Google’s Reasonable Surfer Patent: Why Not Every Backlink Counts the Same

Picture two links on the same page of the same website. The site has the same Domain Rating either way. One link sits in the first paragraph of the article. The other sits in the footer. Most link reports would count them as equal.
This reasonable surfer model reflects the fact that not all of the links associated with a document are equally likely to be followed. Examples of unlikely followed links may include ‘Terms of Service’ links, banner advertisements, and links unrelated to the document.
— Google patent US 7,716,225, quoted by Paddy Moogan at Moz · Moz: The Anatomy of a Link
Google described a system that says they are not. In 2004, Google filed a patent that gives each link on a page its own weight, based on how likely a real person is to click it. SEOs call it the “reasonable surfer” patent, and it is one of the most useful patents a link builder can read.
In this article, I walk through what the reasonable surfer patent says, how the system works, and what it means for the links you build. This page is also the hub for my Google link-patents series. After the reasonable surfer walkthrough, you will find shorter sections on related patents: seed sites and link distance, link timing, affiliated sites, link-spam detection, anchors, and brand searches and clicks.
The history behind why Google slowed ranking feedback is in a separate article: The Problem Google Was Solving: The Spammer’s Feedback Loop.
You can also watch the full video breakdown of the reasonable surfer patent below.
In this guide
- The reasonable surfer patent (US 7,716,225)
- Seed sites and link distance (US 9,165,040)
- Historical data and link timing (US 7,346,839)
- Affiliated sites and node independence (US 8,719,276 and US 7,783,639)
- Link-spam detection (US 7,509,344)
- What an anchors patent actually describes (US 8,959,093)
- Brand searches and clicks (US 7,962,462 and US 8,661,029)
- What this means for your link building
The Patent at a Glance
The reasonable surfer patent is US 7,716,225, titled “Ranking documents based on user behavior and/or feature data.” Google filed it in June 2004 and was granted it in May 2010.
| Detail | Value |
|---|---|
| Patent number | US 7,716,225 |
| Filed | June 17, 2004 |
| Granted | May 11, 2010 |
| Inventors | Jeffrey A. Dean, Corin Anderson, Alexis Battle |
| Assignee | |
| Later versions | Later US applications filed in 2010, 2012, and 2016, including US 9,305,099 (granted April 5, 2016) |
One name on the list stands out. Jeff Dean is one of the most senior engineers in Google’s history. When his name is on a ranking patent, it is worth reading closely.
From the Random Surfer to the Reasonable Surfer
To understand this patent, you need to know what came before it. The original PageRank model, described in Google’s first research paper, imagined a “random surfer.” This imaginary person lands on a page and clicks any link on it at random. Every link on the page gets an equal share of the vote.
That was a good start, but real people do not browse that way. Nobody clicks a Terms of Service link as often as a link in the middle of the article they came to read.
The reasonable surfer patent replaces the random surfer with a more realistic one. A reasonable surfer follows some links much more often than others. The patent even names the kinds of links people rarely click: Terms of Service links, banner ads, and links that have nothing to do with the page they sit on.
That single change matters a lot. If links are weighted by how likely people are to click them, then where a link sits, and what surrounds it, can change how much value it passes.

How the System Works
The patent describes a system that learns which links people click, then uses that to weight every link. It works in four steps.
- Watch what people actually click. The patent gives a simple example. A page, call it W, links to three pages: X, Y and Z. People click the link to X twice and the link to Z once, and nobody clicks Y. Each click counts as a positive example for that link, and each link that got passed over counts as a negative one.
- Record the features of each link. The system looks at where the link sits, what it looks like, and what it says. It also looks at the page the link is on and the page it points to.
- Train a model that predicts clicks. The patent names common machine learning methods for this, such as decision trees and logistic regression. The model learns which features make a link more or less likely to be clicked.
- Turn click probability into link weight. A link people are likely to click gets a high weight and passes more value. A link people ignore gets a low weight and passes less.

The patent also says the model is updated over time. That makes sense, because links come and go, and the way people browse keeps changing.
