Posted by Paul Williams, Senior Specialist, SEO
The mobile era grows increasingly closer as 2011 carries on. Most of our clients have seen over 1,000% increases in mobile visitors to their sites (via both paid and natural entrances), and several now see 5-7% of visitors as strictly accessing their sites from a mobile browser.
As more users adopt mobile devices—and the habits in using them—the closer we get to a collision course with users' behavior around linking and social sharing. For anyone who's been around for the past decade, you've seen "weblogs" evolve into "blogs," and slowly but surely segue into some of the most important places on the Internet to find, explore, and peruse news and information. Bloggers also contribute immensely to online discourse and provide meaningful curation of links with valuable context.
In fact, many bloggers are the Internet's best link builders. Whether they're using footnotes, descriptive anchor text, hyperlinking a reference in-line, or simply doing the "via" reference, bloggers are good at reporting original sources of quotes, articles, images, etc. Since most blogs and bloggers write within the context of an interest (like technology or sports), these links become even more valuable if their destinations are related to the blog’s interest. But what happens when these influential bloggers move their writing and publishing to mobile devices? And what happens when they consider several of their readers' reading habits—several of which are relegated to "read later" services like Instapaper? (If you’re unfamiliar with “read later” services, they provide the ability to run a bookmarklet across browsers and devices that lets you save an article’s text for future reading.) The seeming conveniences of the mobile age actually become issues for link building.
Conveniences Become Inconveniences
Let's take a simple example to demonstrate this problem. Let's say that as a result of readers' consumption habits—using mobile devices on spotty networks to access and read articles, saving articles for later via Instapaper—some bloggers who are commenting or linking to an article of interest are more likely to reference its printout version. They're prone to do this as a nicety and a convenience because:
- Print versions load faster (no ads, no unnecessary navigation links, images, etc.)
- Print versions typically eliminate pagination that occurs on many newspaper sites by displaying the full article
- Print versions make it easier for Instapaper's bookmarklet (or any other “read later” service) to accurately parse (even though it's darn good as-is)
So what's the problem? Traditionally, search engines—especially Google—have relied on links as a prominent signal in their algorithms. Links typically operate as a form of democratic voting that, in aggregate, can assist in the endorsement of authority for a particular page or site. As we move into the mobile era, where most sites are redirecting mobile browsers to mobile subdomains that offer different pages and URLs (oftentimes faster with more efficient, succinct design), democratic link endorsement runs into a jam: Mobile users are tweeting and blogging with the non-canonical URLs. Even if the linked URL isn't a print version but rather a mobile version of the article—perhaps because the blogger visited the site on a mobile device with which he was writing his post and simply referenced the mobile URL—it's still the non-canonical version of the URL. Maybe search engines are smart enough to detect this disparity, but I doubt it. Most sites typically apply <nofollow, noindex> META tags or disallow instructions for their print pages (oftentimes the best practice to reduce duplicate content).
But herein lies the problem: Time has disallowed the /printout/ directory in its robots.txt—it clearly doesn't want search engines indexing and ranking the print version of its pages. At the same time, this won't stop search engines from finding and indexing the page if it's linked to from an external source. And it has been linked to from an external source. Not only did Kottke link to it, but a number of retweets and reblogs (and even this blog post) have also done so. If there is no META tag on the print version of the page that informs search engines to avoid following or indexing it, nothing stops them from doing so.
Finding a Solution
So there's a predicament concerning new behavioral practices made possible through mediums like mobile devices and social sharing. It could lead to a greater problem for SEO as well as search engine indexation because we're moving into an era that encourages users to visit Web pages on mobile devices and, as a result, the mobilized versions of many pages. How will search engines view these mobile or print-only pages that oftentimes have a poor connection to their canonical twin?
We can do our best to inform search engines of the correct flow for the new linking ecosystem. While there isn't a precise method for dealing with this problem in its early stages, we can assume that by leveraging something like the rel="canonical" URL property, we're strongly urging search engines to consider a mobile or print-only version of a page as a convenience URL for users of various kinds—but not the canonical URL.
Applying the rel="canonical" tactic should direct search engines to take the hint and understand that any link equity flowing into a mobile or print version of a page should be appropriately redirected toward the specified canonical URL. We don't want to irritate the search engines—we want to help them. Until mobile sites and lean, print-only versions of articles are more ubiquitous (which, in time, they will be), we can only use the tools provided to us from search engines as best we can.