Insight into company’s experiment to uncover just what is driving Google rankings

To identify these contributing factors, Searchmetrics designed a purpose-built integrated suite of search analytics tools that access the largest and freshest databases of 100 million keywords and 75 million domains.

EARLIER THIS YEAR, we undertook a project to analyse 10,000 hand-picked top keywords, 300,000 domains and several million backlinks, shares and tweets to see what is driving Google rankings. Since its inception, Google has always tended to release new features, signals and filters for Google.com (and to some extent Google. co.uk) first. International search markets follow suit after a number of months, leading some SEOs to assume outdated tactics still work in these markets.

To show how challenging the international search game has become, we conducted a thorough analysis of Germany’s Google.de rankings - one of Europe’s toughest SEO markets.

We believe these factors to be Google (intemational)-specific and that smart SEOs will see them as the minimum threshold of international SEO in 2012 and beyond. Many of the factors that we suspected, based on recent Google updates, do impact rankings but there are some surprising results.

In order to determine the relationship between various factors and rankings, we computed Spearman’s rank correlation coefficient for each of the 16 factors, as displayed in Figure 1. Factors ranged from social (Facebook shares, Facebook likes, and Tweets) to links (number of links, percentage of backlinks rel=no follow, percentage of backlinks with a stop word, and backlink count) to keywords (keyword in domain name, keyword in URL, keyword in description, keyword in title) to position of keyword in tide to AdSense, image count and word count.

The x-axis represents the size of the correlation coefficient; the longer the bar, the higher the correlation. Higher figures on the x-axis (e.g. Facebook shares) have a positive correlation (the more the better) while lower numbers (e.g. the length of the linktexts) have a negative correlation. We can see that the biggest correlation exists between Facebook shares whilst the lowest correlation occurs with the position of the keyword in the tide. With negative figures, we see the correlation between the length of the title in characters and the amount of text on the page: the less, the better the ranking.

So, what can we aggregate from this information? Here is what we see as the most important findings.
printer friendly LAN_NEWS_24email to someone
  Login to rate

You must be logged in to make comments on this site - please log in, or if you are not registered click here to signup