The Worked Example: Weight Changes the Outcome
The patent includes a small worked example with three pages, A, B and C, that link to each other. Each link gets a weight between zero and one, based on how likely people are to click it.
| Link | Weight |
|---|---|
| B to C | 0.9 |
| A to B | 0.6 |
| C to A | 0.5 |
| A to C | 0.4 |

When you run the numbers, page C ends up with the highest rank of the three, at about 0.281, compared with 0.237 for A and 0.202 for B.
Page C does not win just because two links point to it. It wins because one of those links, the 0.9 link from B, is the kind people actually use. The weight changes the outcome, and that is the whole idea of the patent.
The Features Google Can Measure
The patent lists many features the model can look at. They fall into three groups.
The link itself
The first group covers the link. It includes the font size of the anchor text and where the link sits: in the running text, in a list, in a sidebar, in the footer, or above or below the first screen. If the link is in a list, its position in that list counts too.
The model can also look at the link’s color and styling, including grey text or text the same color as the background. It can look at how many words are in the anchor text, which words they are, and how commercial they sound. It can tell an image link from a text link, read the few words right before and after the link, and check whether the link stays on the same site.
The page the link sits on
The second group covers the linking page. It includes how many links are on that page, the words in the page and its headings, and whether the page’s topic matches the topic of the anchor text.
The page being linked to
The third group covers the target page. It includes the words in its web address, how long that address is, and whether it is on the same domain as the page linking to it.

The Rules the Patent Spells Out
The patent gives examples of rules the model might learn. Some apply across the web, and some apply only to certain pages or sites.
General rules
- Links with bigger anchor text get clicked more than links with smaller text.
- Links near the top of a page get clicked more than links near the bottom.
- Links between pages on related topics get clicked more than links between unrelated pages.
Page-specific rules
- Links on pages with popups get clicked less.
- Links to web addresses containing the word “domainpark” get clicked less.
- Links to .tv domains get clicked less.
- Links to web addresses with several hyphens get clicked less.
- Links under the “More Top Stories” heading on cnn.com get clicked more.
The second list is the more interesting one. It shows the model is not locked into one fixed set of rules. It can learn how people behave on a single site, and even on one section of one page.
What This Means for Your Link Building
Here is what the patent suggests for anyone building links. None of these are guarantees, but each one follows directly from the features and rules above.

1. Aim for placement inside the article, and early
A link in the first few paragraphs of the body beats a link in the author bio, the sidebar, the footer, or a resources list at the very bottom. When you negotiate a placement, ask for a spot in the main content.
2. Relevance does double duty
A link on a page about the same topic makes sense to Google, and it is also more likely to get clicked by a real reader. That is one more reason to train your prospecting around relevance instead of raw metrics.
3. Watch how many links are on the page
A link sitting among 150 others has to fight for attention. A page with a handful of well-placed links gives yours a better chance of being noticed and clicked.
4. Keep your anchors natural
The patent measures how commercial the anchor text is. A descriptive phrase that fits the sentence reads better to people and looks better to search engines. I cover this in more depth in why anchor diversity still matters.
5. Pay attention to the sentence around your link
The words right before and after a link are one of the features the system can measure. If you brief a writer, brief them on the whole sentence, not just the anchor.
6. Use the simplest test of all
Ask whether a real reader would click the link. If the answer is no, assume it passes less value than its metrics suggest. Links that pass this test are also the foundation of a natural backlink profile.

Seed Sites and Link Distance
A link’s value also depends on where the linking page sits on the web. US 9,165,040, titled “Producing a ranking for pages using distances in a web-link graph,” describes a system that ranks pages by how far they are from a set of trusted seed pages.
| Detail | Value |
|---|---|
| Patent | US 9,165,040 |
| Title | Producing a ranking for pages using distances in a web-link graph |
| Inventor | Nissan Hajaj |
| Filed | October 12, 2006 |
| Granted | October 20, 2015 |
| Assignee | Google LLC |
| Status (Google Patents, checked November 10, 2026) | Active; adjusted expiration July 7, 2027 |
The patent starts from a set of seed pages chosen as high quality. It then measures distance through the web-link graph and can use the distance to each page’s k nearest seeds. Distance can be measured along incoming or outgoing links. A page with a large out-degree can be treated as farther away, because its vote is spread across many destinations.
