SEO & Redirects
What is PageRank?
PageRank is a link analysis algorithm developed by Google founders Larry Page, after whom it is named, and Sergey Brin while they were PhD students at Stanford University. Published in their 1998 paper “The Anatomy of a Large-Scale Hypertextual Web Search Engine” PageRank introduced the foundational idea that the importance of a web page can be measured by the quantity and quality of other pages linking to it, treating links as votes of confidence where votes from more important pages carry more weight than votes from less important pages.
The core insight of PageRank is that the web’s link structure encodes valuable information about page quality and importance. When a web author links to another page they are making an editorial judgment, deciding that the linked page is valuable enough to send their own readers to it. Aggregating these editorial judgments across billions of web pages creates a quality signal that is much harder to manipulate than simple keyword matching, the foundation on which Google built its search quality advantage over earlier search engines.
PageRank remains one of the most important concepts in SEO despite being over 25 years old and despite Google having evolved far beyond the original algorithm. The fundamental principle, that links pass authority and that the quality and quantity of inbound links significantly affects a page’s ranking potential, underlies all modern link-based SEO thinking. Link juice, SEO equity, domain authority, and similar concepts are all practical interpretations and applications of PageRank principles.
How PageRank works
The PageRank algorithm models the web as a directed graph, a network of pages connected by one-way links, and calculates a score for each page based on its position in this network.
The random surfer model: PageRank is built on an intuitive model of web browsing behaviour, the random surfer. Imagine a user who starts on a random page and follows links at random, clicking any link on the current page with equal probability. Occasionally, with probability determined by the damping factor, the surfer gets bored and jumps to a completely random page rather than following a link. PageRank models the probability that this random surfer would be on any given page at any given moment. Pages that the random surfer is likely to be on, pages with many links from other popular pages, have high PageRank.
The iterative calculation: PageRank is calculated iteratively. Each page starts with an equal initial PageRank score. Then scores are recalculated repeatedly, each page’s score is updated based on the scores of pages linking to it. After many iterations the scores converge to stable values that represent each page’s relative importance in the link graph.
The formula for PageRank of page A is:
Where d is the damping factor, typically 0.85, T represents each page linking to A, PR(T) is the PageRank of the linking page T, and C(T) is the total number of outbound links on page T.
Damping factor: the damping factor, typically 0.85, models the probability that the random surfer follows a link rather than jumping to a random page. A damping factor of 0.85 means 85% of the time the surfer follows a link and 15% of the time they jump randomly. The damping factor prevents all PageRank from accumulating at pages with no outbound links, dangling nodes, and adds the random jump component that ensures every page has some minimum PageRank regardless of its link position.
PageRank distribution: a page distributes its PageRank equally among all its outbound links. A page with PageRank 100 and 10 outbound links passes approximately 8.5 units of PageRank through each link, the 100 multiplied by the damping factor of 0.85, divided among 10 links. A page with the same PageRank and 100 outbound links passes approximately 0.85 units through each link. The number of outbound links on the linking page directly affects how much PageRank each individual link passes.
Convergence: the iterative PageRank calculation converges because the damping factor ensures scores cannot grow indefinitely. After enough iterations, the original paper used approximately 100 million web pages requiring significant computation, the scores stabilise at values reflecting each page’s true relative importance in the link graph.
PageRank and link equity
PageRank is the mathematical foundation underlying the practical SEO concept of link juice and SEO equity. Understanding PageRank’s mechanics explains the practical rules of thumb that SEOs use when thinking about link equity.
High-PageRank pages pass more equity: links from pages with higher PageRank pass more equity per link than links from low-PageRank pages. A link from a New York Times article, which has accumulated enormous PageRank through countless inbound links from other high-authority pages, passes far more equity than a link from a newly created blog with no inbound links.
Links from pages with fewer outbound links pass more equity: the PageRank distribution mechanism means each link on a page passes an equal share of the page’s total PageRank. A dedicated resource page that links to only 5 external sites passes 20% of its PageRank through each link, a substantial allocation. A page with 200 outbound links passes only 0.5% through each, a minimal allocation even if the page itself has high PageRank.
PageRank flows through the link graph: PageRank is not static, it flows through the web’s link graph as pages link to each other. A page that receives a link from a high-PageRank page accumulates PageRank itself and passes it through its own outbound links. Authority cascades through the link network, high-quality external links to a site’s homepage increase the homepage’s PageRank, which flows through internal links to section pages, which flow to individual article pages.
Internal links distribute PageRank throughout a site: the site’s own internal linking structure determines how accumulated PageRank from external links flows to individual pages. A homepage with many inbound links has high PageRank. Internal links from the homepage to key section pages distribute that PageRank internally. Pages deep in the site architecture, many clicks from the homepage with few internal links pointing to them, receive less PageRank than well-linked pages near the top of the hierarchy.
PageRank and redirects
The relationship between PageRank, and its practical equivalent link juice, and redirects determines how much equity is preserved when URLs change.
301 redirects pass PageRank: when a URL changes and a 301 permanent redirect is configured from the old URL to the new the PageRank accumulated at the old URL through its backlinks is passed through the redirect to the new URL. External sites continue to link to the old URL, their links pass PageRank to the old URL, and the redirect transfers that PageRank to the new destination. This PageRank transfer through permanent redirects is the mechanism that makes domain migrations possible without catastrophic equity loss.
Google confirmed that 301 redirects pass PageRank and that the transfer is effectively complete, though the precise percentage retained versus lost at each hop is not publicly disclosed. The practical implication is that a redirect preserves substantially more equity than no redirect, a 404 error on the old URL passes zero equity while a 301 redirect passes most of the equity.
