Google Onpage Search Results Analysis

Originally published at krisbrower.com. You can comment here or there.

I did some analysis of Google’s search results based only on the way Google display’s its results. The analysis is based on the first ~800 well formed unique searchs in search data that AOL released(here is a list I made of the first 5000 well formed queries). I initially was going to do this analysis on “competitive queries”, but I could not find a long list of completive keywords, so I decided to do it on more random search queries. I posted all of the relevant graphs of my analysis on swivel(and you can download the dataset there). Here are the graphs I found most interesting.

This shows the average number of bold tags in the title of each search result(Each result has a title which is also the external link to the actual page. When the search query or any parts of the search query appear in this title, google automatically bolds it). It seems like this is a very important factor in getting ranked well in Google. Unlike some of the other results this trend was consistent for the first 90 search results.

A shorter description(in words) correlates with better ranking

A shorter title(in words) correlates with better ranking as well

This graph looks at the percent chance there is a “…” at the end of the description indicating that there is more to the description that was clipped. Surprisingly not having the “…” correlates with a better ranking.

A lower page size correlates with better rank for the first 20 search results.

Last here is a graph of the depth of a page.(ie google.com has a depth of 0, google.com/adsense has depth 1, … etc) After the first few results it does not seem like there is very much correlation between depth of the page and ranking.