The useful idea for link building is simple: a placement on a site that sits close to independently trusted parts of the web is a different kind of vote than a link from an isolated or unknown neighborhood. When we choose outreach targets, I would rather earn one link on a site with a real audience and a clear place in its industry than a stack of links from sites that only exist to link out.
What this means for your link building. Prioritize publishers that would be trusted even if they never linked to you. Distance in a link graph is not the same thing as a Domain Rating number in a SaaS report, and this patent is not proof of how Google scores sites today. It is a reminder to ask how close a prospect sits to the web’s known, trusted pages before you chase the placement.
Historical Data and Link Timing
Not only where a link comes from matters. US 7,346,839, titled “Information retrieval based on historical data,” describes using the history of a document and its links as ranking signals.
| Detail | Value |
|---|---|
| Patent | US 7,346,839 |
| Title | Information retrieval based on historical data |
| US inventors | Anurag Acharya, Jeffrey Dean, Paul Haahr, Monika Henzinger, Urs Hoelzle, Steve Lawrence, Karl Pfleger, Olcan Sercinoglu, and Simon Tong |
| Filed | December 31, 2003 |
| Granted | March 18, 2008 |
| Status (Google Patents, checked November 10, 2026) | Expired – Lifetime; adjusted expiration October 18, 2025 |
Matt Cutts is sometimes named with this patent. He is listed on the Canadian counterpart, CA 2540573, not on the US grant above.
Among the historical signals the US patent discusses:
- The rate at which a page gains inbound links, including a recent window compared with the period since the page was first found. One illustration uses two documents first found 100 days earlier, and compares a case where 10% of the links were found in the last 10 days with a case where none were.
- Sudden changes in linking, including around news events.
- A change in topic after an expired domain changes hands.
- Domain registration length.
- Anchors that arrive in a coordinated burst, whether the wording is identical or deliberately varied.
- Weaker evidence from pages with little editorial judgment, such as guest books, referrer logs, and free-for-all link pages.
What this means for your link building. From an agency desk, a campaign that drops a large batch of links in one week and then goes quiet looks different from a campaign that earns placements steadily. I would rather sell work that can be sustained than a one-time spike. Expired domains that now cover a different topic are also a reason to be careful about the neighborhoods you buy into. None of this is a promise about how Google times credit today. It is a reason to prefer durable, editorial placements over short bursts.
Affiliated Sites and Node Independence
A group of sites can look like many votes when they are, in practice, one source. Two Google patents describe ways to treat related nodes as less independent than they appear.
The first is US 8,719,276, titled “Ranking nodes in a linked database based on node independence.”
| Detail | Value |
|---|---|
| Patent | US 8,719,276 |
| Title | Ranking nodes in a linked database based on node independence |
| Inventors | Paul Haahr, Michael Kaszkiel, and Amit Singhal |
| Priority | Provisional 60/519,271, filed November 13, 2003 |
| This grant filed | January 4, 2011 (as a continuation) |
| Granted | May 6, 2014 |
| Status (Google Patents, checked November 10, 2026) | Expired for nonpayment, effective May 6, 2026 |
This patent describes identifying affiliated nodes, for example through ownership or WHOIS-style records, and then reducing the effect of that cluster. Two methods it discusses are dividing a vote by the size of the cluster, or keeping only the strongest vote from the group. It also discusses capping the influence of a node that is already treated as a trusted authority.
The second is US 7,783,639, titled “Determining quality of linked documents.”
| Detail | Value |
|---|---|
| Patent | US 7,783,639 |
| Title | Determining quality of linked documents |
| Inventors | Krishna Bharat, Amit Singhal, and Paul Haahr |
| Filed | June 30, 2004 |
| Granted | August 24, 2010 |
| Status (Google Patents, checked November 10, 2026) | Expired; lapsed, effective August 24, 2022 |
This patent describes inferring affiliation from the link graph, traffic, hostnames, and IP addresses, including the leading two or three octets of an IP address. A later family member, US 8,825,645, was granted on September 2, 2014.