PageRank loss through redirect chains: each redirect hop in a chain loses some PageRank, the cumulative loss across multiple hops can be significant. The mathematical model suggests that each hop applies the damping factor again, reducing the PageRank transferred by approximately 15% per hop. A three-hop chain might transfer only 61% of the original PageRank, 0.85³, compared to 85% for a single hop. Minimising redirect chain length maximises PageRank preservation.
302 redirects do not reliably pass PageRank: temporary redirects signal to search engines that the original URL will return. Search engines keep the original URL indexed and attribute PageRank to it rather than transferring it to the destination. Using 302 redirects for permanent URL changes means the destination never accumulates the PageRank that the original URL’s backlinks should provide.
NoFollow links and PageRank: links with rel="nofollow" attributes traditionally did not pass PageRank, the nofollow instruction told search engines not to follow the link for ranking purposes. Google subsequently updated its position to treat nofollow as a hint rather than a directive, potentially passing some PageRank through nofollow links as part of its overall link evaluation. The practical implication is that followed links are still assumed to pass full PageRank while nofollow links pass reduced or no PageRank.
PageRank sculpting
PageRank sculpting was a technique used by SEOs to manipulate how PageRank flowed within a site, using nofollow attributes on internal links to prevent PageRank from flowing to low-value pages and concentrate it on high-value pages.
The original concept: if a page has 10 internal links and PageRank flows equally through all of them adding nofollow to 5 of the lower-priority links would, in theory, concentrate more PageRank on the remaining 5 followed links. By strategically nofollowing links to admin pages, duplicate content, and other low-value pages SEOs hoped to increase the PageRank flowing to their most important pages.
Google’s counter: Google updated PageRank handling to prevent PageRank sculpting from working as intended. Rather than redistributing the PageRank that would have flowed through nofollowed links to the remaining followed links Google instead treated the nofollowed links as not existing, neither counting them nor redistributing their share to other links. The PageRank that would have flowed through nofollowed links is simply lost, it does not boost the followed links.
Modern internal linking approach: rather than trying to sculpt PageRank through nofollow attributes modern SEO practice focuses on information architecture, ensuring that the most important pages have the most internal links pointing to them from high-PageRank pages. This is a natural and sustainable approach to concentrating PageRank on important pages rather than the artificial manipulation of PageRank sculpting.
PageRank in modern Google
PageRank in its original form, the simple iterative calculation described in the 1998 paper, is not what powers modern Google search rankings. Google has evolved enormously in the 25+ years since PageRank was introduced.
PageRank as one signal among hundreds: modern Google uses hundreds of ranking signals, content quality, user engagement, technical performance, E-E-A-T, topic authority, freshness, localisation, and many others alongside link-based signals. PageRank is one input into a much more complex system, important but not deterministic. A page with few links but exceptional content quality can outrank a page with many links but poor content.
Spam fighting evolution: the original PageRank algorithm was vulnerable to link spam, artificially creating many inbound links to manipulate scores. Google has developed increasingly sophisticated spam detection, identifying link farms, paid links, and other manipulative linking patterns and devaluing them or penalising sites that use them. Modern link-based ranking signals are much more resistant to manipulation than original PageRank.
Machine learning integration: machine learning systems, particularly Google’s neural matching and BERT language models, have transformed how Google interprets queries and evaluates content relevance. These systems operate alongside but increasingly independently of traditional link-based signals, enabling Google to assess content quality in ways that are not reducible to link counts.
Toolbar PageRank discontinuation: Google previously displayed PageRank scores for pages in its toolbar, a 0-10 scale visible to anyone with the toolbar installed. This external PageRank signal was discontinued in 2016, Google stopped updating the toolbar scores and eventually removed the feature. The discontinuation removed one of the most visible manifestations of PageRank from the SEO landscape, though the underlying concept remains central to how links influence rankings.
Internal PageRank persists: while the toolbar PageRank was discontinued Google continues to calculate internal PageRank-like signals as part of its ranking infrastructure. Google engineers have confirmed that link-based authority signals, essentially modern descendants of PageRank, remain among the most important ranking factors. The concept lives on even as the specific implementation has evolved far beyond the original paper.
Practical PageRank applications for redirect management
Understanding PageRank principles provides a theoretical foundation for practical redirect management decisions, explaining why certain approaches preserve more equity than others.
Why permanent redirects are essential: the PageRank model explains why 301 permanent redirects preserve equity while 302 temporary redirects do not. PageRank flows to a URL’s canonical version as search engines understand it. A 301 signals that the destination is the new canonical, PageRank accumulation shifts to the destination. A 302 signals the original URL remains canonical, PageRank continues to accumulate at the original rather than transferring.
Why redirect chains reduce equity: the damping factor model explains PageRank loss through redirect chains. Each redirect hop applies the damping factor, approximately 85% transmission per hop. A three-hop chain passes approximately 61% of the original PageRank compared to 85% for a single hop. The mathematical model provides the theoretical justification for the practical recommendation to minimise redirect hops.
Why expired domain acquisition adds equity: the PageRank flow model explains why acquiring an expired domain with backlinks and redirecting it to a primary domain increases the primary domain’s PageRank. The acquired domain’s backlinks pass PageRank to it, the redirect transfers that PageRank to the primary domain, increasing the primary’s effective PageRank accumulation.
Why maintaining redirect sources preserves equity: as long as external sites link to a redirect source URL that URL continues to receive PageRank from those links, and the redirect continues to transfer that PageRank to the destination. Removing the redirect source, by allowing the domain to expire or removing the redirect rule, immediately stops the PageRank flow from all those backlinks. The PageRank preservation argument is the mathematical justification for maintaining redirect configurations indefinitely.