# Google PageRank: Algorithm, Formula and Worked Example

Understand PageRank—Google's foundational algorithm for measuring page importance through links. Learn its history, how it works, and modern relevance.

Canonical: https://backlinkgrid.com/glossary/pagerank


**Google PageRank** is a link-analysis algorithm that estimates a page's importance from the web's link structure. It is a page-level concept, not the same thing as a keyword's ranking position, Moz Domain Authority or Ahrefs Domain Rating. Google says PageRank has evolved and remains part of its core ranking systems.

## PageRank in plain English

Imagine a reader who follows links between pages. A page receives more importance when other important pages link to it. A link from a page with many outgoing links contributes a smaller share in the basic model than a link from an otherwise identical page with fewer outgoing links.

The calculation is recursive: a page's importance depends on the importance of the pages linking to it. It is solved repeatedly across the graph rather than by counting one page's backlinks in isolation. That is why raw link count alone cannot reproduce PageRank.

## The normalized PageRank formula

For a simplified graph with N pages and no dangling pages:

```text
PR(A) = (1 - d) / N + d × Σ(PR(T) / L(T))
```

- PR(A) is the score for page A.
- d is the damping factor; 0.85 is a common teaching value.
- T represents each page linking to A.
- L(T) is the number of outgoing links from T in this simplified graph.
- The scores sum to one in this normalized version.

You may also see a formula using (1 - d) without dividing by N. That expresses a different scaling convention; do not mix values between conventions. Real implementations also need rules for pages with no outgoing links, commonly redistributing that probability across the graph. Damping alone is not a complete description of dangling-page handling.

## A worked three-page example

Suppose A links to B and C, B links to C, and C links to A. Start with all three scores at one third and use d = 0.85. This is an illustrative mathematical graph, not a simulation of Google's current production system.

| Page | Incoming share before damping | Score after one iteration |
| --- | --- | --- |
| A | C contributes 1/3 | 0.05 + 0.85 × 1/3 = 0.3333 |
| B | A contributes 1/6 | 0.05 + 0.85 × 1/6 = 0.1917 |
| C | A contributes 1/6 and B contributes 1/3 | 0.05 + 0.85 × 1/2 = 0.4750 |

The first iteration is not the final result. Repeating the calculation converges to approximately A = 0.3878, B = 0.2148 and C = 0.3974. The total remains one. Notice that A's importance rises because C, which links to A, becomes important too.

The lesson is about network structure. It does not mean two links from any websites are automatically better than one relevant editorial link, nor does it provide a formula for purchasing ranking improvements.

## Is PageRank still used by Google?

Google's [ranking systems guide](https://developers.google.com/search/docs/appearance/ranking-systems-guide#link-analysis) describes PageRank as an evolved part of its core systems. That does not mean the original academic formula fully describes modern search. Relevance, quality and other systems also influence which results appear for a query.

A high score in an old conceptual model would not make an irrelevant page the best answer to a question. Keep the distinction between link importance and search-result ordering clear when explaining the concept to clients.

## Can you check your Google PageRank score?

Google does not provide the old public Toolbar PageRank score as a current webmaster metric. A third-party checker advertising a PageRank number is not exposing Google's live internal score. Check what it actually measures before interpreting the result.

If your goal is to know where a page ranks for a query, use Search Console performance data. If your goal is to review link sources, use a backlink research tool and inspect source pages. Neither task requires pretending that a vendor's domain score is Google's PageRank.

## PageRank vs DA, DR and link equity

| Term | What it refers to |
| --- | --- |
| PageRank | Link-based page importance in Google's research and evolved systems |
| Domain Authority | Moz's comparative domain metric |
| Domain Rating | Ahrefs' comparative backlink-profile metric |
| Search position | Where a result appears for a query, with context and averaging |
| Link equity | An informal SEO term for value associated with links |

Use our [Domain Authority definition](/glossary/domain-authority), [link equity definition](/glossary/link-equity) and [Ahrefs vs Moz comparison](/comparisons/ahrefs-vs-moz) when explaining those distinctions. There is no reliable conversion formula from DA or DR to Google's internal PageRank.

## What PageRank means for internal linking

Use ordinary, crawlable links to connect useful pages. A hub should link to its supporting articles, and an article should link back to its parent and to genuinely helpful related material. Avoid leaving an important page reachable only through a search form or a click handler.

For example, an audit guide can link to a disavow explanation at the point readers need to understand that decision. A tactics article can link to a relevant outreach template. Those links help navigation and discovery even without calculating a numeric equity score.

Do not remove useful navigation merely to reduce outbound-link counts. The toy model is a teaching device, not a reason to make a site difficult to use. Read [internal vs external links](/backlinks/internal-vs-external-links) for practical implementation.

## Paid links and ranking promises

Google's [link-spam policies](https://developers.google.com/search/docs/essentials/spam-policies#link-spam) distinguish advertising links from links sold for ranking manipulation. Paid links should be qualified. A placement can be useful for relevant visitors without promising search-ranking credit.

BacklinkGrid's directory uses sponsored attributes for paid placements. Its [database bundle](/bundle) is a prospect research file, not a package of PageRank or guaranteed links. Relevance and publisher review remain part of the buyer's work.

## What to read next

Start with [backlink fundamentals](/backlinks), continue to [link equity explained](/backlinks/link-equity-pagerank-explained), or compare [referring domains and backlinks](/backlinks/referring-domains-vs-backlinks). The [backlink quality hub](/backlink-quality) explains how to evaluate an opportunity beyond metrics.

## Original research

The historical foundation is [The Anatomy of a Large-Scale Hypertextual Web Search Engine](https://research.google/pubs/the-anatomy-of-a-large-scale-hypertextual-web-search-engine/). The numerical example above is an original teaching example using a normalized three-page graph; it is not a claim about current Google scores.


## Try the free tool

[Try the PageRank calculator](/tools/pagerank-calculator) with your own data. The tool explains its inputs and limitations and lets you export the result.