What this means for your link building. A handful of sites under one owner, on related IPs, or with the same registration trail can be treated as one source, or have their vote split, rather than counted as many independent endorsements. I do not build or buy links from obvious private networks, and I would rather walk away from a placement than add a client to a neighborhood that looks coordinated. Independence is the point, not the raw number of domains.
Link-Spam Detection
Some sites earn links from many unrelated places. Others exist mainly to point at a target. US 7,509,344, titled “Method for detecting link spam in hyperlinked databases,” describes a mathematical way to tell those patterns apart.
| Detail | Value |
|---|---|
| Patent | US 7,509,344 |
| Title | Method for detecting link spam in hyperlinked databases |
| Inventors | Sepandar D. Kamvar, Taher H. Haveliwala, and Glen M. Jeh |
| Priority | Provisional filed August 18, 2003; application filed August 18, 2004 |
| Granted | March 24, 2009 |
| Status (Google Patents, checked November 10, 2026) | Expired – Lifetime; adjusted expiration March 25, 2026 |
The system looks at how a page’s importance score, such as PageRank, changes with respect to a coupling factor. In plain language, it asks how sensitive a page’s score is to the tight coupling of the graph around it. The patent describes link-farm targets showing a large negative normalized derivative, and web rings showing a large positive one. Later US applications in this family were filed in 2009 and 2011.
What this means for your link building. A tight cluster of sites that exist to pass links is a different shape from a popular site that earned mentions from many independent publishers. I avoid farms, rings, and packaged networks that all point at the same targets. If a prospect’s only notable activity is linking out, that is usually enough reason to skip it.
What an Anchors Patent Actually Describes
SEOs often look for a “safe” mix of keyword-rich anchor text. That question is real in day-to-day work, but it is not what US 8,959,093 is about. The patent, titled “Ranking search results based on anchors,” describes something else.
| Detail | Value |
|---|---|
| Patent | US 8,959,093 |
| Title | Ranking search results based on anchors |
| Inventor | Ajay P. Nerurkar |
| Filed | March 15, 2010 |
| Granted | February 17, 2015 |
| Status (Google Patents, checked November 10, 2026) | Active; adjusted expiration November 2, 2030 |
The system can pass quality-of-result or user-behavior statistics from a first document, a search result that acts as a secondary source, through an anchor to a second document being ranked, the primary source. In other words, an anchor can carry implicit quality and engagement signals from the page that hosts the link, not just the words in the link itself.
This patent does not set a safe percentage of exact-match anchors, and it does not define official ratios for branded, naked, or generic text. Unofficial ratio tables you may have seen in SEO guides are industry rules of thumb, not claims in this patent, and I am not treating them as Google’s rules.
What this means for your link building. I still write anchors that a person would click, and I still avoid stuffing the same keyword phrase across a campaign, because that is sound editorial practice. I do not sell a magic ratio. The more useful reading of this patent is that the page you get the link from, and how people use that page, can matter as much as the wording of the anchor.
Brand Searches and Clicks
Links are not the only votes these patents describe. Two others look at whether people search for a site by name, and at how they behave after they click a result.
The first is US 7,962,462, titled “Deriving and using document and site quality signals from search query streams.”
| Detail | Value |
|---|---|
| Patent | US 7,962,462 |
| Title | Deriving and using document and site quality signals from search query streams |
| Inventors | John Lamping and Mark Pearson |
| Filed | May 31, 2005 |
| Granted | June 14, 2011 |
| Status (Google Patents, checked November 10, 2026) | Expired – Lifetime; adjusted expiration September 19, 2025 |
The patent describes queries that “ask” for a particular site or page. Signals it discusses include similarity between the query and a result’s title, URL, or prominent text; a majority of matching-anchor links; or a majority of users selecting the same result. Later US applications in this family were filed in 2012 and 2014.
The second is US 8,661,029, titled “Modifying search result ranking based on implicit user feedback.”
| Detail | Value |
|---|---|
| Patent | US 8,661,029 |
| Title | Modifying search result ranking based on implicit user feedback |
| Inventors | Hyung-Jin Kim, Simon Tong, Noam M. Shazeer, and Michelangelo Diligenti |
| Filed | November 2, 2006 |
| Granted | February 25, 2014 |
| Status (Google Patents, checked November 10, 2026) | Active; adjusted expiration January 31, 2031 |
This patent describes using implicit feedback from clicks, including the difference between long and short clicks and a long-click fraction (LC|C), with safeguards against click spam. Later US applications in the family were filed through at least 2021.
