Google's PageRank Algorithm

PageRank is Google's method of assigning a numerical value to the general importance of a web page that is largely based on the number and quality of the links pointing to that page. The concept of PageRank was the foundation for Google's original approach to ranking, as expressed in the doctoral thesis by the company's founders Larry Page and Sergey Brin when they attended Stanford University. Their theory was that the quality of any webpage could, in part, be evaluated in terms of the number and quality of the links from other webpages that point to it. The theory equated each link as a "vote" for the page being linked to. But instead of a straight numerical count of these links, as other search engines had tried, PageRank also considered the quality of each link.

The original PageRank formula: For any Page A which is linked to by pages T1 through Tn, the PageRank of Page A is determined by the following equation:

PR(A) = (1-d) + d (PR(T₁)/C(T₁) + ... + PR(T_n)/C(T_n))

where PR(A) is the PageRank of the page "A" which is being calculated,
PR(T) is the PageRank of page "T" where each incoming link found by Google resides (noted as sub "1" through "n"),
C(T) is the total number (count) of links on each page "T" above,
and d is a damping factor which compensates for endless loops, etc.

What That Formula Means

For non-math majors, the above formula means that the PageRank value passed by a link is approximately equal to the PageRank score of the page on which the link originates, divided by the number of links on that same page. So, a link from a page with a modest PageRank score but which contains relatively few links can pass on more PageRank value than a link from a page with a relatively higher PageRank score that contains many links. Each web page's PageRank score is then the approximate sum of the PageRank value passed by the links pointing to it.

The exact formula Google uses for calculating PageRank has changed significantly in recent years, but for webmasters this is still the best way to think of PageRank in terms of their own websites. Since the actual algorithm used by Google results in a number with a much broader range of precision than the simple 0 to 10 scale shown in the Google Toolbar, Google must periodically adjust the scale by which their internal PageRank value is converted for display in the Toolbar to account for the ever-growing number of webpages in their index. This is why you'll sometimes see the Toolbar PageRank score of your pages fall a point or two after an update, even though the links pointing to that page appear to have remained unchanged and you see no changes in rankings.

Google engineer Matt Cutts has recently discussed how Google's internal PageRank system has changed recently. He mentions that they now assign different weights to different kinds of links. It's reasonable to assume that menu links, footer links, blogroll links, and similar sitewide links do not pass as much PageRank as links that are embedded within the primary content of a webpage. In addition, Google will often lower the PageRank of low-quality pages. Pages that mainly consist of links, such as link exchange pages and website category pages, will often have no PageRank score at all in the Toolbar PageRank display, which is a strong indication that such pages are largely ignored. Further, some penalties can result in a page or an entire website losing the ability to pass PageRank. This penalty is usually imposed for a serious violation of Google's Guidelines, such as selling links.

Does PageRank Matter?

The value of PageRank in affecting the ranking of a particular page is still very high, but not quite as high as it used to be, which has caused many people to recklessly speculate that PageRank is irrelevant or worthless. In addition, Google has been giving increasing value to the text used within the link's anchor text - that is, the text inserted between the <a> and </a> tags in the HTML code, as well as the text on the page where that link resides. But since you still need a link to get the value of the anchor text boost, which would also contribute to a page's PageRank score, the two are very connected in the ranking potential of every web page. So, while PageRank and anchor text are distinct ranking factors used by Google, the simple fact remains that you can't have one without the other.

Further, the depth and frequency with which Google crawls a site is largely determined by PageRank. This generally means that the more PageRank you have circulating in your site, the more pages from your site will be indexed and the more often they'll be re-crawled and updated in the index. So, you bet PageRank still matters. It matters a great deal.

Toolbar PageRank

What confuses many webmasters is that the PageRank display in the Google Toolbar for Microsoft Internet Explorer and Firefox has two limitations of which you need to be aware. First, as noted above, the Toolbar is only capable of displaying whole numbers from 0 to 10. Google's internal PageRank system has a much finer degree of gradation. It is widely thought that the Toolbar PageRank scale is exponential. That is, each step of Toolbar PageRank is not of equal value. The difference in the value of each step is progressively higher. The second limitation is that the PageRank scores displayed in the Google Toolbar is taken from a database that is only updated every 3-4 months. Google's internal PageRank database is updated continuously as they crawl the web and discover new links. This is one reason why you see Google's rankings change so often and why you can't simply look at the Toolbar PageRank scores of different pages and determine why one page outranks another without considering many other important ranking factors. Note, too, that all of the online tools for displaying PageRank get their information from this same Toolbar database. The only difference is that some of them allow you to check the information on more than one of Google's datacenters, so you can sometimes see if the Toolbar database is in the process of being updated. But only Google has access to their internal PageRank data, which is the information that really matters in this regard.

One last thing you need to know is that Google is constantly changing their methods of counting or ignoring links, and how PageRank is utlimately calculated. While the basic premise of the original formula still forms the foundation of PageRank, the evolution of the web and the need to apply an ever-increasing level of sophistication in determining the quality of individual links and pages means that the effectiveness of simpler methods of building links as a shortcut to better rankings will continue to diminish. Link exchanges, text link ads, blog rolls, footer links, and directory submissions are currently in Google's crosshairs. So your link building efforts should focus on getting more natural links from quality websites whose main topic is related to yours. While it is true that Google's method of calculating PageRank has evolved to include some page quality factors in addition to links, it's still links that matter most for PageRank. And PageRank still matters a great deal in how your site performs in Google.

This page was last updated on October 19, 2011