PageRank and CheiRank - Authorities and Hubs

Audisto calculates PageRank, CheiRank and 2D-Rank within the crawled link graph for all URLs found during a crawl. This allows you to comprehensively analyze the structure of your website and to quickly identify negative effects due to PageRank wasted by internal redirects, error pages, noindex links and so on. Elimination of erroneous PageRank waste can create a considerable uplift for your projects. Please keep in mind that passing PageRank to other domains, within limits, is completely normal for the internet.

Shows the Audisto PageRank overview report including the PageRank flow diagram.

PageRank is a link analysis algorithm to measure the relative importance within a set of URLs by weighting incoming links of the link graph. A high PageRank is a good indicator for an important authority-page. The sum of all PageRank values is one. Audisto shows PageRank as percent-values to make them more readable.

CheiRank is a link analysis algorithm to measure the relative importance within a set of URLs by weighting outgoing links of the link graph. A high CheiRank is a good indicator for an important hub-page. The sum of all CheiRank values is one. Audisto shows CheiRank as percent-values to make them more readable.

2D-Rank combines PageRank and CheiRank to identify the most important URLs. 2D-Rank is calculated by plotting all URLs on a two-dimensional area with Page- and CheiRank indexes as axes. Beginning at (0,0) and wandering simultaneously along the axes, all URLs are numbered in the order of appearance.

In the past we only calculated metrics for HTML pages and redirects using streamlined algorithms for PageRank and CheiRank. As of June 2019 we started using more advanced algorithms that allow us to calculate PageRank and CheiRank in considerably more complex graphs and for all URLs and media types within a crawl.

Example of complex linkgraph with different link types and document types

By using more advanced algorithms we address the fact that search engines also crawl and index non HTML media types like images, videos, audio files, PDFs and machine readable resources like RSS feeds.

Example of different media types in a link graph

The new algorithms allow us to calculate PageRank leakage to dangling URLs like external URLs, URLs that are blocked in robots.txt, and uncrawled URLs. We can also calculate the PageRank bound by URLs with a specific HTTP status code, indexability status or isolation level or on specific hosts.

Screenshot of PageRank details for specific URLs by type

Tip: Consider that you can define clusters for all page types and ranks will be calculated for each segment as well.