Whether we cite the statistic that “more Google searches take place on mobile devices than on computers in 10 countries including the US” or consider the importance placed on Google’s “Mobilegeddon” update, it’s no secret that mobile configuration is a big part of the future of SEO and digital marketing in general.
It could be said that true success in Google is synonymous with having a successful mobile presence. But what kind of mobile presence? When it comes to designing a mobile website, there are three choices: responsive, adaptive and traditional “m.” mobile sites.
Does Google prefer one over another? Is there a difference in preferring a design and using it as a ranking factor?
Before we move forward, let’s quickly establish the difference in the three choices:
- Responsive Design. Often referred to as “RWD” for Responsive Web Design, this design approach uses fluid, proportion-based grids, flexible images and varying CSS style rules to deliver different user experiences to desktop, tablet and mobile devices while maintaining the same HTML and URL structure. The site shrinks or grows according to device.
- Adaptive Design. Referred to by Google as dynamic serving, adaptive design serves different devices using the same URL structure, but it does so by detecting the device and generating a different version of the site’s HTML appropriate for that device. The site has multiple versions that are served through common URLs.
- Separate Mobile Site. Sometimes referred to as mDot (“m.”), this configuration delivers different HTML on separate URLs depending on the device detected. The usual arrangement is to have the desktop site located on the www subdomain and the mobile-friendly site’s pages located on the “m.” subdomain. Pure Oxygen Labs recently reported that 54 percent of the Internet Retailer top 500 brands currently use this configuration.
Google has been nice enough to outline the pros and cons of each design choice in their Mobile SEO Guide’s Responsive Web Design, Dynamic Serving and Separate URL sections.
This is where we get the first hint of which configuration Google prefers, if not uses as a ranking factor. The responsive Web design page distinctly says, “We [Google] recommend using responsive web design because…” and then goes on to list six ways responsive design saves Google resources, delivers improved user experiences and avoids SEO pitfalls like bad redirects and fragmented link presence.
If the primary question is what kind of mobile configuration Google generally prefers, the answer is unequivocal:
Responsive design is Google’s recommended design pattern.
When Google’s big push to reduce duplicate and low-quality content started years ago and canonical tagging became the norm, I often joked that Google will “forgive” a site for bad HTML, but it won’t forgive sites that waste their resources.
I’m reminded of that in their position on responsive design: All else being equal, Google would much rather site designers support the crawling efficiency of RWD’s singular HTML than have many different versions they must index and keep fresh.
More recently (August 18, 2015), Google’s John Mueller used a Google+ post to reinforce that, when pushed, Google recommends responsive:
A good way to make it [a site] work in both worlds [mobile & desktop] would be to have a site that uses responsive web-design techniques to adjust to the size of the user’s device/settings.
Does that put the issue to rest? Well, yes and no. It tells us that Google does prefer responsive design, but it does not tell us if that preference has translated into a SEO ranking factor. Let’s explore that more deeply.
Are There Studies Implicating RWD’s Role As An SEO Ranking Factor?
Ideally, we could reference post-Mobilegeddon, causation-focused case studies where similar-but-different sites with no mobile presence were launched with responsive, adaptive and mDot configurations. These case studies would measure not just the advantage of having a mobile presence, but specifically whether responsive was seen as more positive ranking factor than the other choices.
Two recent studies come close but don’t quite hit the mark:
- 2015 Search Engine Ranking Factors — Moz.com
In a survey of 150 marketing professionals specializing in SEO, the factor respondents felt was the second most influential Domain-Level, Keyword-Agnostic Feature correlating to SEO ranking success was “Use of Responsive Design and/or Mobile-Optimized,” which ranked 6.33 out of 10 (second only to “Uniqueness of content across the whole site”).
While it’s an interesting piece of information, this study is not proof in any way. First, this is a correlation survey, not a causative study. More importantly, it lumps responsive together with simply being mobile-optimized. While the Moz survey is extremely valuable, it does not answer the question we’re asking. To paraphrase, it simply says, “SEO pros feel having a mobile-optimized site correlates to SEO rankings.”
- Proven Responsive Design SEO Benefits [Case Study] – v9seo.com
This case study comes closer to supplying the answer we’re looking for, but not quite. Volume Nine saw a single client with a one-page mDot site benefit from transitioning to a complete, fully responsive design: their organic rankings and traffic increased. At least this is a study rather than a survey and appears to be studying causation rather than correlation; however, it still doesn’t prove specifically that responsive design is a more positive ranking factor than simply having some type of mobile configuration.
