handles the three-way sort the same way it handled the two-way one, by refusing to sort devices at all. Detection taxonomies age. Canvases don’t. Every device the parade introduces re-teaches it, and the parade, per every manufacturer roadmap, is nowhere near done.

The pattern behind the enforcement

Close with the observation that keeps confirming itself, because each instance strengthens the planning value. Google keeps converting user-experience principles into ranking enforcement, and it keeps publishing the list first. Speed in 2010. Setup standards piling up since. Now mobile experience, itemized, with the smartphone crawler as inspector.

The conversions are never surprises to readers of the announcements. The enforcement never punishes organizations that were serving users well already. That is the tell about what the program actually is. The algorithm is absorbing, one itemized failure at a time, the definition of a decent experience.

Serve users well and the enforcement calendar is someone else’s problem. That has been the durable strategy through every conversion, and the mobile edition changes nothing except the urgency for the sites that have been deferring since 2011. The line is drawn, the list is published, the audit takes a day. The near future, as Google likes to call it, is the deadline. It usually arrives on time.

And a scheduling note for the deferral-prone. Fall traffic season sits ten weeks out. The audit run this month fixes findings before the year’s most valuable mobile visitors arrive. The audit run in October fixes them during. Same day of work, very different invoice. The calendar is the argument.