For more than 25 years, how Google structures the web has been one of the internet's greatest unsolved mysteries.
Google is a gateway to the internet that many businesses rely on, yet its constantly evolving algorithms remain closely guarded behind lock and key.
This week, the black box was finally opened.
A trove of 2,500 documents containing highly-anticipated secrets about how Google ranks search result pages began circulating among a small group of SEO experts and was shared widely on Monday, after the company confirmed the materials were authentic.
The already enthusiastic search engine optimization (SEO) community became even more excited, with social media sites and industry forums buzzing about the treasure trove.
The frenzy quickly turned to rage, with some SEO experts saying the document showed that Google hadn't always been honest in answering questions about how it ranks websites.
“This is a new level of competition between SEOs and Googlers,” said Lily Ray, vice president at SEO agency Amsive.
Erfan Azimi, CEO of SEO agency EA Eagle Digital, claimed to have first discovered the document online, and published a dramatic 13-minute YouTube video. For Azim and many others in the SEO community, some of the leaked information seemed to confirm their suspicions: that Google may not have been completely honest about the most important signals that determine which sites appear in the top half of its search engine results pages.
“We've been lied to for over 10 years,” Azimi said, staring into the camera lens. “The truth needs to come out.”
And yet even the most dedicated SEO code crackers have yet to figure out how up-to-date the information is, or which of the supposed 14,000 ranking factors have seen the light of day.
A Google spokesman said the documents lacked context and that the company's systems are subject to frequent change. The company declined to comment on specific areas of the data.
“We want to caution people against making inaccurate inferences about searches based on out-of-context, out-of-date or incomplete information,” a Google spokesperson said in a statement. “We share extensive information about how search works and the types of factors our system weighs, and we also work to protect the integrity of search results from manipulation.”
The leak has further stoked distrust in a Google that is preparing to rewrite the rulebook. With the company promising to “Google you” with AI-powered summary generation, many website owners are preparing for a future in which the company siphons off their content and instead sends them no visitors at all.
“AI is taking the world by storm, but does anyone know how it works?,” said Gareth Hoyle, managing director of marketing firm Marketing Signals. “Who's going to guard the security guards?”
Why Google keeps its searches secret
Google employees are under strict instructions to keep quiet about search: An internal presentation to employees revealed last year during Google's Justice Department search antitrust trial instructed staff to keep discussion of the company's most valuable product “to the bare minimum.”
“Any information we leak will be used against us by SEOs, patent trolls, competitors, etc.,” the document states. “The search issues could infuriate world leaders with power over Google, potentially triggering demands for congressional hearings, etc.,” the document continues.
Here's what we know: At the most basic level, Google uses web crawlers – bots that read websites, map their link structure, and track various keywords. These crawlers are designed to ensure that Google search results return the most relevant and up-to-date information to users.
Other mysteries, such as how Google determines what content is “good” or “useful,” where keywords should be placed, and how high links should appear on a web page, are constantly evolving. In the world of SEO, practitioners conduct rigorous testing, trade tips and theories at conferences, and press Google officials and their dedicated “public search liaisons” about which ranking factors should be most important. For some SEOs, this document suggests they would have been better off sticking to their assumptions.
Measure clicks. SEO experts have long believed that Google ranks websites by analyzing when and how often they are clicked on. The leaked documents use the terms “goodClicks” and “unsquashedClicks,” which SEO experts believe may indicate that Google places more importance on clicks than it has previously made public.
“One thing I've learned from all of this is that Google actually uses click data a lot more than we thought,” said Grace Froehlich, an SEO consultant at digital marketing firm BrainLab.
The document also mentions the symbols “isElectionAuthority” and “isCovidLocalAuthority,” suggesting that Google may rank certain sites with more authority on these topics.
Then there's domain authority, which is a rating of a site's quality and trustworthiness on related topics. Google has said so far that it doesn't use domain authority as a ranking factor, but its documentation does mention a factor called “siteAuthority.”
Or take Google's Chrome browser: The company has previously said it doesn't use browsing data collected by Chrome to rank websites. But the document's multiple references to Chrome have SEO experts convinced that Google does in fact use the popular browser to rank websites. (Google is understandably cautious on this point, given regulators' scrutiny of its possible use of self-prioritizing tactics to boost its search and advertising business.)
“The big picture just highlights the areas where we were right and Google told us we were wrong,” said Michael King, founder and CEO of digital marketing firm iPleRank, who was one of the first to analyze the documents on his blog.
Some in the SEO community are wary of overinterpreting the leak. Aleyda SolÃs, an SEO consultant and founder of SEO firm Orainti, warned that some people may find what they want in the documents, and that it's unclear how Google “weights” factors like clicks and other values.
“I don't even know if all of this is taken into account as an actual ranking factor,” Solis said.
“We are already in a precarious position.”
SEOs and Google's relationship was already frosty. Following two major updates to Google's search algorithm within a few months, some business owners have reported devastating declines in website traffic, while sites like Reddit and Quora have become inundated at the top of search results pages.
Google's cuts have also reduced the number of human resources SEOs have access to. While Google throws lavish parties for its advertising clients, such as the star-studded YouTube Brandcast, it hasn't made the same investment in events for the SEO community. That has some in the SEO community lamenting a breakdown in the search giant's relationship with the experts who helped it sort through it all.
“We already have a very precarious relationship with them,” Amshiv's Ray said.
This comes as Google pushes ahead with AI search generation. A recent test of AI-generated search result summaries in the US drew ridicule for suggesting eating rocks for nutrition or using glue to stick cheese to pizza, referencing satirical websites and Reddit posts. Google's response was met with a similarly dubious reaction from the search community. Google initially claimed that its AI was only spitting out answers to uncommon queries, but later said it was “taking swift action” to manually remove inappropriate answers that violate its content policies.
While the search leak won't dramatically change how websites compete with Google, and it won't necessarily reflect how Google currently ranks the web, SEOs will be watching closely to see whether rules from the document apply to the new world order of AI search.For example, the document suggests that Google is on an “inevitable path” to direct traffic to the websites of big brands over smaller publishers, wrote Rand Fishkin, co-founder and CEO of audience research firm SparkToro.
Eric Huber, SEO director at digital agency Jellyfish, said the leak confirmed that quality content will always win over attempts to game the algorithm.
“That's pretty much the same with generative AI,” Huber says.
For now, Google still dominates the search market, and SEOs have plenty of time to crack the code in the now-fully public reams of documentation. They just can't count on anyone inside Google to help them.
“I think ultimately it's going to make the correlation studies that we're doing in this space better,” King said, “but I also think it might make Google talk to us less.”