What this means for your link building. Earning the link is not the whole job. These patents describe brand searches and post-click behavior as quality signals too. In practice, that means a campaign still depends on the client being a name people look for and a result people stay with. I will not promise that more branded searches or longer clicks will move a ranking. I will say that link building works better when the site behind the links is worth finding.
The Honest Caveat
A patent shows what Google invented and chose to protect. It does not prove that any exact system described here runs in Google’s algorithm today, and Google has not confirmed that these patents do.
Some of these grants remain active. Others have expired, by term or for nonpayment. Google also filed later applications in several of the families. For the reasonable surfer patent, later US applications were filed in 2010, 2012, and 2016, including US 9,305,099, granted April 5, 2016. I am not treating unverified later grant years as facts.
Some of the ideas still line up with Google’s public guidance. For example, Google’s spam policies name white text on a white background and links hidden behind a single character as hidden link abuse. The reasonable surfer patent lists link color and background color as features a model can measure.
Treat these patents as a record of how Google has thought about links, not as a rulebook and not as a ranking guarantee.
Conclusion
The reasonable surfer patent comes down to one idea: not every link on a page counts the same. Google described a system that weights each link by how likely a real person is to click it, based on where it sits, how it looks, what it says, and what surrounds it.
The related patents in this guide add other questions you can ask about the same placement. How close is the linking site to trusted parts of the web? When did the links arrive, and did they arrive in a burst? Are the linking sites truly independent? Does the neighborhood look like a farm or a ring? What signals might travel with the anchor? Do people already search for the brand, and do they stay when they land?
For link builders, the practical lesson is the same one I use in campaigns. Build links that people would actually use, on independent sites, at a pace you can sustain, with anchors that read naturally. Put them in the body of the content, early, on relevant pages that are not crowded with other links.
If you want to go deeper on why Google slowed ranking feedback, read The Problem Google Was Solving: The Spammer’s Feedback Loop. For the basics of how we think about placements, start with Link Building 101.
Want links placed where real readers will click them? Talk to Link Builders about your next campaign.
Sources
- US 7,716,225: Ranking documents based on user behavior and/or feature data, Google Patents
- US 9,305,099: Ranking documents based on user behavior and/or feature data, Google Patents
- US 9,165,040: Producing a ranking for pages using distances in a web-link graph, Google Patents
- US 7,346,839: Information retrieval based on historical data, Google Patents
- CA 2540573: Information retrieval based on historical data, Canadian counterpart, Google Patents
- US 8,719,276: Ranking nodes in a linked database based on node independence, Google Patents
- US 7,783,639: Determining quality of linked documents, Google Patents
- US 8,825,645: Determining quality of linked documents, Google Patents
- US 7,509,344: Method for detecting link spam in hyperlinked databases, Google Patents
- US 8,959,093: Ranking search results based on anchors, Google Patents
- US 7,962,462: Deriving and using document and site quality signals from search query streams, Google Patents
- US 8,661,029: Modifying search result ranking based on implicit user feedback, Google Patents
- The Anatomy of a Large-Scale Hypertextual Web Search Engine, Sergey Brin and Lawrence Page, Stanford University
- Spam policies for Google web search, Google Search Central
About the author
Justin Davis is the founder of Link Builders, which offers link building services to companies and white label link building to agencies, and has been building links since 2015 for more than 150 clients in law, healthcare, real estate, software and other industries. After years of doing prospecting research by hand, he built an AI agent link building system that handles each step, from finding link opportunities and vetting sites to finding contacts and preparing outreach. He now works every project himself, running that system for agencies and in-house SEO teams, earning the links and mentions that help their clients show up in Google and in AI answers, and teaches it on YouTube. He also built AnswerBuilders, the AI visibility platform Link Builders uses to find the pages AI answers cite. Find Justin Davis on YouTube, LinkedIn and X.







