- Category: Search Analytics - September 2015
Every two years, Moz surveys the opinions of dozens of the world's brightest search marketers and runs correlation studies to better understand the workings of search engine algorithms. With the help of Moz's data scientist Dr. Matt Peters, new data partners and over 150 search marketing professionals, more data points than ever could be analyzed this time for the Search Engine Ranking Correlation Study and Expert Survey.
The study consists of two parts: a survey of professional SEOs, who provided expert opinions on over 90 ranking factors, and an extensive study of correlations between page features and ranking on Google. Ultimately, Moz measured over 170 correlations and collected over 15,000 data points from its panel of SEO experts to gain insight into the factors that may help or hurt a website's visibility in search engines.
The key findings include:
• The correlation between the appearance of keywords on a page and how the ranking is getting weaker. According to Moz, this could be because of Google no longer simply recognizing the keyword but other matching keywords, alternatives, and so on, as well and becoming smarter. It is and remains central for websites to meet the needs of users.
• Contrary to rumors, the number of links on a particular page still does not seem to be most important for the ranking.
• Not quite as strong as the links at page level, however, the links to a domain and subdomain feature a relatively strong correlation.
• The number of social shares that combine a site in itself is still a controversial issue for Moz. By trend, there is a positive correlation here, although Google most likely does not use such signals directly in its algorithm, and therefore this only secondarily benefits SEO.
• Sites with lower bounce rates, higher page views and longer dwell times, are associated with a higher ranking.
• There is a proper correlation between domains that include exactly the keyword and the ranking, which is according to Moz, however, due to other factors than just a special preference for such domains by the algorithm.
• There is a low dependence between the type of top-level domain (.com, .org, etc.) and the ranking on Google.
• While page length, hreflang use and the total number of links are available only in a moderate connection to the Google rankings, Moz determined a low positive correlation when using https. Hence, it could tip the scales between two sites. A clearly negative effect is a long server response time as well as the total length of a URL.
Infographic: The influence of ranking factors in Google’s algorithm
Note that these factors are not "proof" of what search engines use to rank websites, but simply show the characteristics of web pages that tend to rank higher. Combining this understanding with both experience and knowledge of search engine algorithms can help lead to better SEO practices.
A similar study was released by Searchmetrics recently, which we checked out this week as well.
By Daniela La Marca