This week, a massive leak of over 2,500 internal Google documents has provided an unprecedented glimpse into the inner workings of the company’s search algorithm. Although the complete algorithm itself did not leak, the documents revealed significant insights into Google’s processes and raised questions about the accuracy of Google’s public statements.
One major revelation from these documents is the suggestion that Google may have misled the public about how it ranks content. For years, Google representatives have denied that user clicks impact search rankings. However, the leaked documents indicate that various types of user clicks do indeed influence ranking, contradicting Google’s public denials. For instance, the documents mention a ranking factor called Navboost that uses searchers’ clicks to elevate content.
“To me, the larger, meta takeaway is that even more of Google’s public statements about what they collect and how their search engine works have strong evidence against them,” Rand Fishkin, a veteran of the search engine optimization (SEO) industry, told The Verge via email.
SEO experts Rand Fishkin and Mike King, who initially shared the leaked information, emphasized that the documents provide evidence against many of Google’s public statements regarding data collection and search rankings. Google confirmed the authenticity of the documents but cautioned against drawing conclusions based on potentially outdated or incomplete information.
The leaked documents include repositories filled with definitions of data Google collects, some of which may inform how webpages are ranked. However, the documents do not reveal how different attributes are weighted, leaving uncertainty about their actual impact on search rankings. Additionally, some attributes mentioned in the documents, such as identifiers for small personal sites or demotions for product reviews, may no longer be in use or may have never been implemented.
“We don’t necessarily know how [the factors named] are being used, aside from the different descriptions of them. But even though they’re somewhat sparse, there’s a lot of information for us,” King says. “What are the aspects that we should be thinking about more specifically when we’re creating websites or optimizing websites?”
One contentious topic among SEO experts has been whether Google uses data from its Chrome browser for search rankings. Google has consistently denied this, but the documents suggest otherwise. For example, they mention “chrome_trans_clicks,” which could imply that the number of clicks on pages in Chrome influences which links appear below the main webpage in search results.
The documents also refer to “Twiddlers,” which are ranking tweaks deployed outside major system updates to adjust content rankings based on specific criteria. Elements like the author’s identity and the authority of websites are also mentioned, though there is a noticeable lack of information about AI-generated search results.
For the SEO industry, this leak represents a treasure trove of information that will likely lead to various experiments and adjustments in website optimization strategies. Publishers, e-commerce companies, and businesses will closely examine the leaked details to refine their approaches to improving search rankings. This could result in changes to how websites look, feel, and function as industries try to make sense of the new information.
Fishkin argues that journalists and publishers should adopt a more critical stance towards Google’s public statements. He believes that uncritically repeating Google’s claims helps the company maintain a narrative beneficial to itself but not necessarily to practitioners, users, or the public. This leak underscores the need for greater scrutiny and skepticism regarding the search giant’s claims about its operations and the factors influencing its search engine rankings.
“Journalists and publishers of information about SEO and Google Search need to stop uncritically repeating Google’s public statements, and take a much harsher, more adversarial view of the search giant’s representatives,” Fishkin says. “When publications repeat Google’s claims as though they are fact, they’re helping Google spin a story that’s only useful to the company and not to practitioners, users, or the public.”