Anecdotally, I’ll add that my agency, PM Digital, studied approximately 40 clients’ organic rankings and traffic before and after Mobilegeddon and didn’t see any appreciable gains or losses when comparing responsive vs. adaptive vs. mDot. (Perhaps today’s post is a kick in the pants for me to revisit that data and publish it as a follow-up… )
Is the simple answer that no study has shown that Google prefers responsive design to the extent of using it as a SEO ranking factor? Perhaps our takeaway is simply that a true causative study is needed.
To be thorough, let’s see if we can find other evidence that responsive design (and not just mobile-friendliness) is a positive SEO ranking factor.
Are There Other Sources Confirming RWD As A Ranking Factor?
Two quotes by Google representatives come to mind as relevant — one in 2013 and another just before Mobilegeddon in March 2015:
- In October 2013, Google’s John Mueller responded to a StackExchange question about responsive design saying, “You may see indirect effects (smartphone users liking your responsive site and recommending it to others), but we don’t use that as a ranking factor.”
- At SMX West in March 2015, Google’s Gary Illyes denied that responsive design provides a ranking advantage. According to a write-up of the event:
Mobile-responsive sites are ‘easier for Google,’ according to Gary Illyes, but there is no difference in how responsive and m. sites are treated – meaning, there is no ranking advantage to having a responsive site. It simply reduces the chances of errors with setups that are common with m. sites.
If we are to believe these quotes are still indicative of Google’s current SEO ranking factors (and I do), our takeaway is clearly what Avinash Kaushik said back in May 2014: “If you have a non-mobile friendly website, a responsive design website will certainly suck less. But that’s it. Suck less.”
What Can Be Said
I think the key to finding what can be said about the SEO impact of responsive vs. adaptive vs. mDot site configuration lies in the indirect effects Mueller referenced back in October 2013. If the configuration itself is not a positive SEO ranking signal, what indirect effects of the configuration would mean a positive or negative impact on organic rankings?
- Bad Redirects. Adaptive and mDot sites are often negatively affected by failure to properly redirect based on device. If Google can’t find your mobile site, they can’t index and rank it. Responsive design does not carry this liability.
- Improper Canonicalization. Since mDot pages have completely different URLs from their desktop counterparts, it is important to signal the relationship between two URLs with rel=”canonical” and rel=”alternate” elements or risk confusing Google as to which version to consider for which index, desktop or mobile.
- Slow Pages. This is an area where responsive sites can sometimes be negatively affected by their configuration choice. Unlike most adaptive or mDot configurations, responsive sites load all objects needed by the full desktop site and do not have the opportunity to streamline page assets to match the device. As a result, responsive sites can often be slower to load than other configurations and break the rule that mobile pages should load in under one second.
- High Bounce Rates. Related to slow pages and bad user experience is that poor mobile configuration can often lead to high bounce rates from visitors dissatisfied with what has been delivered to them on their device. If any mobile configuration fails to serve what visitors expect in a timely, easy-to-access fashion, that configuration can be negatively impacted.
- Responsive Configuration — On the positive side, a responsive design is apt to serve mobile, tablet and desktop visitors with a usable version of the site. On the negative site, responsive sites can seldom vary the content delivered. For example, a responsive mobile site may not properly highlight the store finder or customer service information a mobile visitor likely needs.
- Adaptive & mDot Configuration — Where responsive design will always render a usable mobile and tablet experience, there are those who have argued adaptive and mDot configurations often do not properly serve tablet-based visitors, yielding a high bounce rate that negatively affects organic rankings. My contribution to that potential negative impact of adaptive and mDot configuration: Google says, “we generally do not include tablets in the definition” of mobile. If that’s the case, tablet bounce rates would not negatively impact mobile results.
The conclusion? Until proven otherwise by causative studies, we’re left with the famous line from Altered States: “The final truth of all things is that there is no final truth.”
Each configuration carries its own potential SEO benefits and liabilities. If any responsive, adaptive or mDot site is not set up properly, simply using that configuration will not be a positive SEO ranking factor that overcomes that misstep.
With that said, I do think Google’s overall preference is clear: responsive is the way they’d like webmasters to configure their sites. Is it a positive SEO ranking factor? Not yet.
What do you think? Are there studies I haven’t cited or examples you have seen that prove one mobile configuration is an inherently negative or positive ranking factor?