Ranking correlation studies 2015 – what do they mean?

Avatar

Rob Marsden

Head of Digital - Earned and Owned


International

In the last couple of months both Searchmetrics and Moz have published their 2015 ranking factor correlation studies. These studies are really interesting and give some good insight into what factors are becoming more and less important in SEO.

They should though be read with caution – as interesting as they might be, they are simply correlation studies and nobody is ever going to reverse engineer Google’s complex and multi-layered algorithms with this type of study. I’m sure that wasn’t the intended purpose of them though and the end result is a set of data that is both interesting and insightful. In the main, many of the findings didn’t come as a surprise to us as we’re out there in the wild doing this stuff every day for a lot of clients. This allows us to spot and act upon trends quickly, design tests around different hypothesis and build tools to help us run our own analysis. There are a few notable points though which I thought would be worth highlighting in a blog post.

User experience

In this year’s study, Searchmetrics added in a complete new section on UX. This isn’t surprising as it’s been talked about a lot over the last year or so and it makes complete sense as Google continues to try and return the best possible results for its users. From an SEO point of view we need to think about the user, the intent behind the query and the content/conversion suited to where they are in their research or buying journey. We need to think about SERP click through rate, page load speed, layout, design, content quality, type, structure and relevance, calls to action, navigation and internal linking. We should aim for:

  • High SERP click through rate
  • Low bounce rate with no Pogo sticking (users returning to the SERP to click on a different result)
  • Multiple pages viewed
  • Long time on site
  • Action taken

Yes, the basics of on-site optimisation are still important but by putting the user at the centre of your world you will, in a lot of cases push everything else into place. The role of the technical SEO has moved on as Search Experience Optimisation becomes ever more important.

Links

Searchmetrics showed that the importance of links had dropped since their last study but were still very important; Moz showed that links were still the highest correlating factors to rankings. Our opinion on this is that good PR and content marketing is still very much a major part of SEO and along with links come many other positive signals – referral traffic, citations, increased brand search, social activity and all the other things that make up the footprint of a strong brand. SEO in 2015 is about being a smart digital marketer and strong SEO performance is the output of all the things you do to promote your brand and build your audience.

An interesting point made in the Searchmetrics study was that links from news sites correlated strongly with rankings. That could be because news sites generally have high volumes of traffic which in turn will create referral traffic, site engagement and other positive signals via the link, it could be the brand effect and that being regularly featured on news sites with a positive sentiment increases brand perception to Google or it could quite simply be that news sites carry a lot of authority which will make the link a powerful one. In reality, it’s probably a mixture of these different things.

Another point worth noting is that Searchmetrics said that mentions of a brand or domain without a link are becoming more of a factor. Interestingly just 2 days ago Eric Enge published the results of a test he ran with help from the IMEC Lab team. The results of the test showed that Google doesn’t seem to be using URL mentions as a ranking factor and if they are then it’s a weak signal. Searchmetrics could have found the correlation because as marketers do all the right things, brand coverage is created along with many other positive ranking signals.

Social

Social signals always tend to correlate well with rankings in this type of study in this case it was no different. There are differing opinions in the SEO community about social media and SEO; some firmly believe that social signals are definitely a ranking factor whereas others don’t believe they are. Our thoughts are that social media is a vehicle to amplify great content which in turn can generate natural links, build an audience, promote your brand and drive traffic and engagement to your site. These are the things that help to drive SEO performance rather than the social signals alone. When reading correlation studies ask yourself what might have come first, the position or the social signals – were the social signals the result of people finding good content that already ranked well?

Content

Searchmetrics found that although this factor has lost weight since the last study, longer content continues to perform better in search than shorter content with the average word count of the top 10 results now being 1,285 (I’m at 966, can I pad this out a bit? I’m joking…). Topical relevance and semantically related terms are much more important (and less risky) than simply repeating keywords into website copy. Going back to what I touched on earlier, if you put your user at the centre of your world and truly understand their intent then your content will reflect this and serve your user with exactly what they’re looking for whilst feeding Google with all the relevance signals it needs.

The full studies are well worth a read, you can find the Moz one here and the Searchmetrics one here.


x
Get the latest digital marketing insights delivered to your inbox