Analyzing Google SERPs across Industries and Search Volume
Moz conducted an analysis to study Google’s SERPs across different industries and search volume. The study used their Ranking Factors study data of 2013 to create several buckets.
The study utilized top 50 results based on around 15,000 Google keywords and 100 factors, which included anchor text, links, social signals, and on-page factors, among many others. The mean Spearman correlation was then calculated between the search position and the factor. Usually, higher correlation indicates higher relation of a factor to higher ranking as compared to a lower correlation. However, this does not indicate causation.
Search volume was the first to be analyzed in the study. Three buckets based on AdWords data of US local search volume on a monthly basis were analyzed. The volumes were split into lower than 5,000 searches, 5,000 to 15,000 searches, and over 15,000 searches, all on a monthly basis.
The results indicated that overall page authority along with domain authority and EMD or Exact Match Domain rise together with search volume. This may be due to larger and more authority commanding sites getting a lot more targeted by higher number of search queries.
However, this does not indicate higher correlation with search position. One needs to calculate the mean Spearman correlation for each bucket to find out how search ordering is impacted. A table was used to display factors based on links, brands, social signals, and factors based on keywords.
The table revealed that correlations rose significantly with search volume for factors such as brand, link, and social media. Correlations remained constant for factors related to keywords such as keywords used in names of domains or used on the page.
It seems that Google’s algorithm related to relevance between keywords and documents remains the same for both low-volume and high-volume queries. However, there may be more discrepancy in link or social signals metrics in the SERPs for queries with higher volumes as compared to those with lower volumes.
Google’s ranking algorithm can easily rely on several signals for popular search queries that will provide more relevant pages with several links. However, the algorithm will have fewer signals to work with low volume searches and hence there will be a decrease in correlation. The same analysis can be repeated for various categories in AdWords such as median page authority, median domain authority, etc.
Results from the study suggest that Google may be using the same algorithm related to document relevance for head as well as tail queries. However, link metrics seem to predict SERPs better from popular queries as compared to tail queries. Industries that receive informational and broader queries possess higher correlations as compared to industries that receive queries that are more specific.
