Copywriting for search engines has become less about optimising content, and more about the creation of useful and relevant articles that answer specific questions. Much has been made of the importance of long content and formulas have been presented that supposedly offer the ideal ratios of targeted keywords in relation to the length of a blog. As the Google algorithms have continued to advance at staggering rates (despite ongoing issues with add-ons such as Google News) the importance of creating useful content has increased exponentially.

Our research has shown that detailed, referenced content that is rich in advice and relevant information will rank considerably faster than optimised long-form content. This research  is taken from our own stable of publications and those of our clients and conclusively shows that Google is evaluating contextually the relevance of articles, and using this as a predictor for rankings. Importantly, this isn’t to say that a certain article will rank permanently for a specific search term. In fact, the very opposite appears to be true, with newer and more updated blog posts that cite new information quickly overtaking freshly outdated articles. From the creators and SEO teams point of view of course, it becomes evident that a significant algorithmic change must have taken place but this is usually not the case. In this finding lies an important point around updating articles in line with new information to maintain and improve rankings – but more on that later.

In order to identify the inputs for relevant content, we audited over a thousand high performing blog posts. We defined high performing as being less than three months old and achieving substantial rankings on Google in a challenging category (in this case, a keyword difficulty as defined by HubSpot of 40 or higher.) We then, along with our partners in data analysis, used 153 potential ranking factors to define the potential reasons for ranking or failing to do so. These factors included incidences of grammatical errors, the number of keywords used within the text, the use of inbound links, the number of backlinks and others.

Importantly, the findings were extrapolated and defined by a third party who was retained for the purposes of defining the metrics, without being offered any presuppositions or assumptions. We were careful to ensure that our bias, or any perceived bias, wouldn’t impact on the results. We wanted data we could use, not evidence of what we already thought we knew. It’s important to note that our findings only present a snapshot of a moment in time on Google. We did not allow for industry fluctuations on the SERP or consider any recent or ongoing changes to the algorithm. We also – as was pointed out to us afterwards – failed to take the impact of 401 redirects into account as a measure. This was an oversight.

The findings which follow, we think, help in defining what Google considers to be, “useful content.” We will include a number of assumptions, but the major finding is the importance of depth when creating SEO content.

Key Finding #1 – Reason, Reference, Demonstrate.

Blogs and articles that referenced evidence from reputable locations, and used demonstrative language, ranked considerably higher than other information based blogs. In a recent algorithmic update nicknamed Google Hawk, the search engine changed the way it prioritises rankings for local search. The apparent goal of this update was to offer users a broader range of results on the first page of Google. In other words, rather than a website being able to dominate the front page with AdWords, organic rankings and a local business result on the page, the newly improved algorithm removed or enhanced certain results based on location, and what was already on the page for a specific user. This update demonstrates that Google is seeking to have a smoother search experience for its users, and apparently part of this is ensuring quality results based on evidence are prioritised over sites that have good domain scores through backlinks.

Key Finding #2 – Keywords Are Not a High-Performance Ranking Factor.

No, we did not see this one coming. It ends up that Google has advanced to a stage of relative understanding in relation to articles and the use of homonyms, synonyms and localised language and cultural factors. It doesn’t need to have keywords thrown at it anymore, in fact, if you say what it is you are trying to say, it will probably relate it to the relevant term. Many of the lower performing articles (gradually moving up the SEO food chain) had optimum keyword allocations, but the high performing blogs and articles broke most of our preconceived rules as to how many keywords should be used, and where they should be used. Meta-titles, alt text and even titles were not crucial ranking factors for more than 77% of the high performing blogs.

Key Finding #3 – Length Matters, But Only As Part of the High-Performance Ecosystem.

Creating a long blog post –for the purposes of this research, more than 2000 words– is an important ranking factor, but only if there is considerable depth (refer to key finding #1) in the article. A blog post that has no depth, but is long and uses all the traditional rules such as optimal keyword usage etc, will achieve a level of ranking, but won’t perform at a significantly higher level than average in a short period of time. This again demonstrates that Google notices that an article is long, but doesn’t assign any rank based on that. It doesn’t assume that just because a blog is more than a certain amount of words, that it will be useful and interesting.


This research we consider to be a predicator to the beginning of the end of SEO content creation in its traditional sense. The importance of detail and quality are likely to increase, and opinion pieces and unreferenced articles will achieve rankings based on those qualifiers. Meaning that opinion pieces will be treated as such, while high-quality content that offers advice and information will be prioritised. This means that journalistic quality will become an important factor for all organisations and those websites seeking to achieve page one rankings for challenging keywords and phrases.

It also means that website rankings are now assigned based on more traditional journalistic measures, including the quality of the story, compared to measures with regards to the site. Gary Vaynerchuck made some bold predictions two years ago with regards to the importance of creating journalistic quality content and the importance of making turning your business into a media organisation. Likewise, Dan Lyons from HubSpot, before his very public departure, advocated brand journalism over content marketing. Both of these writers made valid points, but didn’t even begin to touch on just how important quality content was to become from a digital marketing standpoint.

These were the most important findings from our research. We are in the process of creating a report that will list another 16 ranking factors that were defined as crucial by the research. If you would like to receive a free copy of this report, please email admin(at) with the subject, “Research 23”.